USD/Month</strong></th>	vCPU</strong></th>	Memory</strong></th>	Storage</strong></th>	Traffic</strong></th>	Provider</strong></th></tr></thead>
$3.41</td>	1</td>	2GB</td>	20GB</td>	20TB</td>	Hetzner</td></tr>
$4.77</td>	2</td>	2GB</td>	40GB</td>	20TB</td>	Hetzner</td></tr>
$5.00</td>	1</td>	1GB</td>	25GB</td>	1TB</td>	DigitalOcean</td></tr>
$6.70</td>	2</td>	4GB</td>	40GB</td>	20TB</td>	Hetzner</td></tr>
$10.00</td>	1</td>	2GB</td>	25GB</td>	2TB</td>	DigitalOcean</td></tr> </tbody></table> Check out the math on that, seriously. Hetzner’s base tier offers 1 vCPU, 2GB of ram, 20GB of SSD, and 20TB of data transfer for USD$3.41 per month. That’s double the ram and 20 times the transfer for almost half the price</strong> of my current DigitalOcean droplet. The fourth down in the table above, the one I plan on purchasing, costs USD$1.70 more than DigitalOcean and has quadruple the performace</strong> of my current VPS. That’s more than even Docker</em> could need. Ok that might be a slight exaggeration, but it’s still a lot.</p> Hetzner is a green hosting provider</a>, as certified by the Green Web Foundation</a>. DigitalOcean is not. Obviously, climate change is important. If we look at it objectively, it’s one of the only extinction-level threats to human existence at the moment. All that is to say that switching to a green hosting provider is the only thing I need to do</a> according to this online website carbon calculator</a>, and you know I can’t resist online website metrics.</p> Anyways, that’s a pretty good summary of my thoughts on the move. I’m probably going to use the move to make a few other changes to my hosting setup (i.e. `docker-compose</code> to` `k3s</code></a>, umami</code></a> to plausible</code></a>). I’ll keep you all abreast of the latest updates to my hosting in future posts.</p>` `Until then, this is FIGBERT signing off. I should make a more consistent outro.</p>` MacOS Control Center is Better 2020-10-07T00:00:00+00:00 I’ve been using the macOS Big Sur Beta for a couple of months now. So far it’s pretty fantastic, and a surprisingly smooth experience over all. One brand new feature in this update is the introduction of the Control Center. A close relation of its iOS counterpart, they differ in one key way: Control Center on macOS actually turns off services.</p> </span> Your browser doesn't support the video tag, which I use in place of .gifs, and/or the video formats in use here – sorry! </video> Did you catch that? It’s pretty subtle. When you try and turn off Bluetooth or Wifi on iOS, they aren’t actually turned off: they’re just temporarily disabled. The interface still broadcasts, it just doesn’t accept new connections. It’s one of my biggest pet peeves, and I’m glad Apple got it right with macOS. Now if they could just port it to iOS…</p> Going Full Static with Zola 2020-08-22T00:00:00+00:00 Those of you who read my last “I Wrote This” post</a> will know that I was having some trouble with my website. My site was coded using Sapper</a>, a Svelte</a>-based web-app framework I had been using for some time. I had chosen to use Sapper because it allowed me to stay as close to the web-metal as possible, while still letting me do some fancy things like use components, scoped CSS, and server routes</a>. However, after diving deeper into website tests and statistics, I started noticing that my “static” site had a lot more moving parts than I thought. The HTML was crammed full of inline scripts and blob://</code>s, tanking performance, wreaking havoc on my CSP</a>, and breaking the site for people with scripts disabled. I decided to move the site to Zola</a>, a ludicrously simple static site generator made in Rust. Feel free to check out the source code here</a>.</p> </span> I decided I was going to rewrite my site because I have a bad habit of rewriting everything all the time</del> largely because of Sapper’s underwhelming response to this Github issue</a>, which proposes a “strict export” for Sapper sites to remove inline scripts and use of eval()</code>. I think this is a great idea, but it unfortunately has not received much attention (though it appears that as I’m writing this, it has been added to a “Roadmap Triage” project board). I started a new branch and began working to translate my site to Sapper’s main competitor, Routify</a>. Sapper and Routify are not the same thing, but for me they both would serve well enough. After around two days, I had a working MVP of my site in Routify.</p> EDIT:</strong> Sapper is being retired in favor of SvelteKit</a>, which has the same issue</a>. smh.</p> Then disaster struck: I got a bunch of emails from Github. A series of high priority security vulnerabilities had been found in dependencies used by basically all of my web projects. I spent a day force-updating all the dependencies of my web projects – a bit of a pain because npm refuses to natively upgrade breaking changes – and decided to stay as far away from the Javascript ecosystem as possible. I hate that when I install a JS framework, a fundamental tool of modern web development, I install a million other dependencies that could, and often do, have critical security vulnerabilities. I’m thankful that Dependabot caught these ones, but it really killed my enthusiasm for using any JS framework on my site – which means Routify was out of the picture.</p> I decided to use a static site generator</a>. I’d heard of many of the big boys in the past, like Hugo</a>, Jekyll</a>, and Eleventy</a>, but they all had their own problems when I looked at them in the past. Hugo has god-awful templating syntax, Jekyll is Ruby-based and I don’t know Ruby, and Eleventy isn’t even an escape from Javascript! So I decided to use Zola</a>, a “one-stop static site engine.” Zola is made in Rust, so it’s super fast, and it’s designed to be dead simple. Seriously: the CLI has only five commands, everything is configured from one .toml</code> file, and your content is all written in “Augmented</a> Markdown</a>.”</p> The interesting thing is that there’s honestly not much more to the story because of how easy and simple Zola is to use. All of my posts and projects go into the content</code> directory, my CSS, favicon, and miscellaneous files (non-content related stuff like emojis and public keys) go in the static</code> directory, and templates and shortcodes go into the templates</code> directory. If I was using a theme, it’s files would go into a theme</code> directory.</p> My slow Python script to convert Markdown posts to Svelte (which was perfect at first but I then packed full excess tests and sandboxing) is gone. Zola handles that automatically.</li> I got rid of TailwindCSS, and replaced it with custom styles. It’s actually pretty fun to write simple custom CSS, especially with modern tools like variables.</li> Writing new posts is ludicrously easy now. I write a post in Markdown, throw any images or videos used in the same directory</a>, and publish.</li> Zola comes with a whole bunch of features built-in that I didn’t have before, like syntax highlighting and anchor links (the latter of which I have yet to set up). Other things are just handled automatically, like feed generation or i18n.</li> Build times are much faster. Exporting with Sapper wasn’t slow, but it didn’t feel instant. Zola does.</li> </ul> You sacrifice a certain amount of control by using a static site generator, like link</a> properties</a>. You could solve this with shortcodes</a>, or by contributing to the project (which I plan to do).</li> I mean that’s really it to be honest.</li> </ul> EDIT:</strong> You can now add attributes like noreferrer to links automatically with Zola… So I guess there’s no more drawbacks? Yeah that feels right.</p> I’m really happy with using Zola, and I look forward to continuing to work with it in the future. I want to publish my blog’s styles and templates as a Zola theme</a>, but I have to iron out a few kinks (like anchor links, which are still a bit finicky on my end) before that.</p> EDIT:</strong> I have since moved on from this theme on my own site, but you can still use the theme as d3c3nt</a>, a simple, clean, and flexible theme for personal sites.</p> I also have yet to re-implement a bunch of the indie-web features and GoatCounter</a> analytics of my old site into this version. Overall though, I think it’s been a really fun and productive experiment using Zola, and I’d highly recommend using it for anybody looking for a great, no-nonsense static site generator.</p> EDIT:</strong> I fixed anchor links, and I decided to forgo analytics altogether. My site’s better off without JS.</p> Until next time, FIGBERT.</p> Problem Exists Between Keyboard and Chair: How I Spent 2 Days Chasing a Bug that Didn't Exist 2020-07-28T00:00:00+00:00 Post-WWDC2020</a>, I decided to rewrite the backend of txtodo in SwiftUI using the new App and Scene structure</a>. Rebuilding the app from scratch may have not been the best choice, but during that process I have massively simplified the app’s data structure, despaghettified some messy UI code, and spent two full days trying to solve a problem that didn’t exist. This is the story of that last bit.</p> </span> The new app, so far, was mostly the same as the old version but without the AppDelegate.swift</code> or SceneDelegate.swift</code> files (using the new XCode 12 multiplatform app template). I also combined the Core Data FloatingTask</code> and DailyTask</code> entities into one Task</code> entity. By this point, everything was running well enough so I started to migrate more code into the new codebase starting with the fetch request:</p> @</span>FetchRequest</span>(</span></span> entity</span>:</span> Task.</span>entity</span>(</span>)</span>,</span></span> sortDescriptors</span>:</span> [</span></span> NSSortDescriptor</span>(</span>keyPath</span>:</span> \Task.</span>completed</span>, </span>ascending</span>:</span> true</span>)</span>,</span></span> NSSortDescriptor</span>(</span>keyPath</span>:</span> \Task.</span>priority</span>, </span>ascending</span>:</span> false</span>)</span>,</span></span> NSSortDescriptor</span>(</span>keyPath</span>:</span> \Task.</span>name</span>, </span>ascending</span>:</span> true</span>)</span></span> ]</span></span> )</span> var</span> tasks:</span> FetchedResults</span><</span>Task</span>></span></span></code></pre> Tasks are displayed as TaskView</code>s in a ForEach</code> loop on the homescreen, which is simple enough. The TaskView</code> struct, however, is relatively complicated. The purpose of TaskView</code> is to represent and manipulate a single Task</code>. In the previous version of the app (I’m going to call the original version 2.0 and the rewrite 3.0 from now on), this involved passing a number of attributes individually to be manipulated as the view’s @State</code>. When migrating the view, I reduced this to a single @ObservedObject</code>. I also removed some of the text styling, which I planned to port over after I got the UI functional.</p> I ran the app on my device, and this happened:</p> Your browser doesn't support the video tag and/or the video formats in use here – sorry! </video> Well that was unexpected. Instead of checking off the tasks I selected, tasks were checked off starting from the bottom and ascending – obviously not the intended behavior! My first thought was that it was caused by the use of @ObservedObject</code> to declare the view’s task property – I haven’t seen it used to manipulate a Core Data entity before, but it’s worked fine so far in txtodo – so I rewrote the variables to match version 2.0.</p> //</span> VERSION 3.0</span></span> struct</span> TaskView</span>:</span> View</span> {</span></span> @</span>Environment</span>(</span>\.</span>managedObjectContext</span>)</span> var</span> managedObjectContext</span></span> @</span>ObservedObject</span> var</span> task:</span> Task</span></span> @</span>State</span> var</span> priority:</span> Int</span></span> @</span>State</span> private</span> var</span> config </span>=</span> TaskConfig</span>(</span>)</span></span> //</span> UI...</span></span> }</span></span> </span> //</span> VERSION 2.0</span></span> struct</span> floatingTaskView</span>:</span> View</span> {</span></span> @</span>Environment</span>(</span>\.</span>managedObjectContext</span>)</span> var</span> managedObjectContext</span></span> @</span>ObservedObject</span> var</span> task:</span> FloatingTask</span></span> @</span>State</span> var</span> completed:</span> Bool</span></span> @</span>State</span> var</span> name:</span> String</span></span> @</span>State</span> var</span> priority:</span> Int</span></span> @</span>State</span> var</span> deleted:</span> Bool</span> =</span> false</span></span> @</span>State</span> private</span> var</span> editingText:</span> Bool</span> =</span> false</span></span> @</span>State</span> private</span> var</span> editingPriority:</span> Bool</span> =</span> false</span></span> @</span>State</span> private</span> var</span> viewingNotes:</span> Bool</span> =</span> false</span></span> @</span>State</span> private</span> var</span> confirmingDelete:</span> Bool</span> =</span> false</span></span> //</span> UI...</span></span> }</span></span></code></pre> Still no change. It was getting pretty late at this point, but I decided to stick it out for just a bit longer. I rewrote the TaskView</code> struct from scratch two more times</em> to no avail. Something was wrong, but I had no idea where it was and there was no way I was going to figure it out at two in the morning by coding it again the exact same way.</p> The next morning, I took a look at the code again. If the problem wasn’t in TaskView</code>, where was it? The only other thing in the UI was the button to make a new task, which looked something like this:</p> Button</span>(</span>action</span>:</span> {</span></span> let</span> newTask </span>=</span> Task</span>(</span>context</span>:</span> self</span>.</span>managedObjectContext</span>)</span></span> newTask.</span>name</span> =</span> "</span>test</span>"</span></span> newTask.</span>priority</span> =</span> 3</span></span> newTask.</span>notes</span> =</span> [</span>String</span>]</span>(</span>)</span> as</span> NSObject</span></span> newTask.</span>id</span> =</span> UUID</span>(</span>)</span></span> newTask.</span>date</span> =</span> Date.</span>init</span>(</span>)</span></span> newTask.</span>daily</span> =</span> true</span></span> do</span> {</span></span> try</span> self</span>.</span>managedObjectContext</span>.</span>save</span>(</span>)</span></span> }</span> catch</span> {</span></span> print</span>(</span>error.</span>localizedDescription</span>)</span></span> }</span></span> }</span>)</span> {</span></span> Text</span>(</span>"</span>Add</span>"</span>)</span></span> }</span></span></code></pre> Some of you may have figured it out by this point. At the time, I was still confused – this was the exact method I was using in my previous app, but with preset values – how could it be broken? I modified the generation slightly so I could tell the difference between tasks, and hopefully get to the bottom of the issue:</p> let</span> newTask </span>=</span> Task</span>(</span>context</span>:</span> self</span>.</span>managedObjectContext</span>)</span></span> newTask.</span>name</span> =</span> String</span>(</span>UUID</span>(</span>)</span>.</span>uuidString</span>.</span>prefix</span>(</span>Int</span>.</span>random</span>(</span>in</span>:</span> 5</span>..<</span>9</span>)</span>)</span>)</span></span> newTask.</span>priority</span> =</span> Int16</span>.</span>random</span>(</span>in</span>:</span> 1</span>..<</span>4</span>)</span></span> newTask.</span>notes</span> =</span> [</span>String</span>]</span>(</span>)</span> as</span> NSObject</span></span> newTask.</span>id</span> =</span> UUID</span>(</span>)</span></span> newTask.</span>date</span> =</span> Date.</span>init</span>(</span>)</span></span> newTask.</span>daily</span> =</span> Bool</span>.</span>random</span>(</span>)</span></span></code></pre> I ran the app again and saw this:</p> Your browser doesn't support the video tag and/or the video formats in use here – sorry! </video> The tasks weren’t being marked off in ascending order. They were being moved to the bottom automatically when marked as complete, which I couldn’t see because a) all the tasks were identical and b) there were no animations to indicate that was happening. They were sorted by the FetchRequest</code> with a NSSortDescriptor</code>, to make sure that the unfinished tasks are the first thing the user sees. The “glitch” I had spent two days chasing down was entirely by design, and I had just forgotten.</p> There were two main things I learned from this experience. First, it’s incredibly important to be able to take breaks. The difference between spending two days trying to fix a non-existent glitch and realizing it’s a feature you implemented could be as simple as a nap – it was for me. Secondly, your test and placeholder data is more significant than you might think: garbage in, garbage out</a> definitely applies here. If all your test data is the same, your tests are not good tests.</p> To make the sorting more clear, I randomized the tasks’ priority, name, and category (as seen above) and added an animation with .animation(.easeIn(duration: 0.25))</code>. The current prototype looks something like this:</p> Your browser doesn't support the video tag and/or the video formats in use here – sorry! </video> This has been a really fun blog post to write! A got a big laugh out of this bug chase, and I hope you’ve enjoyed reading it.</p> Till next time, FIGBERT</p> I Wrote This #3 2020-07-04T00:00:00+00:00 It appears I’m not that great at keeping a weekly schedule. Right after I published my last post, I started a class on text adventures and have been living and breathing in Inform7</a> ever since. I’ve also spent some time working on this site – though I’ve got even bigger changes coming in the future – learning a few new languages, and listening to a whole lot of music. I think something happened in cybersec</a> too? Oh also I learned how to drive.</p> </span> These past few weeks, my main work has been on interactive fiction as part of a summer class for my school. The first week was spent learning the history of text adventures (as I will refer to them for the rest of the article, becuase calling them “interactive fiction” makes me sound like a dick</del> a snob). I gathered a collection of games to play in my free time (if I ever have any more of that), from the very first of the genre – Adventure – to modern ones like Lost Pig.</p> </p> In the second week, we all made our own text adventures. It was a really interesting experience – drastically different than any other programming I’ve done. The main thing I struggled with was the semi-NLP style of Inform7</a>. Rather than telling the computer what to do, I felt like I was making suggestions. I also struggled with the editor a bit. Inform7 isn’t open source, which means there’s only one IDE you can use to program in the language. That would be fine if the IDE was high quality, but I often found it really bad and draining</del> mediocre. This was made a little better after I changed the editor colors to be much more vibrant so I wouldn’t fall asleep</del>, but I still felt like it needed some work. Regardless of any struggles along the way, the payoff after the game was finished was huge. Watching my little sister play a text adventure for the first time, one I had made</em>, was awesome.</p> Without further ado, I present to you: One Angry Wizard, or the Tentacular Adventures of our Brave Hero through a World Most Strange and Foreign</em>. You can download the .gblorb here</a>.</p> I made some changes that you guys can see, and some other ones behind the scenes.</p> I first encountered webrings on Hacker News</a>, where I find most of my interesting internet things. From there, I decided to join both the geekring</a> and the Hotline Webring</a> – you can find links to the webrings in the nav on every page. Joining the Hotline Webring was designed to be supremely easy, but the geekring was a little more complicated. I’ve never been a big IRC person (though not for lack of trying), and I ended up joining via the HTML form and getting my key via email. I may or may not have confused my number with my private-ish key, and had to use a little bit of bfg</a> magic to fix that.</del></p> EDIT:</strong> I would now very much consider myself a big IRC person.</p> My site was previously hosted using a continuous deploy on Netlify</a>. This was a pretty great system for me – all I had to do was commit my code to the GitHub repo, which I was already doing, and the they would take care of everything and serve a great site – here’s the kicker – for free</strong>. This was working really well for me for a while, until I decided to add security headers to the site. I tried doing this in two different ways: using Helmet</a> as described in the Sapper Docs</a> and using Netlify _headers</a>. The Helmet middleware didn’t work with Netlify, and I’m honestly not sure why – possibly because of how they host the file server (I’m not sure what/how they serve the site), or maybe I just f*cked up somewhere idk JavaScript is hard. Using the Netlify config didn’t work either because I’m not a paying subscriber, which wasn’t too clear at first. This would probably be solved if I just gave them money, but I’m not going to do that any time soon.</p> I decided to move to DigitalOcean</a> and host the site myself (or I guess not totally myself – I don’t feel comfortable running it on a Pi from my house just yet) with Caddy</a>. I went with DigitalOcean partially because of their dev-focused business model, but mostly because they’re cheap (my droplet is USD$5/month and I get $50 credit with them from the GitHub Student Developer Pack</a> – my first year of hosting for only $10).</p> Unfortunately, I didn’t quite run the transition too well. It definitely wasn’t zero downtime – it was probably more like a-few-hours downtime. I doubt this really affected anyone because of my small audience, but I was still up late working on it. If it had happened a few days later, it might have gotten mixed up with the Cloudflare downtime</a> (I don’t use Cloudflare).</p> My current setup uses a custom CaddyServer configuration to serve my site, which I am absolutely loving so far: it is so much easier to run advanced setups with Caddy than Apache or Nginx. I wanted to move away from Javascript analytics (via GoatCounter</a>) to log-based analytics (via GoAccess</a>) but found that Caddy’s structured logs are not supported natively. I’m hoping they’ll add it</a> soon though! Also, I stumbled on this article</a> while writing this, and may check that out.</p> I’ve got two major todos for my site in the near future. The first thing I want to do is Dockerize. Using Docker will significantly clean up my current setup, and allow me to host more fun things on the same machine (for example, my own analytics</a> and maybe git</a>). With that said, I probably should get a little better at Docker first… For beginners, Flavio Copes’</a> has a few pretty great posts on the subject.</p> The second thing I want to do is a bit of a larger project, and one I’m not yet sure how I’m going to solve. I’ve run into some problems with the frameworks I use to develop figbert.com</a>. Sapper generates inline scripts and blobs, which tanks the security of my CSP. Their proposed solution</a> is to inject nonces with JS middleware, but this doesn’t work with CaddyServer. I would much rather Sapper avoid inline scripts</a> altogether, but this doesn’t seem likely. I really like the freedom that writing static sites in Svelte provides me versus other static site generators, but this could be a dealbreaker. I might talk a look at Routify</a> and see if that’s any better. I’ve also been reconsidering using TailwindCSS, however useful it is, after reading these</a> three1</a></sup> articles</a>. I attempted to replicate the current look of figbert.com</a> without TailwindCSS, and failed. So I’m probably going to rewrite the site again, though I’m not sure how. Through iteration, we will arrive at a stable version – this one, unfortunately, is not yet it. If I do leave Svelte, which is probably a 50/50 chance at this point, I would probably go either back to Hugo</a> (with some variation of the incredible archie</a> theme) or to Zola</a>.</p> So, some pretty crazy things have happened. For one, Twitter got hacked</a> by some people from OGUsers who got access to some internal managment tools. This hack, though not necessarily the most technical, was super high profile. There were a few really good articles</a> written about the hack, but I also noticed a few people acting in ways that I thought were not appropriate. More specifically, Brian Krebs</a>, who reacted to the hack by repeatedly</a> doxxing</a> the hackers, many of whom are still teenagers. He’s received relatively little criticism</a> for doing this, though it appears to be a pattern of behavior</a>.</p> There was also a brief mention of a new macOS malware</a>, but judging by the lack of further publicity it’s probably not that common in the wild. That, or people don’t care because it’s spread mainly through torrenting sites and they don’t want to protect people they perceived to be morally beneath them – which is wrong, obviously.</p> I learned how to drive! Or rather, I got kind-of licensed to drive a car in the US. I’ve been driving around a lot lately as something fun I can do to pass time in quarantine, and last week I finished my online driver’s ed course and got my learner’s permit. It was a surprisingly simple and COVID-friendly process! The most complicated/least COVID-friendly part was going to the DMV to take the written test, but I wore a mask and gloves and glared at anybody who got remotely close to me. The next steps are much more dangerous to do during a pandemic, but we have a year to complete them so we’re going to delay them (some in-person driver’s training and the driver’s test) for a bit. For now though, I have a piece of paper that says I can learn to drive!</p> I’ve also taken some time to focus on decentralization. I’ve had a Mastodon account</a> for a while, but I rarely used it. I didn’t have a mobile client, and I almost never use social media on my computers. To help change my behaviour I downloaded Mast</a>, a paid Mastodon client for the Apple ecosystem. I went with Mast mostly because it’s absolutely gorgeous</em> – and open source</a> – which is something I value in the apps I use. Apps that look really nice are fantastic inspiration for my own apps’ designs, and this one is no exception.</p> I’ve also started using three other apps much more heavily recently: Feedly</a>, Octal</a>, and Element</a>. Feedly is a fantastic RSS reader, which helps me keep track of all the interesting blogs that I find online (I read somewhere about self-hosting an RSS reader, which sounds pretty cool, but I forget what it was called). Octal is a HN client for iOS, which is really handy for keeping up with my favorite tech news when I’m away from my computer. Element (previously Riot), is the first-party Matrix</a> client that makes using a decentralized E2EE chat service feel better than Discord. I love it. In an ideal world where I could choose the tools I use to communicate with people, I would only use Signal and Element/Matrix (for replacing iMessage and Discord/Slack, respectively).</p> Oh my god there’s so much new music.</p> Jonny by Bar Tsabari</a></li> Kukuriku by Eden Ben Zaken and Omer Adam</a></li> Ta’azri et Atzmech by Dekel Vaknin</a></li> Sivuvim by Eden Hason</a></li> Lecha Dodi by Moshe Peretz and others</a></li> Haravot BaPita (Album) by Peled</a></li> Magevet BaAvir by Noroz</a></li> Nadav_15 by Shekel</a></li> Moshe by Narkis and Miri Mesikah</a></li> Tik Tok by Mohamed Ramadan and Super Sako</a></li> Aliyato VeNiflato Shel Shem Tov Hevi by Tamir Bar</a></li> Falafel Pop (Album) by Quarter to Africa</a></li> Lama Kacha Atzuva by Idan Raichel and Stav Beger</a></li> Kapara by Kevin Robin and Rotem Cohen</a></li> Ehad HaAm by Aya Zahavi Fayglin</a></li> </ul> I hope you enjoyed reading this update! I know it’s much longer than usual, and I’ll try and keep them shorter than this from now on. I plan on writing more short, topic-focused articles rather than just updates, but those will come soon. For now though, I need to focus on knocking out some of the larger projects that I’m working on. I keep finding cool and interesting things to do, and I need to work on finishing the ones that I’ve already started!</p> Salamat, FIGBERT</p> This previously linked to https://edvinleander.com/2020/07/15/stop-the-overuse-copy-pasting-and-unecessary-libraries/</code>, but the site has been down for years now and I neglected to archive it. Such is the nature of the internet. ↩</a></p> </li> </ol> </section> How to Replace Keybase in 3 Easy Steps 2020-07-03T00:00:00+00:00 Ever since Keybase was acquired by Zoom</a>, a company</a> with</a> a</a> very</a> bad</a> history</a> with</a> security</a>/privacy</a>, people wanted an alternative</a>. There have been a few different alternatives proposed: this is the best</del> mine.</p> </span> Before we talk about replacing Keybase</a>, we should have a good idea of what Keybase actually is. It’s main features are as follows (ordered as on the website):</p> E2EE chats and messaging (people and teams).</li> Cryptographic identity verification from around the net.</li> KBFS</a> (Public signed file hosting, private E2EE file storage w/ sharing, Static site hosting??</a>)</li> Git repositories? Crypto? An alternative to PGP?</a></li> </ul> I’m not the first person to try this, obviously. Some brave folks have tried to build Keybase alternatives, such as keys.pub</a> and the brand-new Keyoxide</a>. I’ve tried both, but found that though they both are good in their own right, they are not the solutions that I am looking for.</p> There are a few great pre-existing options for encrypted messaging: Signal</a>, ProtonMail</a> if you want to go full email, Telegram</a>, and WhatsApp</a>. However, they all have their problems (though I use the first two on a daily basis). Signal requires a phone number, and is more of an iMessage/text replacement than a Slack-style chat app. Protonmail is literally not chat – it’s email. Telegram is (debatably) not</a> secure</a>. If you use WhatsApp for security you might be crazy – I only use it because it’s the way</em> to communicate with people in the Middle East and Africa.</p> Instead, I would recommend you use Matrix</strong></a>. Matrix is an “open network for secure, decentralized communication,” and it’s the perfect replacement for Keybase’s chat and I would argue most other chat apps too</del>. It utilizes E2E encrypted messaging, and can be self-hosted as well or if you’re cheap like me just get your friend to host</del>.</p> In addition to a Matrix server, you also need a client. For this, I recommend Element</strong></a> – though Nio</a>, once stable, will almost surely be my go-to. Element is a beautiful Matrix client with a bunch of awesome features, including Slack-like integrations, and apps for pretty much every major platform (Linux, MacOS, Windows, iOS, Android, and a web client). Plus it looks a lot like Discord.</del></p> Replacing Keybase’s original function</a> is probably the most difficult part of this tutorial: cryptographically verified identity proofs is a great and innovative idea. I would swap this out with an IndieWeb</strong></a> profile – one part</a> of the larger microformats</a> HTML structure. There are some pretty great tutorials out there (I would recommend this one</a> by the fantastic Kev Quirk</a> and this one</a> by Brian Wisti</a>), so I won’t go into too much detail about exactly how to do that. However, it’s important to note that though Kev recommends hiding your h-card with the display: none;</code> property: don’t do that</a>. I just merged my about and contact pages onto my homepage, and added the microformats classes to my existing markup.</p> </picture> Replacing KBFS is easy to do, but hard to get right. Swapping to Google Drive</a> is probably the move that most people would make, but that abandons the entire security/encryption aspect of Keybase. There’s also Dropbox</a>, but that has the same problems as above. ProtonDrive</a> has potential, but it’s not out yet.</p> EDIT:</strong> ProtonDrive is now in beta for paying subscribers, but I am shifting away from the Proton ecosystem for other reasons and do not recommend it. More in a future blog post.</p> Enter Syncthing</strong></a>. Nikita Tonsky</a> wrote one of my favorite posts of all time about Syncthing – go read it. One reason Syncthing is so great is that it’s not the same thing as KBFS or any of the other “Drive” solutions. Instead of being a file hosting system, it’s a “continuous file synchronization program” – aka p2p. You have no data limits other than your storage and no third-party to worry about. Plus, sharing folders is also incredibly easy. Just read the article.</p> It would be a shame to talk about text chat, or really any form of communication, in this new pandemic age without talking about video chat. After all, the whole reason I’m writing this article is because the new videocalling giant Zoom</a>. So, how have I replaced Zoom and how does that relate to replacing Keybase? Well, Matrix happens to have a fantastic Jitsi Meet</a> integration</a>. Plus, the folks over at Jitsi are working on E2E encryption for their calls</a>. I’ve integrated Jitsi Meet into my self-hosted instance of Matrix, and now all my videocalls are just that – mine!</p> Swapped chat to Matrix and Riot.</li> Swapped identity verification to Indieweb.</li> Swapped file storage/sync to Syncthing.</li> Added videocalling to chat program via Jitsi.</li> </ul> Keybase is a great service, and the people who work there should be really proud of what they’ve built. However, given Zoom’s aquisition of the company, the stability and security of the product have been called into question. So, ever one to hop on a hype train, I jumped ship. I’m really happy with my solution, and I’d love to hear your thoughts as well.</p> Keybase Has an Onion Address 2020-07-03T00:00:00+00:00 Little known fact about Keybase</a>: it has two onion addresses.</p> </span> The short fncuwbiisyh6ak3i.onion</code> and the longer keybase5wmilwokqirssclfnsqrjdsi7jdir5wy7y7iu3tanwmtp6oid.onion</code>. Huh.</p> I Wrote This #2 2020-06-27T00:00:00+00:00 Hi! So, I just started this series and I already missed a week. Oh well. This week was a crazy week for Apple and their operating systems, as they announced iOS 14 and macOS 11 Big Sur at this year’s WWDC</a>. I’ve also started working on a new game project, another little side project, and finding a good alternative to Keybase</a> now that it’s owned by Zoom</a>.</p> </span> When I was a kid – or at least, a younger child than I am now – I had a friend who jailbroke his phone. I thought it was super cool, and wanted to do the same. At around the same time I was watching TechSource</a>, I got deep into those Top 10 Amazing Jailbreak Tweaks</a> compilations. However, I wasn’t allowed to jailbreak my phone because my parents said it wasn’t allowed, and that was the end of that.</p> Until last week, when I decided to jailbreak my iPhone 7 running iOS 13.5.1 using the checkra1n</a> hardware exploit. To do this, I couldn’t use my USB-C to Lightning cable</a> due to a known issue</a> and instead swapped to a USB-A to Lightning cable with a USB-A to USB-C adapter. After jailbreaking the phone, I began to install a host of tweaks. One of the most interesting things I learned about the jailbreaking community is that charging upfront for apps is quite common for premium and well-known tweaks, which stands in high contrast to the iOS App Store’s ubiquitous freemium payment model (a free download followed by a monthly subscription) especially in the light of the Apple-Hey.com controversy</a> of recent weeks.</p> All in all, I spent around USD$21 on tweaks, which is infinitely more than I have spent on the App Store – so maybe charging upfront does work. Huh.</p> Your browser doesn't support the video tag, which I use in place of .gifs, and/or the video formats in use here – sorry! </video> Soon after completing the jailbreak, WWDC</a> happened and new developer betas were released. Never one to miss jumping on a hype train, I updated both of my main devices to the new software</a>. These new betas come with a whole host of bugs, as is to be expected, which I plan to write about in a blog post coming soon. So far, however, I’m liking these updates – the PiP on iOS is incredible</em>, and the new design on macOS is growing on me.</p> Some interesting resources:</p> An interview of Craig Federighi by MKBHD</a></li> These</a> three</a> issues</a> on GitLab for iTerm2 users</li> These</a> two</a> posts for Firefox</li> This Github issue</a> for Homebrew users. so basically everybody.</del></li> </ul> Interestingly, I was having a bunch of issues with brew (curl</code> was broken?) after updating to Big Sur but now it seems to be working flawlessly. Very nice.</p> For my next medium-large project, I’m going to need to make some 3D art and models. One problem: I don’t know how to do that, or rather, I didn’t. Thanks internet! I powered through the incredible Blender Beginner Tutorial Series</a> by the Blender Guru</a>. Here’s my final animation:</p> Your browser doesn't support the video tag, which I use in place of .gifs, and/or the video formats in use here – sorry! </video> Now that Keybase is compromised</a>, I thought it would be good to seek out an alternative. Initially, I thought that keys.pub</a> would be perfect – I thought it was a slightly-tweaked fork. However, this does not appear to be the case. Instead, it seems kinda like garbo – it doesn’t have the budget Keybase had, and as a result, is majorly lacking in features. However, I’ve replaced Keybase completely, and am happier with this setup than I ever was with Keybase. My new setup uses Matrix</a> and Riot</a> for messaging, Jitsi</a> for video chat, Syncthing</a> for file storage, and an Indieweb</a> h-card for identity verification. I plan on writing more about this at a later date, but I’m really pleased with the results so far.</p> I’ve got a few interesting projects in the pipeline at the moment, not counting the few articles that I’ve already promised to write: Glitches in macOS Big Sur and How to Fix Them</em> and How to Replace Keybase in 4 Easy Steps</em>. My next major project is a dreidel game/simulator for mobile devices – I plan to code it with Godot</a>, make the models with Blender</a>, and do any music/sound stuff in FLStudio</a>. Naturally, I’ll keep you all updated on any progress there.</p> I’m also working on a much smaller project that should take around a day or two to complete. The working title is “howmanyisraels.com” – essentially, it’s a website that takes the size of the State of Israel, and compares it to the size of a given location and tells you how many times Israel could fit inside it. All I need to do is find a suitable mapping API (preferably not by Google) that will tell me the size of different locations and I can whip it into a static site!</p> I’m also planning on revamping the txtodo macOS app with Mac Catalyst, which I feel I haven’t properly explored. I’ve also discovered a bug in the app that causes issues when tasks are modified by CloudKit instead of the user, so I’m going to have to work on that as well. Shouldn’t be hard!</p> </picture> I’m not sure what tense to write my headers in – I currently write them in present continuous</a> tense but I am considering writing them like I do Git commits. Huh. I’m also committing to writing more non-update blog posts (i.e. the two I mentioned above), but we’ll see how that goes. Writing is hard, and you shouldn’t make false promises. I’m also working on implementing webmentions to the site, which is a little difficult with a static site but I’m going to give it a try. Lastly, I really need to learn how to use vim but it’s so goddamn hard. I’m writing this blog post with it but I’m definitely not using it to its full potential at all.</p> Wow, these have been a great two weeks for the Israeli music scene. In no particular order, here are some good new songs (basically in the order they appear on the trending tab):</p> Habib Albi – Static and Ben El</a></li> Rak Banot – Itay Levi and Stephane Legar</a></li> Shvuaim – Eden Hason</a></li> Yom Huledet – Eliad</a></li> Hatzi Madleket – Dudu Faruk</a></li> Hayde – Peled</a></li> Ze Mi She Ani – Peled</a></li> </ul> I know music isn’t tech, but I like it and now it’s a recurring section.</p> So, that’s been this week’s update. I’m really enjoying writing updates on this blog, and I hope you stick around. Until then, farewell!</p> – FIGBERT</p> I Wrote This #1 2020-06-13T00:00:00+00:00 So, I’m FIGBERT – I wrote this. nailed that intro.</del> I’m a highschooler working in the tech world, currently stuck at home like pretty much everybody else who’s not trying to die. I plan on using this blog to writing mainly on my experiences in the tech world, either in focused articles centered around one topic or more summary-style roundups like this, the “I Wrote This” series.</p> </span> I’ve always found statements like that kind of funny in their generalness, and was thinking the other day that if I ever made a game studio I would name it “I Made This Game.” Then, when I was trying to think of what to name this series (The Weekly Fig? Fig Talks?), I thought it would be funny to do something similar. I plan to release these once a week on Saturdays, if all goes to plan. Without further ado, here’s a brief summary of interesting things that happened this week.</p> Some people reading this may be familiar with my most recent project, txtodo</a> – you can read a full writeup of the project here</a>. The point of txtodo was to create a todo list application without any feature bloat I found in other competing solutions – every task deletes at midnight</em>. This week I launched the MacOS companion app for txtodo, completing my long-term “cross-platform” goal for the app. I went from not knowing SwiftUI at all to having two published apps that talk to each other! Making computers talk to each other is, like, the hardest thing you can do with computers! Ok, technically a lot of it is just Apple APIs but still, let me have this.</del> I think there is probably one more major change I could make to the app (moving the tasks from an @State</code> variable to an @EnvironmentObject</code>), but after that it’s pretty much bugfixes and… the bank? I don’t know.</p> I while back I saw that when I editing my project README.md using the GitHub web app, it got a little green “verified” badge next to the commit message. I looked into it for a bit today, and found this GitHub help article</a>. One GPG install and Protonmail-key-download later, now all my commits have a green badge! Neat!</p> I made the switch from nano to vim a few months ago, but up until a few days ago I was using the barebones default macvim</code>. Not anymore! I DuckDuckGo-ed “vim configuration,” and downloaded what is supposedly “The ultimate Vim configuration.” from amix</a>. It’s a pretty good configuration out-of-the-box: I’m using the Awesome version with the fantastic Jetbrains Mono</a> font. I’m still not that good at actually using vim to its fullest potential but I’m confident that I’ll improve with a little practice and ascend to elite vim status in no time.</p> </picture> Speaking of elite programmers: Rust! I’ve been thinking about learning the language for a while, and now that I have no main project (other than maintain my finished apps and sites), it’s as good a time as any to learn a new language. It’s idiomatic syntax combined with low-level abilities makes Rust seem a little bit like Python+ to me. It also helps that it’s skyrocketing in popularity – maybe I’ll do the cliche Rust developer thing and post about how much I love the language on the orange site</a> once I learn it. ok so I was planning on linking a bunch of posts like that here, but when I looked for them on HN I actually couldn’t find any, but I distinctly remember reading them so I stand by this statement.</del></p> I made the switch from Brave</a> to Firefox Developer Edition</a> this week, after accumulating around 50 USD in BAT. I started using Brave almost when it came out and loved it – the chromium feel with a powerful built in adblocker? It sounded too good to be true – and perhaps it was: over time the browsers performace began to slow down, they hijacked links</a>, launched a weird Zoom competitor</a> with no fanfare or announcement, and refuse to get rid of their required KYC</a>. Anyways, I dropped it. Now my dock is all blue!</p> </picture> On a non-tech-related note, a bunch of good new music came out recently. In no particular order, here are some good new songs (largely out of the Middle East):</p> Coronavirus by Mohamed Ramadan</a></li> Zot Ani – Ela Li Lahv</a></li> Million Dollar – Noah Kirel and Shachar Saul</a></li> Naughty – Tzagay Boi</a></li> TROLLZ – 6ix9ine & Nicki Minaj</a></li> </ul> So, that’s been this week’s update. I’ve got some exciting ideas for my next posts, so I hope you stick around for what’s to come. Until then, farewell!</p> – FIGBERT</p> txtodo 2020-05-20T00:00:00+00:00 txtodo</strong></a> is a minimalist open-source todo list app inspired by Jeff Huang’s One File to Rule Them All</em></a> and Mac-assed Mac apps</a> like IINA</a>, Sketch</a>, and Safari. It’s available for download on iOS and macOS via the App Store.</p> </p> txtodo lists your immediate, short-term tasks to help you get things done without overthinking it. At midnight, all tasks are discarded so you can start fresh tomorrow. You can also create up to three long-term tasks that “float” with you from day-to-day.</p> On 15 February 2020, I deleted my account with Trello. A solo developer, I know my projects inside and out. I know what features need work, and which graphics need a refresh. I don’t need a massive crudware</a> suite to keep track of every little move. I need something small.</p> </p> A dead-simple UX and midnight expiration date prevents the overwhelming buildup of tasks we delay over and over again, and helps you deal only with tasks you can get done that day</em>.</p> Download the app on the App Store</a>, and explore the source code on my forge</a>.</p>

FIGBERT

Google Refreshes Brand Icons to Make Them The Worst

2026-06-17T00:00:00+00:00

I mostly don’t use the Google Suite—but when I stumbled on these through Sheets, I genuinely thought there was a rendering error in Safari. ★</a>

Like Someone Who Actually Just Wants to Talk About the Settlements

2026-06-03T20:00:00-07:00

Like Dreamers by Yossi Klein Halevi. Finished on May 22, 2026. Rating: ★★★★★ 

Update: I wrote my final paper for POLISCI114</a> on this book, which I’ve uploaded for you to peruse</a> should you be interested. The original, much shorter, and somewhat less formal review is preserved below (with a tweak to the penultimate paragraph, which contains a new direct quote).

When I first picked out this book, I anticipated that “the story of the Israeli paratroopers who reunited Jerusalem” would have a lot more to do with the actual battles of the Six Day War. But as I now know: anyone looking for that book should look elsewhere. Like Dreamers is about these paratroopers and that war, but only insofar as Halevi believes the war to have been the most pivotal moment in modern Israel’s history and the paratroopers a useful example of engaged, elite citizens shaping the national project.

Halevi instead focuses on the Six Day War as having introduced a previously unthinkable issue in the Israeli discourse: that of “should”. Israel fought a war and won—far surpassing the historic precedent of merely surviving—meaning that for the first time the nation thought not only of what it “could” have but what it “should” have. Like Dreamers is about the ongoing adjudication of this issue. In essense, it tackles the following question, lifted in this instance from the book’s inner monologue written for Yisrael Harel:

…what nation would turn its back on its ancestral heartland, won in a defensive war against attempted genocide? </blockquote>
My stance on this issue is pretty deduce-able. In our more settled times (compared to ’67), my mental model of the country has always been defined by the places I’ve actually been and where the vibrancy of the country’s life is concentrated—so settlements are less important, to the point that my pragmatism toward an eventual cessation of hostilities means I would give them up in a heartbeat. That puts me somewhere like “sympathetic to the above sentiment but it’s irrelevant.”1</a> I nevertheless enjoyed that Like Dreamers exposed me to a vast spectrum of dialogue that I had not yet encountered.
In this book there is a secular, weed-smoking, kibbutznik musician who builds an intense and personal relationship with God in a country largely restricted to religious binaries. There is a religious Zionist yeshiva-leading Rabbi who fiercely defends secular Israelis as Jewishly righteous and condemns violations of secular laws in religious terms. There is a man who fights for the kibbutzim, another for a free market, and a third for the public to draw a line of descent between the socialist communes and settler villages. Getting to know these characters means constant engagement with the brilliant and bewildering entanglement of Israeli society, all too recognizable to those who know it well:

His love for the land of Israel was inseperable from his famous love for women: “There are those more beautiful than her / but none as beautiful in the way she is.” </blockquote>
As Israel’s ultra-religious population skyrockets and the country’s long-standing conflicts with its neighbors seem ever more intractible, effective solutions are going to have to get a lot weirder. That makes a book like this all the more important. Like Dreamers resists simple categorization—it is not an endorsement nor condemnation of the settler cause. I mean, it even contends with one of my favorite wacky solutions of all time: “keeping settlements intact under Palestinian sovereignty.” If the solutions people came up with decades ago were viable they would have worked decades ago. Only the weird ideas are going to work from here on out. Just look how weird we’re already getting: sending laser defense systems</a> to a desert kingdom to fend of nuclear-addled Iranian attacks, fighting terrorists hand-in-hand with the Lebanese</a>, and witnessing the rebirth of a nation at the hands of al-Qaeda’s most Hoover</a>-pilled soldier. Is it that unfathomable to see a future where Jewish religion and secularism converge</a>? I am holding on to my hope for an incredible future.
Like all good, focused book reviews, I will end with a tangent. Like Dreamers formalized my conception of one of Israel’s most impactful practices: “the IDF’s doctrine of preemption, of taking the war into enemy territory.” In the United States this is not something one needs to think about, as America’s enemies are oceans away. But in Israel it’s a big deal—fighting on the (tiny and densely populated) home front means untold Israeli civilian casualties. This doctrine’s impact is obvious in the case of 1967, when a then-smaller Israel fought the Jordanians in the West Bank (annexing Jerusalem), the Syrians in the Golan Heights (now also annexed), and the Egyptians in Sinai (returned for peace). But the doctrine is still very much relevant: just look at the maps of the current conflicts in Gaza</a> and the North</a>. Despite what the American zeitgest may imply, I do not believe that Israel is a radical expansionist power in the Middle East. Rather, it’s implementing the same doctrine as always. The West is just profoundly forgetful.
In any case, as you may have observed already, I quite liked this book. I found it rather engaging: it managed to skip the part of most Israel-oriented books where they start by trying to introduce you to the conflict and then end before they can talk about anything else. Consider giving Like Dreamers a read—and if you do, let me know your thoughts.

The irrelevance of the above is especially true in a political order that pretends borders are static—which maybe prevents war on net (not in Ukraine?) but perpetuates long-standing injustices by ruling out a whole class of problem-solving methodologies (see: Kurdistan)—and once more when you factor in what I grew up calling “Ben Gurion’s triangle” but from some research appears to be more commonly referred to as the Israeli trilemma (aparrently coined at least in part by Tom Friedman, whom I had the pleasure of hearing speak last year). ↩</a> </li> </ol> </section>

ME163*

2026-04-18T18:00:00-07:00

In ME103 you build a “high-fidelity product or object… with the PRL’s</a> manufacturing resources.” I knew from the outset that I wanted to make something a bit humorous but that I would actually use in my day-to-day life. So in the funky spirit of co-opting projects to do things I want to do anyways</a>, I set out to learn something of what I’m missing out on now that ME263</a> is no more. Ultimately, I built a beautiful piece of furniture: an artistic, layered stool with three angled aluminum legs, six statues “struggling” to hold up your weight, and dark walnut detailing throughout.

While the stool may seem simple, there is a surprising amount of complexity hidden beneath the surface. Most of those challenges have something to do with scale: there are 49 drilled holes, 27 screws, 18 tapped holes, and 9 threaded inserts in the body. The structure is 14 inches across and 24 inches tall—near the limits of how large you can make something in the PRL, and in many ways well over them, meaning I had to pull quite a few tricks to get it all working.
The stool is made up of four major components, which are (from bottom to top): the legs, the mount, the plate, and the statues. I’ll cover their manufacture individually below.

These were perhaps the simplest part of the project. I cut external threads onto three sections of 1“ tubing and called it a day: even though these insert into the mount at an angle, the tapped threads in the mount are tilted at the same angle meaning that the legs themselves don’t need to “know” about any of that complexity. That does mean that the base of the legs is at angle to the floor, however—so I ended up printing custom foot pads out of flexible plastic</a>.

Offloading the complexity from the legs means it has to go somewhere, and that’s here. How do you drill and tap 1“ holes on the mill at an angle? I first tried the indexing head—which can tilt and rotate—but found that it couldn’t hold anything near big enough. I thought of the rotary table, but quickly realized that having to adjust the fixturing of the piece after rotating it would negate the consistency of its rotation. And all this was after I had split the mount and plate into two parts from my original design in order to make it easier to work with!
I decided that all my pondering was holding me back from actually getting my hands dirty</a>: I mounted a sharpie in the mill, drew out the centers of the holes on the flat plane, and then moved the part to a 15° tilt table and eyeballed the rotation. Drilling each hole required a sequence of seven different tools to properly drill and tap. Ultimately this worked.


This enormous disk marries the legs to the rest of the seat. It has two sets of through-holes: three in the center which attach it to the mount and twelve around the perimeter for the six statues. These sets are countersunk in opposite directions. The central three holes were smooth going, but working on the outer rings I quickly became familiar with the range of motion of our mills—slightly over 11 inches, which becomes even smaller when you factor in the overhang of the 14 inch disk itself interfering with the movement. This was a problem.
You may begin to sense a theme here: I returned to my trusty eyeballing strategy, calculating the coordinates of each holes, carefully keeping half of the circle in-bounds, and then performing one risky rotation to make the second set of six. I employed a clear, laser-cut jig on this step to aid the accuracy of my rotation and get all 12 holes aligned. Sensing another theme, this worked quite well.

Inspired by Atlas, the videogame Necropolis</a>, and Art Deco style, these human figures are the crux of the stool’s humor. Lying on their back, pushing up against you with braced arms and legs, one must imagine their job as at least somewhat unpleasant. Similarly, making them I had to leave my favorite CAD program</a> behind (it lacked a drafting</a> feature) in favor of the more feature complete but thoroughly unpleasant Fusion</a>.
I ended up casting three times, starting first with a single test part before upgrading to a family of three statues per pour. Filing and polishing them was the worst thing ever, so I forced Vivek</a> to help me, and now they look nice and contrast beautifully with the darker colors of the stool.

After all this machining, I could see the finish line. I whittled</a> three covers for the leg joints, to give them more substance, and fixed them in place against the legs with set screws. I planed down a seat from the same enormous walnut board, and set six threaded inserts in it to fix it on top of the statues. I became best friends with an orbital sander—thanks Sam PW for that—before whipping out my trusty spray paint to turn the polished aluminum matte black. And, all of a sudden, it was done.
I’ve come away from my late nights and early mornings in the PRL with a tremendous amount of technical knowledge and an intuition for working with metal to make professional hardware. But the true purpose of ME103 is to get you to work on one totally unmanufacturable project so that you get all your mistakes out of the way and never send a machinist something ridiculous in the real world. I am happy to report that in this regard I was beyond successful: I got a profound number of mistakes out of the way over the course of this project. You’re welcome, machinists.

All the time I spent creating could really be divided into two categories: exploration and implementation. Any time I approached a process to realize an idea, the first time around was long and arduous—and then repeating it to make additional components was quick and easy. Further, I would make many, many mistakes in the exploration phase. On its face, this may seem like an endorsement of the prevailing idea that ideas matter more than implementation: if I could only start with the right idea, I would skip the hard part of messing up repeatedly first. But I think this experience goes to show the exact opposite. At the highest level, my idea was simply to build a stool. I could only come up with the right sequence of “good ideas” to make that happen through the feedback I got attempting my implementation over and over again. My “mistakes” guided my solutions. Letting the process guide you, as such, is a sort of humanity and craft that I hope to bring to products I have a hand in long after I am no longer turning the wheels on the mill myself.
I have never worked as hard as I worked on ME163* ever in my life. The stress alone has taken years off my life, and that’s not even factoring in the grease and aluminum shavings I’ve eaten. I was alone with my thoughts</a> machining, working out of a greasy paper notebook. And I made something from nothing</a>. Now it lives happily in my dorm room.

Coconut Water

2026-04-09T15:00:00-07:00

It took me two decades to learn that coconut water does not taste like sweat. Rather, some bottles on the grocery store shelf are full of sweat while others are not. I tried them all so now I can tell you which is which. 

Once Upon a Coconut</a></li>
Coaqua</a></li>
Harmless Harvest</a></li> </ol>
These are the kings of coconut water. Somehow, they manage to taste toasted. I don’t understand how you do that. Delicious.
Notes and caveats: Do not try the flavored versions. Good coconut water is like good sushi—simple is better. Also Coaqua’s brand copy is extremely cringe. And why is Harmless Harvest pink?
C2O</a></li> </ol>
I really like the corrugated metal look but the flavor is, in my view, unremarkable. Nageena’s favorite.

Taste Nirvana, Parrot, etc.</li> </ol>
I would rather have normal water than these, unless I’m feeling desperate or it’s been a while.

Vita Coco</li> </ol>
Do not buy under any circumstances. Despicable, vile drink.

Stanford Quarterly Reflection (Y3Q1)

2026-03-30T15:45:00-07:00

Welcome back to our regularly scheduled programming! Junior year has been downright blissful. Let me tell you about it. 

I discovered two important things about scheduling this quarter:

If you plan things right—or just get lucky and stumble into it, I’ll let you guess how I found out—you can have no finals at all. This may be the consequence of a necessarily project-focused physical major, but while we’re picking poisons I’m going to drink this one.</li>
Advanced courses with low unit counts are the greatest! You cover exciting and fresh material in a relaxed environment. Ideally every class would work like that—one can dream.</li> </ol>

This course was the cornerstone of my quarter. I learned heaps about rapid manufacturing, interfacing between parts, how hardware—which contextually, I have learned, basically means screws—works, and more.
Most importantly, though, it just got me making things. When Vivek</a> and I bought a couch but couldn’t get it flush against the wall because of an inconveniently located pipe, we opened Shapr</a> and made new legs to lift it over the obstacle and into the position we had initially imagined. When the gear guard on my bike cracked and started stabbing my pant legs, I rapidly iterated a replacement. This course helped reinforce for me how malleable the world around us can be.

Of course, the major tangible outcomes</a> of this course were the projects: I’ve posted about both</a> already</a> elsewhere on this site. I’m proud of the results and how much I learned in making them. Working with my design partners, Vivek and Trun</a>, was a joy. But I do expect the attitude change to be far more productive than any physical output in the long run.

I tried to avoid this course entirely, but the bureaucracy won out and I enrolled. The first project centered on a PSA to induce impossible behavioral change rather than addressing the root issue via a relatively simple systemic change. The second project was decently fun: my lovely group and I built an escape room entitled Jank Ass Spaceship, designed at least one totally novel puzzle, and used a simple val</a> to remote-control a laptop. The third project I commandeered to hang an unauthorized photo collage of folks in the PRL</a> outside of AMPS as a means of community building. It’s still up! I may expand it soon.


The incomparable Professor Sallam</a> described this course as “PoliSci Coachella.” He wasn’t kidding: we learned about nationalism from renown motorcycle enthusiast Frank Fukuyama</a>, democracy and it’s tumult from Larry Diamond, the unique case of Latin America from Alberto Díaz-Cayeros, the international promotion of democracy from Mike McFaul, and numerous other subjects from Didi Kuo and Sallam themselves. How fortunate I am to be in a place where these are normal opportunities. The sessions I enjoyed the most were, of course, case studies, and it was generally wonderful to have an island of writing and analysis in my otherwise technical quarter. Also, with the overlap in material and the more engaging teaching style, I think this course should take the place of POLISCI1.

Three good things came from this class:

It satisfied the first third of my Domain Focus for the Design major.</li>
I learned that palm trees are, in fact, a grass.</li>
We had one awesome class where we talked about The Winds. I don’t think anyone else in class was sufficiently appreciative of the fact that the products of Hadley Cells and the Coriolis effect were the stuff of myth, discussion by pirates, and the primary drivers of the economy and globalism up until fairly recently.</li> </ol>
My general knowledge of natural ecosystems has increased by a mild-to-notable amount, and I am now perhaps 0.5% more likely to read Braiding Sweetgrass.

Portfolio building class! Awesome that there’s time carved out for this in the program. It probably doesn’t need a teaching team of six. If, however, it manages to get me a job at LoveFrom I will sing its praises until the end of time.

I spent this quarter as the only president of JSA on-campus, which was a rewarding endeavour. The community is thriving: Grupo Benji continues to be the bane of the Coupa Cafe line, and Special Drake was not only super fun but a widely congratulated theme for our Fall Quarter party (surpassing my expectation that I would be the only person truly enthused about it). We’ve also adopted Basecamp</a>, which I hope will help us stay organized, democratize participation even further, and leave a trail of documentation for those that come after us to help them take up the torch. The formal establishment</a> of the Israel Studies Program</a> this quarter has similarly helped to formalize the legacy and continued existence/stability of Jewish life on campus. I was honored to be invited to its inauguration, and I’m excited for where the program will go in the future.

I got up to plenty of shenanigens with my friends as well. Daniel’s FLiCKS was yet again a raging success, with a screening of Austin Powers</a> followed by an evening of discussion with Jay Roach himself. Daniel also starred, along with Henry, Carter, Jack, Odin, and myself, on The Quambos intramural indoor volleyball team. I don’t know that I’ve ever had more fun than playing our first game, winning through spirit rather than skill, and designing our preposterous kits. The league will certainly fall to us next quarter. Living in French with Vivek has been awesome: it seems it’s really as simple as liking the same snacks (and perhaps videogames). And Nageena and I went on a whole host of adventures—from kayak camping</a> in Point Reyes, to Smuggler’s Cove</a> and the Little Red Window, to Tahoe with no snow.
Of course, many other people helped make this quarter so beautiful: my family, Dani, Tommy, Lulu, Aaron, Yoni, Miles, Omry, and perhaps you, dear reader, as well.
I love you all!

Aspirational Surfing

2026-02-09T11:51:00-08:00
Sweetness and Blood by Michael Scott Moore. Finished on February 01, 2026. Rating: ★★★☆☆ 
A little over three years ago when I read The Desert and the Sea</a>, I noted that Michael Scott Moore</a> had also written a book about surfing. As the life-long Californian that I am, the sport and its aesthetic associations are familiar to me. But it took until a recent tropical</a> obsession</a> to convince me to pick up the book and follow his winding tale of “how surfing spread from Hawaii and California to the rest of the world, with some unexpected results.”
I enjoyed his world tour and learned something everywhere we went. Some of the locations did just seem like they were “strange” in that they were poor, but others had more depth. The chapter on surfing in Israel and Gaza was particularly interesting to me, of course: I’ve swum in the shore breaks of Tel Aviv, and seen the confluence of people who ride boards a short walk south in Yafo where the shore is unprotected. Embarassingly, I had never heard of the hasake</a> before this book. I’ll keep my eyes peeled next time I’m in the country. Cuba and São Tomé and Príncipe were fascinating reads in the much same manner as Israel: locales where surfing is really foreign, really out of place. And his writing on Indonesia and Morocco made me want to bum there on the beach when I have some time off.
When I was a kid I went to surf camp</a> and was too scared to stand up on the board. Since then, I have justified my non-surfing with the ocean’s frigid temperature along the northern Californian coast. The Germans in Moore’s book put me to great shame by continuing on nonetheless:

“Remember, we had no wetsuits!” Drath said. “My God, we froze!”… Yet in those days, more people on Sylt went swimming… “Why?” I said. Father and son, in unison, said, “Central heating.” </blockquote>
You can see the beginnings of Moore’s later work in this book. As always, he weaves a larger narrative alongside his journalist’s personal touch—telling a history of surfing as much as a story of a friend’s cancer. I think there’s a lesson there in the ills of code switching: Moore leaves no part of himself behind, and the book is better for it.
Curiously, it was challening to get my hands on the book. The physical edition is not too common these days and I had trouble finding a digital version. I’m not sure where the EPUB I found came from.
Perhaps, having finished the book, I’ll now find some time to actually get out on the waves.

Recurse Center

2026-01-03T20:20:00-08:00
Began: 30 Jun 2025 Concluded: 19 Sep 2025 
Many people say the job market is a mess. Fewer people find themselves on a layover relocating for a job only to have the second leg of the trip cancelled as a war broke out</a> mid-flight. This was the situation I found myself in one evening in New York—and I began to form a new plan. There was a mysterious place I had heard of, from internet celebrities</a> and Stanford peers</a>, where programmers could go to build… things. It wasn’t all that clear to me actually but everyone I knew who had done it was very cool and I had yet to try living on the East Coast so I checked the window for applications, saw there was about a week left before the summer batch, and the rest is history</a>.

Recurse</a> gave me the resources—both technical and personal—to do the best work I’ve ever done. I was able to hack on projects I had long imagined but never had the time for, and pursue new ideas with wild abandon. On my first day in the office I showed up to a hackathon partner-less, and said that I was looking to make progress on a code forge</a>. Someone else had shown up alone as well, and volunteered. Josh</a>, fountain of git</code> knowledge and the world’s premier ZIP expert</a>, became a good friend. On my last day in the office, I one-shot a prototype dreidel game</a> at a hackathon for the Playdate console</a>. In between these bookending coding sprints, I discovered that everyone at the Recurse Center had arrived with their own profound specialties and interests. I was no exception: some things</a> I</a> made</a> there</a> have</a> ended</a> up</a> on this site.
Some notable occurences and accomplishments from my time at the Recurse Center have yet to make their way into the public eye. Frank</a> nerd-sniped me into a hardware project of grand scale that I am still working on</a>. When it’s done it may very well be my magnum opus. Keith</a> tricked me into learning Gleam</a>, and I have gone on to become a huge fan of the language. I also picked up Ruby and Rust, though I didn’t end up trying a Lisp as I had hoped. I met Filippo</a>, creator of age</code></a>, and he mailed me stickers that are now on my bike. David</a> planted the idea in my mind that “project people” should have a truck, and four months later I test-drove a Rivian R1T…
Working at Recurse is a high-agency playground. If you enjoy shooting off on productive tangents</a>—as I do—there are enough toys there to keep you entertained for a long time. Apply yourself and you can turn the 3D printers, retro computers, ESP32s, and ultrawide monitors around you into something wonderful. Or somethings wonderful! And while doing so, you’ll pair with incredible folks, present your work to a large audience, and hopefully become a better maker yourself. You are the only one in charge and there is no guide. I enjoyed that. In many ways it felt like a natural extension of the d. school’s philosophy of “learning by doing.”

I was chatting with a good friend of mine over summer, describing Recurse, when she posed a good question: “How is this going to change what you’re looking for in a career?” It’s true that I have yet to see a job posting for Person Who Builds Funky and Impactful Open-Source Stuff and Eats at a New Restaurant Every Night. But there are aspects of Recurse that I will look for wherever I end up next. I enjoyed the variety, the personal responsibility, the creative control, and the balanced blend of impressive goals with relaxed attitude. These qualities made the the work feel special. If you’re working on something with purpose and you think you can provide the above, do reach out</a>.
The Fall of a Great Product 2026-01-03T13:00:00-08:00 Many years ago, my family discovered a secret about the Grand Wailea</a>: amongst all their wonderful amenities, none managed to surpass the quality of their in-room body wash. A good product and a great one are worlds apart, and until recently, the Honey Mango scrub</a> was one of the best. I want to take a moment to appreciate what it was and honor a beautiful work by observing its passing. Three things made the body wash an excellent product: it smelled phenomenal, it lathered</a> well, and it had a shining amber color. Combined, these qualities made the exclusive wash not just the best in terms of performance—cleaning dirt and imbuing scent—but also in terms of operation. It was fun. I have fond memories of carrying the nuclear-orange bottle home from the package center and watching it glint in the sun. I enjoyed smelling like Maui every day. This was the deal—quality for inconvenience. For half a decade, we would trade money (in literal cost) and time (in slow shipping to the mainland) for a superior product from their online store</a>. We were happy with the arrangement; not compromising on little, day-to-day experiences</a> compounds and happiness is a good deal at any price. I am grateful for the years of excellent body wash in a way that most readers will likely find ridiculous. It wasn’t to last. Last year, the product line suffered a refresh. Each of its finest aspects was decimated: the smell was less sweet, the lather unimpressive, and the color—now hidden behind dark blue plastic—approached transparent. But the consultants who had made a mockery of this wonderful soap did not stop there. They found new insults. The 32oz size, the most economical option, was eliminated. And the new pump was designed with a sharp edge that sliced my fingers repeatedly. This last offense in particular seemed an omen. I conferred with my dad, who revealed that he had also cut his fingers on a bottle of the scrub, and we came to a solemn conclusion: it was time to move on. I do not expect I will find something quite as magical as the original soap anytime soon. But I have found a replacement</a>: a lather as good—if not better—than the original, a slightly inferior scent and color, and a much improved retail experience. It will do. A soap may seem too trivial a thing to memorialize. But what is a life but a compendium of little wonders? This was one of mine. Projectile Launcher 2025-12-21T00:00:00+00:00 Flinging an object a distance is perhaps humanity’s oldest challenge—and it’s one I faced with my teammates</a> for the final project of ME102: Foundations of Product Realization. Building a machine to achieve this feat covers a fairly vast problem space: fortunately, we were provided with some constraints. This narrowed our work to answering three fundamental questions. How might we propel a ping pong ball eight feet into a 4“-diameter hole using springs?</li> How might we reload after each shot (at least 5 times) without touching the balls?</li> How might we do each of these tasks consistently and smoothly?</li> </ol> After an exhausting week practically living in the PRL</a>, we had our answer: Our first three prototypes focused on the firing mechanism. Our designs evolved from a simple tube and rod system into a hand-held ping pong gun. We primarily iterated on the geometry of the barrel and housing of the springs, developing many features that would be central to our continued work: an octagonal shape, side-mounted modules, and an ergonomic handle. Though functional enough to be fun to shoot at each other, this version didn’t quite satisfy the project requirements: our continued design work focused on mounting the system. We were particularly interested in ways that this new form could make its operation less manual and more consistent. We bounced between gear systems, levers, cranks, and several more automated release mechanisms that made the device more manual and less consistent. Observing this perplexing result, we halved the number of springs (down to four, which yielded identical results in terms of distance while allowing us to simplify our launcher mechanically) and returned to the functional firing system we began with. Finishing the device necessitated a whole host of upgrades: a reload system built from an inclined tube, heat-set inserts throughout the chassis, and a reimagined stand manufactured on the laser cutter rather than the 3D printer. The reservoir in the reload system has pivoting guards above it which restrict the movement of the balls within: without the protection of the guards, our extremely light ammo would be jolted out of the device by the recoil from each shot. Meanwhile, the new stand boasts a guiding track which limits the wobble of the striking rod, neutralizing the vertical and lateral variability that might otherwise arise from a combination of human falliability and machined tolerances. Crucially, we dialed in the functionality through testing. This experimentation involved firing (and sometimes losing) whole clips of ping pong balls. We tweaked clamp and weight placements, varied the angle of the barrel, and trained to hold the handle</a> properly in order to get a reproducible shot. Part of what’s wonderful about working with a team you really enjoy, as I had the great fortune of doing on this project, is that the work itself becomes play. This is where real creativity surfaces. In the case of the launcher, this meant that we wanted to go beyond simply making a device that achieved the maximum performance score. We wanted something that would make us proud: it needed style. Our first addition to that end was the projectile iron sight. This small etched acrylic circle—aligned with a small notch at the rear of the reservoir—employs the parallax effect to give the operator a direct line of sight along the ball’s trajectory. It was mostly a joke: like putting a spoiler on a Prius. But it was actually pretty useful on testing day, which was a welcome bonus! Second, we designed the five-shot reload capability to approximate a bolt action mechanism. The handle is located by the mouth of the barrel, and physically prevents balls from descending from the reservoir in its resting position. When retracted, one ball enters the barrel, and a large chamfer on the front-facing side pushes any extra balls back into the reservoir as you conclude the loading process by returning the handle back to its extended state. The launcher is definitively non-lethal, but you hardly feel that way when you’re slamming the bolt around after a shot. Looking at a complex machine, it can be challenging to grasp how all of its minute components form the whole. In our CAD work for this project, we developed a visceral understanding of how that understanding might happen: piece by piece. A table-mounted ping pong cannon is too big a problem to solve all at once. But chunk it up, and it becomes approachable—a barrel is a cylinder, a handle is two perpendicular rectangles, and so on. This sort of understanding is further facilitated when you avoid over-engineering things. It is tempting to look down on simple solutions. Yet these are often the most reliable! Our ultimate product uses far more of these solutions than the complex, fallible ideas we often start out with. In iterating on the device further, we are curious about maximizing precision and consistency and minimizing size. Though this could be performed through experimentation, we want to give some attention to prior art. We are keenly interested in reading materials on the complex mechanisms we implemented (such as chambers and reloading, striking/launching projectiles, and springs in low-tolerance environments), and will seek them out before making any improvements. We also want Timotheé Chalamet to see it and use it in the promo for his new movie</a>. I Think This Might Have Been Epistolary 2025-12-15T00:00:00+00:00 The Making of Prince of Persia by Jordan Mechner. Finished on November 29, 2025. Rating: ★★★★☆ Stripe</a> is bursting with side quests these days. In addition to providing the “financial infrastructure to grow your revenue” that we all know and love</a>, they run a print magazine</a> and a publishing company</a>. Scrolling through the website of the latter and feeling jealous of their stunning visuals, I stumbled on a book about videogames. Or rather, one videogame</a> in particular. Having played the first level of Aladdin on the SNES and enjoyed the movie adaptation of Prince</a>, I thought I might give it a read. The journal entry form made it an easy read, and the concision of each note reminded me—for my own practice—that you don’t have to write a novel for every entry to build a compelling narrative or get meaningful content onto the page. I found Mechner’s attention to craft, especially now in an era where the love of the game has dried up in the shadow of grifters and generic fortune-seekers, inspirational. A random sample: The PC version is maybe 50% of what it should be. I can’t even tell these guys what to fix… it’s a million little things, and they’re just not up to the hassle. That kind of attention to detail is why the Apple version took me two years. This version is probably the best I’ll ever get out of them. Oddly enough, this makes me more psyched to do the new game. It reminded me why I’m good at this – of what I can do that others can’t, or won’t. </blockquote> His discussions of money are bizarre and interesting. Either things are very different now, or I have underappreciated the value of intellectual property and the amount of work people will do for you with the promise of an eventual commission. Likely both. In particular: he makes a chunk of change selling the right to port his games to other platforms. Modern game engines abstract away much of this process, and even without this help there are only three desktop platforms and three consoles worth thinking about—strictly less than in his day. But I will keep in mind the value of an agent shopping your work around, should I ever decide to do something like screenwriting, as well as the benefits and drawbacks of working on a personal creative project within the bounds of a corporate partner (as he does with Prince of Persia via Broderbund). It’s a thoroughly Bay Area book as well. To interpolate a seasonally-appropriate phrase: reading this, I am presented with a narrative of events from his time but in this place. It’s trippy: he eats at Marin Joe’s! I suppose I should have assumed from the general vibe of that restaurant that it had been there forever. To top it all off, there’s this lovely paragraph: Man, it’s great to be back. When I got into my rented car and hit Highway 280, I laughed out loud, I was so thrilled. The trees, the colors, the quality of the sunlight… San Francisco is the most beautiful city on the planet. </blockquote> Should Mechner ever feel compelled to release a sequel from the period in which Prince was being adapted to a feature film, I would immediately give it a read. For someone so in love with both making games and writing films, I’m sure it was a crowning acheivement. Developments on the Git Forge Horizon 2025-11-18T23:00:00-08:00 Steve Klabnik, as found on Lobsters</a>: some of my friends have started a new company, ERSC</a>, which is going to be building a new platform for developer collaboration on top of jj</code>. </blockquote> GitHub delenda est. I currently</a> have no self-hosted solution</a>, so I’m stranded in the digital bazaar. The above seems neat but far away. Tangled</a> looks promising: it hits the self-hostable but accepting broad constributions sans-account bullseye, and manages to do so without even resorting to Drew DeVault’s insistence that email can work again. But the frontend needs work and I don’t want to push via SSH. ★</a> Vault 2025-11-13T11:30:00-08:00 For the second project of ME102: Foundations of Product Realization, you are assigned a mechanism and instructed to build something with it. My group was assigned a 19th-century spiral feed system</a>. We reimagined this historical mechanism as part of a vault door locking system, using the spiral’s gradual rotary-to-linear translation to control four locking arms. This new use-case captures the physical character of the original mechanism while taking it in an unexpected new direction. And it looks fantastic: Before any physical experimentation, we sketched and ideated a number of possible mechanical arrangements and functions, with ideas ranging from tidal and wind movement to drills and water wheels. Ultimately though, the idea to transform the drill arm into a lock bolt came from a visit to a restroom in the Cantor Arts Museum</a>. There, we saw the horizontal motion of the stall’s bolt directly creating a fixed rotational change in the external signage (red/green correlating with the stall’s occupied status). Our mechanism performs that same transformation in reverse! We made two prototypes over the course of the project. The first was a 2.5D laser-cut mock, which we used just to get an intuitive sense of the mechanism. For our second iteration, we migrated to the 3D printer. This version was our first vertical spiral, and the plastic significantly reduced friction over the earlier wood model. It sported a wide base for stability and a LEGO-inspired plus axel to latch the rotational arm to the mechanism. After these two functional experiments, we refocused our work toward building for a specific use-case: the vault. Our final design built on the successes of our second prototype while upgrading to mixed fabrication methods. This iteration combined a 3D-printed spiral and wheel components with laser-cut acrylic plates and metal fasteners. The inclusion of real hardware allowed for smoother motion, adjustable tolerances, and a sturdier overall assembly. This was a professional device. Our final CAD model combines four different materials, three manufacturing processes, and two complex joints in one digital source of truth. The device just barely fits inside an 8“ cube (in case you have one of those laying around). Two interesting notes that might otherwise get missed: We designed our 3D printed parts in the center around the 2D constraints of the laser-cut acrylic, which we stacked (0.236” per layer) to create the transparent walls.</li> Embedded in a layer of the door’s acrylic and the plastic wall of the safe body are a set of magnets! These add a subtle click that makes the door quite satisfying to open and close.</li> </ul> This project built on a lot of the previous knowledge we had acquired prototyping in ME102. For example, before cutting the final acrylic parts to construct the wall, we performed a number of test cuts to examine the tolerance. This made working with fasteners much easier, and is a practice we adopted after our introduction to friction fits in the first project. Looking toward the horizon, we are of course interested in improvements to security. That would mean replacing our plastic parts with metal—so that you can’t defeat our vault using blunt force—as well as an actual locking mechanism, which we deemed out of scope for this version. Perhaps you’ll hear more about that from me and my incredible teammates (Vivek</a> and Trun</a>) at some point in the future. How to Format a micoSD Card for the Nintendo Switch 2025-10-29T00:00:00+00:00 I burnt vastly too much time on this today, after having already figured this out a week ago, so here it goes in writing. In one sentence: format everything in exFAT and use the Master Boot Record partition scheme. Spoiler for my Y3Q1 quarterly reflection: this year my dorm room is awesome. One of the new goodies that we have is a TV, and so I brought my basically unused Switch out of retirement to see if console gaming might be entertaining now. It is! We’ve been playing the hits (Mario Kart, Overcooked, etc.) and some indies</a>. And as the Switch’s popularity among my friends has grown, so has the demand for more games. This, of course, presents an issue: games take up much space, and the onboard storage of the Switch is not infinite. So I ponied up for an upgrade to my existing external storage (8GB) and bought a 128GB microSD card off Amazon. Of course, I had at this point forgotten how I set up the first card. I took a big, painful detour trying to create a disk image of the old card with Disk Utility, SuperDuper</a>, CCC7</a>, hdiutil</code>, and dd</code> before realizing that I could just copy the data like any other files (Nintendo conveniently puts everything in one top-level folder). After that, the partition scheme was just trial and error. I’m looking forward to the expansion: my first round of purchases includes Stray</a>, Factorio</a>, and A Short Hike</a>. If the habit proves more permanent, I’m sure my now-dormant Backloggd</a> will make its way onto my homepage. Two Pockets 2025-10-09T22:30:00-07:00 There is an ancient Jewish story told about a wise man who used to walk around with two pieces of paper, one in each of his pockets. On one piece of paper was written “the world was made for me” (Sanhedrin 4:5), and on the other, “I am dust and ashes” (Bereshit 18:27). The ability to hold on to two mutually contradictory truths equally tightly is a core aspect of Judaism, in my mind, and a skill I find myself relying on more than one might expect. As such, I wanted a physical reminder of this ancient wisdom. I modeled two bracelet charms in Shapr3D</a>, tweaking them until I was satisfied. I printed out a number of tests at different scales at the Recurse Center</a>, which allowed me to play with the size on my wrists as well as refine the text so that it wouldn’t come out a muddled mess. None of the prototypes came out perfect, but I hoped that the increased accuracy of resin printing would make the cast come out a smidge cleaner—and in any case, I had resolved to make an attempt at realizing the vision. There are a number of print-on-demand services. The first one I encountered was Sculpteo</a>—but I didn’t ended up using them, as my choice was guided primarily by material options. I wanted something gold (to match my existing jewelry) but without paying the hefty fee that entails. This is seemingly a niche desire. So: plated silver and Castimize</a>. The metal pieces are perfect—honestly I’m a bit surprised! They’re so good that I’ve almost entirely left my watch behind, and picked up some new work on the side</a>. I’m not in love with the yarn bracelets: I’m afraid that salt water and time will make them brittle and they’ll snap, so I’m in the market for a good set of chains. But what I’m happiest about is that I have made art for its own sake. And not just any art: Judaica! Text Conversation, Transcribed 2025-09-25T23:35:00-07:00 The Counterlife by Philip Roth. Finished on September 17, 2025. Rating: ★★★☆☆ i just finished the counterlife Oh wow Thoughts? I honestly don’t remember it all that well anymore </blockquote> i understand that the whole point is that the book was an interpolation of the same story in many different interpretations but it made it a very confusing read What was the conflict between the brothers One of them was having an affair? </blockquote> they were both involved in various differing affairs i would say that the main conflict is about conceptions of the “good life” the playboy vs the father, the jew and the heretic, the diaspora and the independence I should read more of the Zuckerman books </blockquote> On the Recommendation of a Friend 2025-09-25T23:30:00-07:00 An Odyssey by Daniel Mendelsohn. Finished on June 12, 2025. Rating: ★★★★☆ I should really read the source material. But a touching journey nonetheless, which I am glad the author has shared. Cloudlab 2025-09-25T17:20:00-07:00 In the vast apparatus of the internet, I have my own little hideaway: an Alpine Linux</a> box hosted on Hetzner</a>. The website you’re reading lives on this very server, alongside numerous other personal services. Running these myself allows me to be completely independent of major service providers like Google, a choice with serious benefits for my privacy. It has also helped me accrue real-world experience with system administration—and the crises</a> that occaisionally entails. Finally, owning a fixed amount of hardware means that there is zero marginal cost for additional services crammed into my alloted resources: this means the monthly cost of running my many programs remains well below market price. My server setup would historically have been best described as a rickety shack, barely standing, upon which I would occaisionally try and haphazardly tack new services while hoping the whole thing wouldn’t come crashing down. This was sufficient for at least five years</a>. But at Recurse</a>, I became fed up with the status quo. I explored a few radically different systems. Frank</a> pointed me in the direction of Talos</a>, and I played around with setting up a one-machine cluster. But Josh</a> asked me if I really needed all that baggage, and truth be told I didn’t. So I returned my focus to trusty Docker Compose</a>—and toward minimalism, pattern, and clear structure. That structure employs a number of axioms: “All services are run in containers” (file storage</a> and backups</a> having been the big exceptions previously), “configuration and secrets</a> get committed”, “no building images on the server”… Really, I can’t say it better than I already did</a> during development, when I wrote long commit messages and thorough documentation. I also upgraded my server during this transition. I investigated a move to the United States, but found that resources prices here vastly outpaced the rest of the world. So I made a different move: to the ARM architecture. Miraculously, I now get double my old resources at the same price. More than enough for my email, calendar, websites, and anything else I’ve thought of throwing at it over the years. There is still room for improvement in this setup. As part of a renewed commitment to elegance, I dropped my janky self-hosted git</code> setup and do not yet have a good alternative</a>. And getting the server prepared entailed some minor setup work which should be controlled in a declarative and versioned fashion alongside the rest. But that rabbit hole</a> seems too deep for even me at the moment… It will have to wait until next summer. Check out the code</a> for your own self-hosting inspiration! Stickies 2025-09-17T21:00:00-07:00 When Apple announced that visionOS 26 was bringing wall-mounted widgets to their face-mounted headset, I was very excited. For one thing, I am most interested in the device for its ability to create interactions between the digital and the physical worlds—which this feature enhanced in a particularly compelling way. But also, I was excited because as I looked around at my walls I saw at minimum 100 sticky notes across every previously-blank surface. It was time. There were many peculiarities in its development, not the least of which being that I don’t own a Vision Pro. By the time I tested it on-device (thanks to generosity of Jesse Sheehan</a>), the app was already feature complete! In this phase, I ended up making significant adjustments to accomodate what the radically different controls actually feel like—your eyes are simply nothing like a mouse and keyboard. That sounds obvious, but it’s a fun challenge to translate that into tangible design choices. Side note: the walls in the simulator don’t count for sticking widgets on, and .toolbar</code> is actually .ornament</code> on visionOS. For anyone else embarking down this road, I hope I’ve saved you a handful of hours pouring over Apple’s documentation. With Stickies, you can create and edit boards of different sizes in the app and pin them to your wall as widgets. This comes with a number of benefits beyond being pretty cool to see: your notes will never come unstuck, you can edit them infinitely, and perhaps most excitingly you can pin the same board to multiple locations and keep them in sync. Stickies is coming soon to an App Store near you! In the meantime, feel free to check out the code</a> for yourself. Cordially 2025-09-17T20:11:00-07:00 The de-facto standard for event invitations at Stanford is Partiful</a>. It’s a great platform: it looks cool and makes gathering RSVPs easy. But what if you want less of a party vibe, and more of an air of mystery? Less “join the herd” and more “you, individually, have been chosen.” For that, you might need Cordially</a>. There was of course a time before Partiful, when Paperless Post</a> was dominent. And they got something right: you would receive a skeumorphic envelope in your inbox, and there was a fancy letter-opening animation on their website. That felt special. So, after throwing a handful of parties myself and hand-writing an awful lot of invitations, I decided I could take the best ideas from pop culture and these competing platforms to make something better. Then I decided to wait several years until I was at the Recurse Center</a> and staring at a deadline for the Palantir Build Challenge</a>. And thus naturally, I ended up building my fancy-shmancy invite platform in a mad two-day sprint on top of a data platform best known for its military-industrial use-case. With the strict time constraints, I didn’t get to make Cordially the visually stunning platform that I had initially envisioned. But I learned loads: I built the app in Python’s FastHTML</a> as another waystop on my roving quest to find a good web framework, and I was able to concretize Palantir’s product in my mind, where otherwise it would just be a swirling, intangible, mildly sinister idea in my head. One day I may rewrite Cordially to be gorgeous and production-ready. But for the time being, you will have to be satisfied with the code</a> and a video walkthrough</a>. Rapid Riter 2025-09-17T20:00:00-07:00 I enjoy many things. Two in particular are alternative display technologies</a> and using the right tool for the job. These combined in a truly joyful project when I wrote a script for the Rapid Riter. When I first walked into the Recurse Center</a>, the large red LED matrix mounted on the wall caught my eye immediately. I’ve been in a good number of hacker spaces, all strewn with exciting toys in various states of decay—this was the first space I’d experienced where the toys worked. I knew then I would have to make something for it. This, as you may have surmised by the label, is the Rapid Riter. To get my program working, I orchestrated a series of tools in a beautiful symphony. Or at least, that’s how I felt writing it at midnight on a Saturday with wine glass full of water. The graphics are written in p5.js</a>, an animation DSL, and uploaded directly to the Riter through Greg’s</a> developer portal. The number is calculated via the Zulip</a> API, and served from val</a>—which I had heard of as a long-time follower of Tom MacWright’s</a> blog but never used before. Writing little type-safe scripts that are live instantly, with proper secrets management? I loved it, and will be tucking it away for future projects. When my script was done, I presented it during demos at Recurse</a>. You’re encouraged to send daily “check-ins” during your time there: the final number I missed was much higher than the number above. If you want to see how high, you’ll just have to apply! Night 2025-06-21T20:15:00-04:00 Night by Elie Wiesel. Finished on June 20, 2025. Bad Book Good Ideas 2025-02-22T15:15:00-08:00 How to Do Nothing by Jenny Odell. Finished on January 07, 2025. Rating: ★★☆☆☆ I was assigned Jenny Odell’s How to Do Nothing as part of the Three Books</a> program. I didn’t take a COLLEGE course in the fall</a>, so I have relatively less of an idea if the program was taken seriously, but from my perspective it was comprised of a) an email that told me it existed, b) a free book that was—I think—presented to me on my arrival to campus, and c) no further mention of it ever. A long while later I decided to read it. There are discussions of great merit in How to Do Nothing—but they are both poorly executed and tied to other beliefs that I disagree with, which ultimately has created a work that mirrors the program that recommended it: a core of goodness enveloped in the overgrown vines of bad. To address the former portion of my critique, I will turn to the common adage that the hallmark of true understanding is being able to explain something simply. Jenny Odell has not done that here. This book is a wandering text with no driving argument. Whatever tidbits of knowledge are to be found within it are sprinkled throughout at random intervals, and then repeated, each time as if she has forgotten about any previous mention. I felt as if I was reading a first draft the entire time. To the latter half of my critique, what arguments are articulated in How to Do Nothing are themselves something of a mixed bag. There are portions that I resonated with strongly: I am personally unsatisfied with untrained attention, which flickers from one new thing to the next, not only because it is a shallow experience, or because it is an expression of habit rather than will, but because it gives me less access to my own human experience. p. 119 </blockquote> Poswolsky writes of their initial discovery: “I think we also found the answer to the universe, which was, quite simply: just spend more time with your friends.” p. 34 </blockquote> But they are often followed up with just terrible addenda. I suspect it comes down to the following fundamental disagreement between myself and Odell: …I find existing things infinitely more interesting than anything I could possibly make. p. 5 </blockquote> This is what leads her to follow up potentially interesting suppositions: The first half of “doing nothing” is about disengaging from the attention economy; the other half is about reengaging with something else. p. xvii </blockquote> With nonsensical conclusions: That “something else” is nothing less than…. bioregionalism… p. xvii </blockquote> This exchange in particular is typical of a type of thinking in the book that I do not understand in the slightest, which prizes non-humans over our own species and asceticism over greatness. That stance smacks of a particular branch of Christian thought that sees “lesser” states as inherently more “pure.” The plant is prized over the person because it is incapable of sin. The mendicant over the industrialist, because why seek out anything in this world when the world to come is what matters. This view does not move me. Let all of the righteous acts that Odell proposes be done at scale as a testament to our glory. One thing I have learned about attention is that certain forms of it are contagious. When you spend enough time with someone who pays close attention to something… you inevitably start to pay attention to some of the same things. p. xxiii </blockquote> There is no such thing as a clean break or a blank slate in this world. p. 53 </blockquote> “The explorer who will not come back or send back his ships to tell his tale is not an explorer, only an adventurer.” p. 55 </blockquote> There are no limits to what we are capable of. Attention is the tool by which that potential is made manifest. Blind Date A Book 2025-01-19T17:20:00-08:00 Blind Date A Book</a> is part ecommerce website and part art project. I first encountered its physical counterpart at a Barnes and Noble some five years ago: I remember thinking to myself even then that it was a perfect candidate for digitization. At long last, using the approach to making I picked up in DESIGN11</a>, I’ve hacked together a boutique version of this experience accessible for anyone with an internet connection. And at the Recurse Center</a>, I overhauled the site for a more fun, correct, and cost-effective developer experience. This is the story of how I became the proprietor of my very own bookshop. At first I was confused by the lack of digital equivalent to the physical stands. Surely ordering books online is a solved problem. I mean, that’s what Amazon was founded for—and a custom frontend that just displays things differently seems a trifling project. Turns out it’s hard to buy things with code. So it actually came out a bit more personal</a>: A list of books chosen by people I care about.</li> Descriptions they hand-wrote.</li> A funky little website.</li> Me going to buy the real book and sending it you myself.</li> </ol> More of a sensitive touch, perhaps. The tech stack of the site bears further analysis. I’ve long puzzled over the best way to go about building web projects, going so far in my complaints-sans-solutions as to coin Benji’s Rule of Frameworks</a>. For CLIs</a> and iOS/macOS</a> apps, the answers have been easy: but building for the internet has always seemed more experimental. In the first iteration of this project, I tried out a framework called Web Origami</a> and ran the commerce via Shopify. It wasn’t all sunshine and rainbows—the product was far too JavaScript-y for me—so I continued poking around until eventually I found Gleam</a>. Building in Gleam is weird and exciting, which is exactly what I’m looking for! I like it so much that I gave a talk about it at Recurse, the slides and notes</a> for which are available to read here. I also moved from Shopify to Stripe, which has allowed me to a) have a single source of truth by using key-value metadata in Stripe as a database and b) save $500/year by paying a percentage of profits instead of a fixed and pricey subscription. Check out the website</a> for yourself to find your next read and explore the code on GitHub</a>! It's Time to Build New Hardware 2025-01-14T17:57:00-08:00 There is a feeling coursing through the populace.</a> It is just now bubbling over, spilling through the cracks and taking its many shapes. The world changed forever with the introduction of the iPhone in 2007—it is a fool who thinks it can’t change again. I am a fool who thinks it is going to change very imminently. Let’s be honest, making a new primary device is effectively impossible. We probably won’t see a smartphone-killer or a computer-killer ever. Louie Mantia, Jr.</a> </blockquote> These two sentences represent the prevailing wisdom on mobile computing. Not just that consumer hardware is hard—it is, with even great successes like Pebble</a> eventually collapsing—but that it’s not worth pursuing: that we’ve already figured it out. It’s the most dominant form of technology worldwide. Billions of people make their digital homes on their phones. Surely that means the field is done. To my mind, this has always sounded ignorant. Look at your phone: it’s a rectangular screen. What is this optimized for? What does this form factor make easy? The screen is a blank canvas. It enables Jobs’ vision of a device that doesn’t need physical updates to gain new functionality. But it’s also profoundly lazy. It is the lowest common denominator of interfaces. A game for everyone is a game for no one. Arrowhead</a> </blockquote> The screen is also uniquely condemnable for the way it inherently hijacks our attention. Our brains are ill-prepared to encounter the light of a thousand suns or colors more vibrant than any Amazonian tree frog, but that’s what we see when we check our phones at 3am. Try and have a conversation with a phone in your peripheral vision. Play a video with the sound on mute: can you avoid distraction? Alternatively: could you actually recognize the contours of your palm? The common turn of phrase would suggest this to be the most familiar thing in the world. I suspect we all know our home screen much better. *wakes up and looks at phone* ah let’s see what fresh horrors await me on the fresh horrors device @missokistic</a> </blockquote> At the same time as I am thoroughly steeped in technology, I hate it. I want to maximize the amount of time I spend present with the people I love: my friends, my family, random kind and exciting people. The time I spend “in” a device—the fact that such a turn of phrase even exists—is despicable, in my eyes. And I’m not alone in this: If you’re looking at a phone more than someone’s eyes, you’re doing the wrong thing! Tim Cook</a> </blockquote> The status of this crusade, as it stands, is a bit of a mixed bag. On the one hand, there are a metric ton of reasons to despair: there’s a whole generation of “iPad kids” out there who have grown up breaking Cook’s ironic axiom above. I myself just quit a Reels addiction. The status quo is riddled with the ills of our devices, from the wasteland of social media to the phantom notifications plaguing our pockets. However, there is some cause for hope and excitement: that feeling I mentioned earlier. People don’t just live amidst all the garbage—they’re starting to feel it. Some of these people are seeking to do something about it all. Throughout the industry there are new takes on what the coming device paradigm might look like: new minds finally tackling the question of what it should look like to live alongside our technology. Obviously, none of them have come close to supplanting the phone or the laptop or the screen at all. Not just yet. But I hope they will, eventually, and I want to be a part of that. There are numerous significant and cool initiatives in this area these days. Humane</a> has been universally panned but I’m extremely interested in rapidly developing their laser display technology as an alternative to traditional display panels. Origami Computing</a> is actually the reason I’m writing this post, because I promised Sarvasv I would do so over six months ago and I want to reply to his email. I steal his line about the love-hate relationship we have with our devices all the time. The Apple Vision Pro</a> may initially seem like its going the exact opposite direction of what I’m advocating in this post, but if you look at it closely you realize that it too is seeking to eliminate the role screens play in our lives today. tinyPod</a> is such a compelling reason to buy an Apple Watch that I might actually do it. I own the reMarkable 2</a> and use it every day for school: for years it replaced all my paper use, though I’ve built up a nasty habit of of sketching and journaling on dead trees again. I met an angel investor in Daylight</a> on a flight to New York—I’m most interested in their tablet because it runs full Android, and there are good apps I use on my phone that I wish I could use on an e-ink-ish display. I have similar feelings toward the Palma</a>. I’m not sure that Freewrite</a> belongs among this new wave, because they’ve been around for a while, but it is yet another e-ink device that I want to buy. Limitless’ pendant</a> and Avi’s friend</a> strike me as products similiarly oriented around the question “what if a microphone was maximally intelligent?” Spatial Pixel</a> gives strong tldraw</a>/todepond</a>/Ink & Switch</a> vibes, as well as relying on cool projection technology that we’ve previously established I’m a fan of. The Oura Ring</a> is not as much of a general computing device as the rest of these, but it has the great quality of delivering utility</a> to me on a daily basis while not having a screen, so it has won a spot. And frankly: Roll Call</a> was inspired by this very scene/movement and was explicitly</a> an expirement in alternative display technology! Gruber wrote the sharpest critique of these sorts of efforts two years ago when the Ai Pin was first announced, in a post entitled If You Come at the King</a>. In essence, he says that Humane’s founder, Imran Chaudhri, may be right that our phone addictions are sad and problematic. However, he points out, people don’t care. The objective truth has no bearing on the fact that everyone loves their phone. Addicts love their compulsions, even when they know they’re self-destructive. I agree with Gruber. But I think that the need is too dire to simply give up. We need a Pareto improvement</a> in our mobile devices: new digital companions that are not only healthier for us, but that we love even more. People won’t change their behavior for abstract concepts like well-being, but they will absolutely change their behavior for better. And something better is possible. We do not have to take the good with the bad. A more humanistic future is out there: one with technology that is helpful and ubiquitous but totally dissolved into the background of our existence. We have designed our present, and we can design a future with all of its benefits and none of its downsides. We must right our priorities by building a better system on the back of the one that came before. There is no area of more impact than hardware, and no better time to experiment</a> wildly</a>. It is time to build new hardware. It is time to turn our focus to the world we were given. Chronological Sort Considered Positive 2025-01-12T16:08:56-08:00 Robin Rendle, himself responding to another post on “digital gardening”: How many bangers on this here website are lost to the feed of new stuff and are only hard to find because of the chronological list I’ve slapped together? There’s so much stuff that I’d like to revise or tweak or scratch out—but keep a record of that change of mind for the future. Wouldn’t that be so neat? Going back over twenty years of work and watching your thought process adapt over time? </blockquote> That record of evolution—of a minding shifting and evolving—is in fact chronological sort’s raison d’être. If you change your mind about something you’ve written, if you’ve grown in your thinking, or even if your position has concentrated into a stronger version of what you already believed: write another post. I would propose that contrasting those posts over time will probably provide a better understanding of one’s “process adapt[ing]” than just converting your personal website to a wiki. ★</a> Stanford Quarterly Reflection (Y2Q1) 2025-01-04T00:00:00+00:00 Last year my friends would often pointedly remind me that I was a freshman. So it’s been a little weird this quarter no longer holding that title. However, as far as beginnings go, I’d say it’s off to a fine start. This quarter is what Stanford is supposed to be. I learned. A ton. I was enrolled in: MATH 21: Calculus</li> COMPLIT 37Q: Zionism and the Novel</li> SUSTAIN 345: Sustainable Societies Lab</li> DESIGN 11: Visual Thinking</li> CS 148: Introduction to Computer Graphics and Imaging</li> </ol> The extent to which I was actually in MATH 21, despite my enrollment, is somewhat questionable. I attended very few lectures, and I think possibly two discussion sections. On balance, I think this was alright. I obviously didn’t derive as much as I could have from the course (pun intended), but I performed adequately and had more time to devote to things that interested me far more. This is despite my liking the professor: when I did manage to show up for lecture, I felt a bit like I was participating in the mathematical world of yore. The class did manage to have a statistically significant impact on my sleep though! See if you can guess when the problem sets were due: Zionism and the Novel was a real treat. The syllabus</a> was so interesting that I felt guilty any time I had to skip a reading, and promised to catch up on it at a later date for my own good. I’m still working on that. I also ended up sharing a lot</a> of my writing</a> from the course on this site, which is a very nice plus. The Sustainable Societies Lab continues to be a blast (I participated in another incarnation of it last spring</a>). We heard from a number of deeply knowledgeable speakers from a broad range of fields. I was blown away by Marc Tarpenning, which also made its way</a> into some posts</a> here. I also strongly disagreed with the founder of Beyond Meat, and will attribute any future failure to his attitude and any future success to pure luck. The latter half of that sentence is a joke. DESIGN 11, my old nemesis</a>, was possibly my favorite class. It was certainly the most impactful—I think I will never look at the world the same way after taking it. The class forced me to become comfortable sitting in the liminal space of making things. There are very few ventures that you can be sure of before embarking upon them. Most entail some space in between, where each step from the start takes you further toward your ultimate destination. Some steps are misplaced, mistakes that you have to recognize and adapt to instead of abandoning your progress and starting again from scratch. This process must feel like home to you if you want to make something. The class instigates this change in its students by forcing you to do the things you’re bad at, over and over again. First the stakes are low: building little one-off thingamajigs and prototypes. And then you’re thrown directly in the fire and made to combine those little skills you didn’t even realize you were learning to make something real. We did that three times: carving and constructing an automaton</a> from foam core, abusing Figma</a> to make a functional digital vending machine</a>, and the final open-ended project that ultimately coalesced into Roll Call</a>. I am grateful to Patrick Fenton, who made this a comfortable place to grow, and to Vivek for pushing me through the course when I needed an extra shove and making building things a total blast. Also I sort of learned to draw: CS 148 was another fun class—I took this one at Naama’s behest, and found the final project</a> in particular a good opportunity to implement the techniques for making that I had picked up in DESIGN 11. The class was scheduled at the same time as DESIGN 11, so I never actually attended a lecture, but I’m theoretically a fan of the professor because he has two Oscars for inventing the math behind CGI and was a professional bodybuilder</a> in his youth (scroll to the bottom of that link). It had been a long time since I worked with Blender</a>, so it was fun to get back at it! I felt a lot more confident building things independently this time around, which I would attribute to my growth in many contributing areas and skillsets—quite satisfying. This quarter lacked a structured avenue for me to engage with political thought. Zionism and the Novel came close, and I certainly engage seriously on a myriad of political issues in discussions with my friends, but I still felt something of a void. All of which is to say that I declared a PoliSci minor on the International Relations track, which will hopefully keep me entertained in this respect going foreword. They also gave me a funky shirt when I declared, which Design didn’t, so extra points to their department. This was the quarter of Junch and the Sleepoverers. Odin and my membership in both groups means that we technically won the quarter. Every week was marked by me getting progressively more excited until I actually arrived at Junch and then seven days of comedown until the next one arrived. Our more ambitious schedule for this upcoming quarter is going to be epic; I just hope Naama and Jordan can keep up. The Sleepoverers were my people for these past few months. If I had downtime, I was with Nageena, Lulu, and Daniel. What a squad we were—the half of our contingent that are going off to Paris this upcoming quarter will be dearly missed. My roommate this year is no longer random: I’m living with Vivek! Who would have thought that we would be living together, four years after we first met each other and thought the other one was super weird. Our room is a total blast. It really is as simple as Vivek told my dad: “we like the same snacks.” Here’s to next quarter, when my morning classes might finally force me to adopt his sleep schedule. I’ve created some clean delineations above for the sake of the post, but of course reality is much messier. Jack and Zach, besides rhyming, are pillars of my existence here at Stanford. I look forward to more GSB library study sessions and rapid costume creation, respectively. Adri, Deit, Reis, Kato, and Bradley welcome me back to the Phi Psi house like a returning champ every time. Sophia is a boundless source of creative energy for activities and I am very sorry for spreading the false rumor that she broke her leg playing laser tag. Vedant and Hasan show up all the time in such unexpected places—I look forward to the next time I’m summoned. Nate continues to outlift me in the gym, and Trun continues to do things with computers that shock and surprise me. I am routinely impressed by Daniel’s FLiCKS. This was a different quarter. I enjoyed it. This next quarter, I’m going to build on the positive institutions that were seeded here—continuing my perpetual search for better and more. Academically, I’m going back to the basics with a lot of foundational requirements. I’m looking forward to it. In the meantime, I’m going to go celebrate getting this out on time. Stanford Quarterly Reflection (Y1Q4) 2024-12-25T00:00:00+00:00 I continue to set new records in how long-overdue something can be. Nevertheless, I refuse to let this summery meditation slip away. I have written already at length about my internship</a>. This can, for the most part, take the place of the usual section on academics. Too much time has passed for me to recall any other projects I pursued that might be grouped in with my work to some alternative “productivity” category. I met with Adam Nash</a>? But that wasn’t really work. It’s hard to say. I lived in two apartments. The first was in the Marina: it had a little backyard that was wonderful to sit in, and heated floors in the bathroom that made walking around in the early mornings significantly more pleasant. The second was in Cow Hollow: it was masterfully decorated, had the most excellent kitchen I’ve ever lived with, and was situated literally one block from Union Street. I think I’ve come to witness, within my lifetime, the fall of Union Street and the rise of Chestnut, and now the rebirth of Union as Chestnut grows stale. Which is to say that though both were lovely, the latter apartment was a far more lively neighborhood. Living somewhere, for the first time, on my own and with a kitchen, I was able to participate in perhaps my favorite tradition: Shabbat. I hosted about seven in all, dashing out of work early on Fridays to the grocery store and then to the stove to throw together the week’s menu by the time the guests would arrive. I had high-minded and silly inspirations</a> in mind while throwing them, and they largely lived up to those expectations. I placed an inordinate amount of weight on and thus care toward crafting cover images and descriptions for each Partiful</a> invitation. And I was honored that people came! To have had you eat of my food and drink of my drink and be together in one place, laughing late into the evening, was the greatest reward and made me quite happy. I am also particularly fond of one story, in which Zach arrived an hour late—he drove up from Stanford—and came through the door to hear, “You’re just in time! We’re bringing out the food now. Come sit!” I was not always the most punctual. I also got away ever so briefly to New York, Oregon, and Mexico. I spent the first running around seeing everyone who was available—which meant quite a few people, given that Stanford basically splits in two between San Francisco and New York for the summer. In Oregon I lived in lakes and rivers, which was great because it might be impossible to have a bad time in the water. In Mexico, I had costras de queso for the first time, which was a life-changing experience I can highly recommend. The ant larvae at La Gruta were a life-changing experience I am less enthusiastic about. Of course, people were the most influential in making my summer wonderful. My whole family came to every Shabbat, baking the challot and creating amusing Partiful personas. Daniel and Emily were on the couch of my apartment waiting for me when I came back from my first day of real work ever. Daniel also hosted me in New York—the city where we first became friends—which was extraordinarily kind. Davita, Hasan, and Zach were the most consistent Shabbat attendees and thus gave shape to the most influential institution of those months. Odin crashed on my couch for ten days during the one period in which I was working remote and proceeded to induce The Most Fun and get me on creatine. Kelly and Joven I will thank for foolishly jumping on a plane with me time and time again, and Longhair for hosting us (and calmly weathering anaphylaxis). Trun for our meal at Lavash. Vivek and Nate for tolerating my god-awful tennis. Sophia for her rooftop DJ set, and Jordan and her mother for a great Effy’s brunch. Guy the locksmith for his locksmithing, though he will never ever read this post. Now I will immediately begin writing the next reflection, so that I can hopefully have it done before I return from winter break. If I’m lucky. Great revelry awaits! Biting the Hand That Feeds You 2024-12-24T00:00:00+00:00 Om, on his blog: Unlike search engines, AI platforms are built on precision and summaries. They’re unlikely to be a major source of traffic or advertising revenue. What about getting paid for the summaries OpenAI serves in response to prompts? These prompts will vary widely, as will the responses, breaking the traditional mass-media revenue model. As I explained in my “future of browser” article, information itself is being atomized, which will likely upend the web and media as we know them…. OpenAI, Anthropic, and Perplexity… won’t need the media for long. </blockquote> What does it mean to not “need the media” anymore? Chatbots know nothing of the observable universe—they are fed media, during and after training. OpenAI et. al. do not experience events as they happen. Reporters are on-site witnesses and produce original analyses that comprise both the news we consume and the content AI is fed. The internet was originally envisioned as a place for connection, collaboration, and discovery. But over time, it has been distorted by business models that prioritize engagement metrics over meaningful interaction. Discovery has long been the open web’s greatest challenge, with search engines turning it into an SEO game and social platforms creating algorithmic echo chambers. AI platforms are making discovery almost irrelevant. </blockquote> This is because the greatest challenge has always, in fact, been monetization. Before the internet, you could sell people the paper your news was printed on to fund its production. When the internet first arrived, it was a toy and money wasn’t a factor. When real people came online, money made everything bigger and better but also introduced ads, engagement, and the rest of that garbage fire because we’ve all agreed that web pages are free. If ChatGPT takes over the internet and we say that news companies can’t make money through their existing channels anymore, they will need to find yet another way to pay their staff. I hesitate to say that we are running out of business models, but we are certainly burning through them at an alarming rate. More importantly they all seem to suck. You stay still, but your AI agent goes out and fetches, distills, and synthesizes the content and renders it in whatever format you want — audio, video, or text. This is the future. </blockquote> Not if there is no content to synthesize, because they have been driven to bankruptcy. Mass media is an invention, not a necessary constant. If AI obsoletes the news by providing people with a mouthpiece that they prefer, without providing a backstop for the companies that produce its sources, it will run out of material to synthesize. ★</a> A Screed of the Old Style 2024-12-19T13:36:42-08:00 Israel Alone by Bernard-Henri Lévy. Finished on December 19, 2024. Rating: ★★★☆☆ I got this book (and had it signed) on Election Night. Lévy had come to visit the Stanford Hillel, to philosophize with students and watch as the votes were called. A student asked him at one point how he could be so certain that the West and its allies would triumph over the rising Axis of Resistance. I remember I was struck by his response: “Of course they will. It is good against evil, the evil must lose. It will be defeated. That is what happens.” This book will do little to convince those who may disagree with what it has to say. It presents few novel facts to those who share his positions. It is a position piece: a statement, for anyone who is interested, to read and comprehend the contours of Lévy’s beliefs. Two further notes: However much of a conversational tone I believe myself to impart on my writing, Lévy has vastly more.</li> I have never appreciated how vastly different the Francophone world is from its Anglophone counterpart.</li> </ol> I would be interested in hearing how Lévy’s thoughts have evolved in the time since the publication of Israel Alone. It is not quite that the book is outdated, but with the pace of change in the Middle East today, there is certainly much more to discuss. Roll Call 2024-12-17T14:52:00-08:00 Roll Call is a jury-rigged receipt printer that sits on my desk and dictates my morning routine. Every morning, it prints out a briefing of all the critical information I need to catapult me into the day that awaits. Phones are great—but I don’t like checking mine when I wake up. I want my brain fresh and undistracted. I first ideated what would become Roll Call on a post-it note on the wall of my dorm. When the final project of DESIGN11</a> rolled around and they asked us to “fix a bug in our lives,” I knew I had to take advantage of the opportunity. After a flurry of on-paper ideation, I began with the software. It’s all open-source: the bulk is a couple hundred lines</a> of bespoke Python, supported by an iMessage library I customized</a> to allow me to determine if a message was unread. I call out to a few startlingly</a> good</a> libraries to fetch the data that will go on the receipt: the day of the week, the weather, my Oura Ring stats from the night before, unread iMessages, my todo list, and my calendar events for the day. When that was done, I called every single possible retailer in the Bay Area who might sell me a receipt printer—no dice. I even asked a couple restaurants that I was eating at if I could pawn theirs, which was similarly unsuccessful. So I ordered one on Amazon (guided in my purchase by this blog post</a>), married it to my code, and have been using it every morning since. Roll Call is one experiment in my long crusade to replace the phone. Our current era of mobile computing has brought benefits that are difficult to fully comprehend. But they have come along with deep ills. The combination of hardware and software in Roll Call has allowed me to entirely eliminate my phone from this component of my daily routine. In this slice of my life, I experience all the benefits of technology with none of its downsides. However, this only works because I can still rely on my phone during the rest of the day. I can replace my phone with a receipt in the morning and feel better because of it, but later in the day I still need to fall back to the harsh light of the screen to respond to messages, mark todos as complete, and more. What the Roll Call experiment allows us to do is free our minds to imagine what removing the phone might look like from the rest of the day. This part? It’s already taken care of. We’ve solved for this use case, one that is normally solved with our phones. How can we solve for the rest? We have an alternative way to read data, how can we rethink writing? How do we redesign the rest of our phone-based experiences to avoid falling into Instagram and a hundred other things that melt away all the presence and the possiblity of our beautiful lives? This part of the pressure is gone, now let’s deal with the rest. A theoretical, optional pitch that loses people but I think is conceptually neat and cool </summary> Roll Call is paper, with a digital brain. Every morning I look at a slip of paper—for all intents and purposes the same piece of paper—and instead of remaining static it autonomosly reflects the changing state of the world around me. The state of the world exists in a Platonic ideal somewhere in conceptspace. Nowadays we represent all that information about our lives, that typically is just held in our heads, in various databases. That information is displayed back to us in our calendar apps, and todo lists, and all the rest. Those apps transform the database's zeroes and ones, which are themselves representations of ideas, into a form that we view on our phones all the time. Roll Call's receipts are equally valid transformations of those zeroes and ones. </details> As a last component of my time in DESIGN11, we made little advertisements. You can check out mine here on YouTube</a>! Remind me never to film when I have a stuffy nose ever again. Iteration continues. One random day at the Recurse Center</a>, my friend Frank</a> brought in a small device that caught my attention—a receipt printer, in miniature! After one glance at its cute cat visage I knew I couldn’t pass up the opportunity. This version of Roll Call exists parallel</a> to the primary implementation, and is interesting as a proof-of-concept of how extreme portability might impact the device (it fits comfortably on top of my MacBook Air trackpad). It ditches the cables in favor of Bluetooth—I wrote a bespoke Python package</a> to render receipts and submit them for printing using the device’s largely undocumented communication protocol. Ultimately though, as entertaining as the new form factor may be visually and as rewarding as it was to overcome the progamming challenges in making it functional, I think this evolution of Roll Call is likely a dead end. There remain gaps in the current implementation worth exploring in hypothetical future versions, however. Roll Call’s deep reliance on my laptop, for iMessage and Contacts as well as to drive the printer, is a flaw in my screen-free vision. Ideally, all this would be replaced by a simple button-computer that could talk to a cloud macOS server as well as the receipt printer. As it happens, I have six weeks left at Recurse and a whole bag full of microcontrollers… More from Tarpenning 2024-12-16T19:40:00-08:00 Rocket by Marc Tarpenning. Finished on December 15, 2024. Rating: ★★★☆☆ I stumbled onto Altamira</a> during my usual daily web surfing, though exactly how I can’t quite remember nor discern from my browser history. What I do remember is that I was there for In The Dark</a>. Rocket</a> was a delightful surprise. It’s not hard to discern from my previous writing</a> that I think Marc Tarpenning is a smart fellow. I was quite excited to discover another opportunity to hear what he has to say. I’m also a sucker for bespoke hardware, e-ink screens, and reading books by people you’ve met. After reading, I can affirm that Rocket is an altogether pleasant experience. The fundamental insights about “measure twice, cut once” that I have already written about apply just as well to NuvoMedia as they did to early Tesla: Tarpenning writes that they “never varied the plan” for what they were going to build. His focus on business models—later important to proving that a market for high-end electric vehicles might exist—is also present, reflected at NuvoMedia in the discussion they had around the role of publishers and internet distribution in the world of e-books. Rocket places a stronger emphasis on a specific method for narrowing down addressable problem-space than the talk of his that I attended. The methodology is described succinctly as follows: Our thesis, if you will, was looking for products that didn’t exist, that the technology was just getting good enough to enable. </blockquote> The examples Tarpenning provides illuminate fun alternate histories that never quite made it. The form that this book takes is also quite notable: it is barely longer than 50 pages in total, heavy in font but light and thin in the hand. It would not be out of place as an article or a blog post somewhere on the net. I am a strong proponent of the pamphlet as an art form, so I don’t begrudge the physical manifestation of Rocket. I did not, however, enjoy the layout: were the Federalist Papers written in interview format? What I've Learned from Stella Levi (So Far) 2024-12-16T18:28:00-08:00 One Hundred Saturdays by Michael Frank. Finished on December 15, 2024. Rating: ★★★★★ This is the most vibrant, imagery-inducing work that I may have ever come upon. The first half of the book is dedicated to breathing life into the long-since vanished world of Rhodes’ Juderia. I have always been prone to feelings of nostalgia for times that I never experienced.1</a> This book exploits that tendency to the greatest degree, accentuated by stylized and interesting art interleaved within the text. Since I first cracked open its cover, I have been thinking ceaselessly about the structural features of the Juderia that induced specific behaviors and tendencies, which of those I admire and which of those I think are counter-productive, and how one would go about recreating an idealized version of this sort of enclave. Certainly an unexpectedly great book for people interested in the ways that urban environments impact their residents. Or people who like cities that sound like Acre or Jaffa. Of course, it is also serious work about an issue of much import: that of the total, intentional destruction of life in the Juderia and all that that implicates. I used the word “vanished” above to describe what happened to the world we witness in the first half of the book, but this is wrong: it was annihilated. This makes the middle of One Hundred Saturdays a rather heavier read than its beginning. It is also what makes the book worthwhile—life is not a walk in the park, and there is beauty in its challenges. This work would be incomplete as an analysis of life in the Juderia without a discussion of its end. The two are inseparable. One Hundred Saturdays is not even actually about historical events, so much as it is about Stella Levi herself. It is through her life—which so remarkably intersects time and again with the greatest focal points of change—that we learn about the shifting world. Stella is a woman of unfathomable resilience. This book is a treatise to her, a shocking and kind gesture from a close friend. The greatest gift one could divine from these pages would be a fraction of her insight. The book was also valuable to me in that it forced me to confront my own biases. I am in many ways very anti-diaspora.2</a> One Hundred Saturdays managed to challenge my anti-diasporic stance. Were I to apply to it the same standards as I do elsewhere, I should have been far more dismissive of life in the Juderia, Judeo-Spanish, and the culture expressed in the book. Instead, I found myself enamored. This is probably attributable to a combination of my unfamiliarity with Sephardic history between the Inquisition and 1948, personal connections and feelings of shame toward the Ashkenazi diasporic experience, and the charisma that emanated from every page of this book. I think the adjustment that I should take away is not a total reversal of my stance—the fate of the Rhodeslis must ensure that—but a softening. I would finally like to extend a huge thank you to my wonderful mom, who recommended me this book. EDIT: I sent this book to my TA from HISTORY81B last year</a>, because I vaguely remembered that her area of research was something to do with Ottoman Jewry and thought she might be interested. Turns out she did all the Ladino/Judeo-Spanish transcription. Nutty. Or, for that matter, never existed, as is the case with my feelings toward a whole host of Disney movies. ↩</a> </li> This of course clashes with the fact that I live in America, but: I recognize that what is correct for the individual on a case-by-case basis may differ from the needs of the whole/many, am not above the hypocrisy of adjudicating myself as a special case, and also think that the American acceptance of Jews does make it something of a special case. It is worse than Israel, but it is better than everywhere else where we are totally unwanted. ↩</a> </li> </ol> </section> Baymax on Memory Lane 2024-12-14T17:31:45-08:00 The above image was created as my final project for CS148</a>, under the auspices of renown bodybuilder and occasional professor Ron Fedkiw</a>. Everything in the scene was made from scratch in Blender, while the final image was rendered with the Cycles engine. All my work in the class was done in collaboration with my good friend Naama. The scene was inspired by the movie Inside Out, in particular when the character Joy ventures into the annals of her host’s memory. We appropriated the concept of a colorful shelf of memory orbs, and then brought another charismatic Pixar character into the scene, Baymax. We would like to think that he’s having a happy moment perusing good times past. The stars of the show when it comes to this 3D model are the memory orbs. Most impressively, there’s actually only one canonical orb in the scene! The rest are procedurally generated duplicates (same with the shelves). There is some simple math to figure out how to place them, and then a color is randomly selected from a weighted list of options: The most notable benefit we get from ray tracing in our scene is once again in the memory orbs lining the shelves. These orbs glow many different colors: yellow, red, green, blue, and purple. However, the light source for each orb is actually the same—a bright white light in the center of the ball. The light source is surrounded by the material of the orbs, which is transparent but tinted, and thus changes the color of the light ray after emission and before it hits the camera. Other benefits include the slight light penetration through Baymax’s inflatable body as well as area and sky lighting. On a more personal tack: making this scene vindicated a lot of the philosophies that I’ve picked up in Stanford’s Design program. During my first forays into Blender</a>, I was paralyzed to the point of inaction by the fear that any step I took—extruding a face, tweaking a texture—would mess up what I’d built so far or would build going forward. I was limited to following tutorials, which restored some of the structure I was familiar with from programming to the complex world of modeling things that looked good (technical term). But the d.school has changed my perspective</a>. The will force you to spend hours and hours sitting in the in-between phases, acting first and iterating, abandoning ideal perfections to make something real. With practice, you get comfortable in this space. That willingness to screw up and CMD-Z</code> was crucial on this project. Waking the Public to Waking Lions 2024-12-14T13:53:42-08:00 Waking Lions by Ayelet Gundar-Goshen. Finished on November 28, 2024. Rating: ★★★★☆ Author’s Note: This review is an adaptation of the final paper I wrote for Russell Berman’s wonderful Zionism and the Novel</a>. For another artifact of my work in the class, check out my review of The Illusion of Return</a>. This paper seeks to investigate Ayelet Gundar-Goshen’s novel Waking Lions as a work of engaged literature. A thorough analysis of her writing is conducted to extract the specific political positions she advocates. The analysis of the political commentary contained in her fiction is then paired and contrasted with contemporary anthropological scholarship on the issue of Eritrean migrants in Israel to build a deeper understanding of the context in which the book was created and where it now stands, approaching a decade after publication. The paper ultimately claims that Gundar-Goshen’s writing opens a wide view into the everyday strife and overlapping conflicts and harms of Israel’s myriad communities and urges a remediation of the Eritrean migrant crisis in Israel through integration and acceptance. The novel’s framing places the issues at the core of the text beyond the scope of the question of Zionism: rather, the continued implementation of Zionism is what has given rise to the status quo, thus necessitating Gundar-Goshen look beyond the philosophy—toward human values and compassion—to find the source of the mandate to aid the Eritreans. This dissection of Waking Lions is important precisely because the book has been treated as an unimportant work by the scholarly community. The text doesn’t touch the question of the Palestinians, and thus is sidelined. But just as is seen with the communities in the book, nothing is black and white: Israel is more than its conflict with the Palestinians. Waking Lions’ unflinching spotlighting of the myriad internal issues faced by the diverse Israeli polity make it one of the most important texts to come out of the country in recent years. Waking Lions is not only engaged literature insofar as all texts are engaged literature, dealing with subjects that are covered by the broad tent of political opinion, but instead goes above and beyond in provoking a political conversation around, and interrogating the political ramifications and engagements of, the Eritrean refugee issue in Israel. This is by design. Gundar-Goshen’s intention in creating this work of fiction was to make her political lamentations appear more compelling to a broader audience by disguising them in the contours of characters with depth and dimension and emotion that readers can recognize from their own lives. A political diatribe with specific policy criticisms and recommendations will only be consumed by a very specific subset of people. A work that engages more directly with the universal instinct of storytelling can better evangelize its message. This is how Waking Lions was crafted from its inception. An important component of Waking Lions’ political engagement is the length to which Gundar-Goshen goes to impress upon the reader the physical realities of Eritrean migrants in Israel, seeking to highlight their suffering through its personification in the character of Sirkit. Sirkit, the Eritrean woman at the center of the novel, is a complex figure—at once despised for her manipulations of Dr. Eitan Green and yet seemingly using her Machiavellian abuses for the betterment of the lives of overlooked refugees. It is only after forming a relationship with Sirkit that the reader is introduced to the details of her living conditions: a caravan, one cramped room filled with 8 mattresses and dirty dishes, parked behind the gas station where its residents work.1</a> Sirkit herself scrubs floors during the day. None of the Eritreans who cram themselves together into the caravan to sleep on the floor night after night are paid the minimum wage nor given the traditional benefits of employment. Additional emotionally difficult information is held even further, with the reasons for Sirkit’s flight from Eritrea hinted at only in the very last chapter of the novel. Here, in her internal monologue, she refers to the African nation as “the land of the dead children.”2</a> She reflects on a “well near the village that one day, simply had no more water,” on soldiers that stole their flour, on the trek over land to Egypt, on abusive Bedouin smugglers, and on Israel, the place where “she stopped. From [which] she would not move.”3</a> The strategic delay in the delivery of this information is done to counter the jaded desensitization of the receiving audience. There is the age-old adage: “one death is a tragedy, a million is a statistic.” By giving the physical conditions of the Eritreans a familiar face in the form of Sirkit, the theoretical plight of a people is turned into the tangible plight of a person. Beyond the individual suffering, the systemization of the Eritrean struggle in the Jewish State is given an embodied form in the text through Eitan’s visit to the Holot Detention Center. Having come to visit Sirkit, Eitan looks out across the vast desert yard and observes: “Any one of those people could be Sirkit.”4</a> This is an explicit declaration of the above determination that Sirkit is a stand-in for the plight of Eritreans in Israel more broadly—Gundar-Goshen states that Sirkit is equivalent to the other Eritreans in Holot, and thus that all the detainees are as human as the book has portrayed her to be. It is also a further description of the dehumanization that Eritreans are subjected to at the hands of the Israeli government. Eitan continues: “They looked as alike as a herd of sheep. Of cows…. When he looked at them together, a crowded collection of bodies, he felt that they had lost every drop of selfhood, and all the small differences that made each of them who they were had been eclipsed by that large mass of identical flesh… the overcrowded space stripped them of their personalities and made them a single entity—Eritrean women…. They were Eritrean women waiting to be deported…”5</a> In his brief visit to the Holot Detention Center, Eitan is used as a tool to convey the banal brutality of the destruction of Eritrean individuality in the national system. The ultimate resolution of the book’s moral challenge through Sirkit and Eitan’s deception reveals Gundar-Goshen’s preference for how to resolve the political quandary of Eritrean migrants in Israel. The roiling truth of the story behind the illegal hospital in the desert—Eitan’s initial crime of vehicular manslaughter, Sirkit’s blackmail, the robbery of hospital materials, peripheral involvement with the Bedouin criminal underworld, hints of adultery, and everything else that ultimately comprises the narrative of Waking Lions—is smoothed over. Sirkit does this of her own initiative. Speaking with Eitan’s wife, Liat, she papers over the reality of the situation that had brought his family and marriage so close to collapse: “Before the Bedouins had surprised them, Eitan had gone to treat her injuries. He’d left Yaheli’s bed and driven two and a quarter hours to get there. Only an angel would do something like that.”6</a> With the full context Gundar-Goshen provided during the action, the additional layers are revealed. Eitan drove down to perform the surgery in part because of his romantic feelings toward Sirkit as well as her continued power over him due to the potential for her testimony. But in the end this is resolved—his service to the Eritreans is simplified: “He felt guilty about the silence she had imposed on him concerning Zakai’s bribes. He wanted to atone…. It was illegal. And dangerous…. And [Liat] realized suddenly why he had been so interested in the investigation of that Eritrean’s death. Those people weren’t just a newspaper article for him. He knew them. He was helping them.”7</a> Gundar-Goshen does briefly broaden the scope to include some of the other narratives in Israeli discourse, describing a café scene after Eitan and Sirkit’s revisionist story hits the news: “Several people began arguing. We can’t have all of Africa coming here. If those bleeding hearts have their way, we’ll end up without a country.”8</a> But this broadening is done primarily for the purpose of foregrounding the opposite narrative—her narrative— as expressed by a woman who approaches Eitan to say, “We need more people like you in this country.”9</a> Through Liat’s acceptance of the morality of her husband’s actions and the Israeli public’s endorsement, Gundar-Goshen lends her own voice to the idea that Eritreans in Israel should be accepted, appeased, and integrated. Her political preferences, expressed through these varied characters, are paired with an additional pithy quip, a sort of condemnation of the idea that the hard work her politics would mandate can be ignored. It is the final line in the novel, and it comes from the mind of Eitan Green, newly freed to return to his old life and—apparently—bury his head in the sand: “How beautiful the earth is when it moves properly. How pleasant to move with it. To forget that any other movement ever existed. That a different movement is even possible.”10</a> Gundar-Goshen’s political commentary through Waking Lions of course exists in the context of Israeli reality. This has continued to evolve quite rapidly in the seven years since the publication of her work of engaged literature. As recently as last year, there was violence in the streets of Tel Aviv between different Eritrean factions resulting in the injury of over 100 individuals as well as significant arrests. The conditions in Eritrea that led so many to choose “liminality in Israel over forced conscription (often until death) in Eritrea or ethnic cleansing by Arab groups in Darfur” are the same or worse as they were at the time Gundar-Goshen released her novel to the world.11</a> Migrants largely remain in limbo, governed under the conflicting mandates of the Prevention of Infiltration Law and the 1951 Refugee Convention. Aspects of Gundar-Goshen’s humanitarian ideology have received broader adoption. The Holot Detention Center, a location that played a significant role in the psyche of the Eritreans of Waking Lions, was shuttered in 2018.12</a> The Deposit Law, which mandated 20% of asylum seekers’ salary be deposited in a bank account only accessible at the airport when leaving the country, was struck down by the Israeli Supreme Court in 2020.13</a> Eritreans have developed their own community centers, educational structures, and institutions that “attest to the agency of the… community in Israel.”14</a> Migrants are building lives in the country, living on visas that need renewal every 2-3 months.15</a> This itself is a massive political victory for Gundar-Goshen’s school of thought: as Waking Lions conveys, the daily lives “of Eritreans in Israel are not apolitical.”16</a> Their continued existence in Israel is a testament to the political success of the ideology expressed in Gundar-Goshen’s work of engaged literature. In spite of these developments, Israeli society remains broadly hostile to the presence of the Eritrean migrants in precisely the ways that Gundar-Goshen opposed in Waking Lions. The country is governed by an extreme right-wing coalition which harbors significant anti-Eritrean sentiment, though internal and external factors have made continued attempts to address the migrant crisis low on its list of priorities. The anti-migrant position resulted in the Prime Minister reneging on a negotiated settlement with the UN refugee agency to give permanent status to around half of asylum seekers in Israel in exchange for resettling the other half in other countries.17</a> Such a compromise was perceived as too soft and harshly criticized. Integration of Eritrean migrants as full members of the Israeli polity, as citizens, is perceived as undesirable. Emigration is encouraged, and the mass departure of Eritrean migrants remains the state’s preferred outcome. The specific interplay of the international system in the Eritrean refugee crisis in Israel is a particularly rich topic for dissection. In many ways, the sovereign Israeli system is set up in direct opposition to the presence of Eritrean migrants, but is countered by international refugee law that exists “precisely because states are often inclined to act differently than how the law prescribes.”18</a> The particularly proactive role that international governance plays in the day-to-day experience of Eritrean migrants in Israel is made greater by the already-enhanced focus of the international community on the Jewish nation—a fact which the community has at times used to its advantage. Eritrean society is not traditionally organized around the concept of race or color, but rather ethnolinguistic groups and tribal affiliation.19</a> In most migratory scenarios, individuals who may be racialized as Black “attempt to highlight their immigrant background or national origin to escape negative stereotypes” associated with this new identity.20</a> The Eritrean community in Israel has become an exception to this trend, opting to make the strategic decision to self-identify as “Black.” This particular language is intended to garner support from abroad by contextualizing the experience of Eritrean migrants in Israel in a foreign framework, so as to make it more intelligible in the international arena and increase pressure on the Israeli state to halt deportations. The role of nonnative forces in the Eritrean-Israeli crisis goes largely undiscussed in Waking Lions, to its detriment. Ayelet Gundar-Goshen’s Waking Lions paints a colorful picture of a pressing, contemporary Israeli issue, presenting the reader with a clear call to action through its expert personification of the problem. The narrative is thoroughly grounded in modern Israel, making the intentional decision to place the question of Zionism squarely in the past. The Jewish State already exists—Gundar-Goshen’s narrative interprets the Eritrean migrant crisis within its borders as a question to be answered by the generic State portion of the Zionist dream, not the Jewish (and by extension Zionist) part. The novel implores its audience to take action to embrace and integrate Eritreans into the fabric of Israeli society. Its detailed description of the suffering of the migrant community, in desperate poverty and constant fear of state action, is gracefully described with its day-to-day complexities—its members are not pure good, nor evil—while ensuring that it is clear that such suffering is a moral failing of the state. Gundar-Goshen believes this can be solved. It is in turn a failing of the scholarly community that this is the first paper to seriously engage with Waking Lions. The discussion of international conflicts, and their portrayal in literature, is flashier. Such analysis allows the author of a paper to connect with the oldest of human traditions: myths of wars and conquests waged throughout the eons. It is, bluntly, dramatic and fun. In the Israeli context in particular, there is no shortage of conflicts and fictions about them to dissect; the Palestinian issue in particular provides a sure and stable base for study. It is, however, a poor academic that allows themselves to fall prey to sampling bias. Israel is far more than its conflict with the Palestinians. Indeed, for the Eritreans under constant threat of deportation—and, for that matter, Dr. Eitan Green—the matters highlighted in Waking Lions take precedence. The dialogue in this work of engaged literature must be given space, instead of letting the Palestinian issue take all the oxygen. With an opportunity to thrive, to engage with and impact a large audience, Gundar-Goshen’s novel could catalyze real progress on one of Israel’s most serious internal conflicts. Ayelet Gundar-Goshen, Waking Lions (London: Pushkin Press, 2017), pt. 2 chap. 3. ↩</a> </li> Ibid., pt. 2 chap. 16. ↩</a> </li> Ibid. ↩</a> </li> Ibid. ↩</a> </li> Ibid. ↩</a> </li> Ibid., pt. 2 chap. 15. ↩</a> </li> Ibid. ↩</a> </li> Ibid., pt. 2 chap. 16. ↩</a> </li> Ibid. ↩</a> </li> Ibid. ↩</a> </li> David Clinton Wills, “A Home at the End of the World: Eritrean and Sudanese Asylum Seekers in Tel Aviv, Israel,” Sanglap: Journal of Literary and Cultural Inquiry 3, no. 2 (2017): 321-349, https://sanglap-journal.in/index.php/sanglap/article/view/223</a>. ↩</a> </li> Ibid. ↩</a> </li> Itamar Dubinsky, “Digital Diaspora: Eritrean Asylum Seekers’ Cyberactivism in Israel,” African Diaspora 12, no. 1-2 (2020): 89-116: 10.1163/18725465-bja10002 ↩</a> </li> Ibid. ↩</a> </li> Clinton Wills, “A Home at the End of the World” ↩</a> </li> Dubinsky, “Digital Diaspora” ↩</a> </li> James Yap, Hilina Fessahaie, and Enbal Singer, “Populism’s Global Impact on Immigrants and Refugees: The Perspective of Eritrean Refugees in Europe and Israel,” Maryland Journal of International Law 35 (2020): 189-201 ↩</a> </li> Dubinsky, “Digital Diaspora” ↩</a> </li> Amanuel Isak Tewolde, “Becoming Black: Racial Formation of Eritrean Migrants in Israel,” African Diaspora 13, no. 1-2 (2021): 183-203, 10.1163/18725465-bja10006 ↩</a> </li> Ibid. ↩</a> </li> </ol> </section> Marc Tarpenning on Innovation 2024-12-01T00:00:00+00:00 I was very fortunate to have the opportunity to hear Marc Tarpenning, cofounder of Tesla, speak at the beginning of this quarter. It was the most exciting and impactful talk I’ve attended at Stanford thus far. This is really saying something. In his talk, Tarpenning elucidated a theory of innovation and disruption that feels to me to have been “lost to time.” It is a measured, idea-first strategy that is totally foreign to the current generation of startups and to a world in which “founder” is itself a career. Sara Singer</a> condensed his theory into a wonderful diagram that I have further developed into the following list (broad principles are bolded, with the specific examples from Tarpenning’s time at Tesla in the normal weight): Identify a real problem. Prove it exists with real-world data. Narrow down your search area by taking the age-old wisdom seriously: “if something can’t go on forever, it won’t.” Create the solution that will replace it. At Tesla, Marc and his cofounder Martin Eberhard wanted to help solve the issue of climate change. The DoE projected in 2008 that over 70% of U.S. oil demand came from the transportation sector, and others have calculated that decarbonizing road transportation “could feasibly reduce global emissions by 11.9%” (The Carbon Almanac</a>, pp. 67). Without any action, climate change would end the world. The scenario they identified thus fits the above model—a problem that exists and a solution that must arrive eventually. So they set out to build a car without emissions.</li> Evaluate your solutions. Prove that it will work, mathematically, before you start building. Your product cannot be better along just a single-axis if you want to change people’s behavior; make a Pareto improvement</a>. Do the hardest thing first to learn if your solution is possible. Marc and Martin evaluated two things: the technology and the market. First, they exhaustively investigated possible power sources before deciding that batteries had reached the right power density and mass-producibility to be—provably—the most effective alternative to gasoline. Even powered by coal generators, an electric car is less polluting per mile than an ICE vehicle due to the efficiency of the electric motor (significantly less energy wasted as heat, thus more energy toward movement). On the market front, they found that upper-class Americans were seeking out environmentally-friendly vehicles as a status symbol—with the Prius eating into Lexus sales, and GM having to physically claw back their EV1</a> from leasers. Without scale, their first car would be expensive, but it being the only vehicle to truly satisfy this niche meant if the tech worked they would find buyers. Finally, their solution wasn’t just technically better for customers with its radical efficiency: it was fun, sexy. The rise of electric cars has made them feel normal, but Tarpenning described the “electric smile,” a meme they threw around at Tesla in the early days to describe the reaction that drivers had from the very first moment they pressed the pedal and felt the instant torque.</li> Change the world. The shape of progress is not inevitable, we make it. Success is assured only in retrospect.</a> Tarpenning advocated for a goal of “sustainable abundance,” forged from a fusion of technological advancement, government buy-in, and behavioral change.</li> </ol> A lot of software businesses have none of these qualities. Daring Fireball on Love 2024-11-01T17:28:15-07:00 A funky list of two for my own use mostly. Facebook’s VR Platform ‘Horizon’ Sucks</a></li> The Things They Carried</a></li> </ol> The Novel of Palestinian Disillusionment 2024-11-01T17:25:51-07:00 The Illusion of Return by Samir El-Youssef. Finished on October 23, 2024. Rating: ★★★☆☆ Author’s Note: I write everything in a conversational tone, so I doubt anybody reading this will notice much difference, but this review was originally a presentation. It is one of two I gave in Russell Berman’s wonderful Zionism and the Novel</a>. It was first written in iA Presenter</a>. The title I have given the review here on the website was taken from our syllabus. For another artifact of my work in the class, check out my review of Waking Lions</a>. Alright close your laptops and let’s do this for real.</del> To get you up to speed, let me tell you succinctly everything you need to know about the book. Let us start with the fact that it’s semi-autobiographical. Though it is a novel, and elements are fictionalized, the life of our protagonist broadly echoes the trajectory of our author. Which is a great intro to our Beginning with our Author-Protagonist. He’s Palestinian, born in Lebanon, and lives in London. He has no name! We do know a lot about him though. Ali has the same origin story, lives in the US, and is drug buddies with the protagonist. Ali is the closest of the bunch to the protagonist, because the other two basically thought that the protagonist wasn’t mentally up to snuff. Maher has the same origin story as well. He is also a Communist. George is a Lebanese Christian, still in Lebanon as far as we know, and a big fan of abstract philosophy (specifically Heidegger). He is the only non-Palestinian of the bunch. In the prologue, we witness the protagonist approaching the fifteenth anniversary of his departure from Lebanon. He gets a call from Ali, who has a layover in London on his way from the US back to Lebanon, and he wants to say hello. This throws the protagonist into a whirlpool of emotions, as he had pretty much severed contact with his past up until this point. The real meat of the novel is narrated to us in a series of flashbacks to the last night the gang of four spent together, revealing progressively more and more about their overlapping lives and relationships. Finally, in the prologue, we get to witness the discussion between the protagonist and Ali in the Heathrow Airport and a little bit of the latter’s perspective on things. And, before we address the fun socio-political commentary of the novel, we can take a small more-mundane dive into one of the big questions of the course: El-Youssef answers clearly that, whether we like it or not: yes. The stories in this book are deeply tragic. Our protagonist had a sister named Amina. Amina was physically abused by her other brother Kamal, and ends up killing herself. The protagonist is haunted by her memory, and his family refuses to even say her name. Obviously, this is deeply tragic. It also exists in a political context! Amina shoots herself after Kamal sees her kissing a man and threatens to kill her himself for dishonoring the family. The freedom of women in society—regretfully, a political issue. The man Amina was kissing was a fellow member of a Palestinian resistance movement. Her family found her membership in that deeply troublesome, primarily because of their uniform of military fatigues. After her suicide, the movement covers up the cause of her death and honors her for dying in the fight against “the Zionist enemy.” All deeply political! Ali’s had a brother Sameh. Sameh was gay, and because of this a faction of Palestinian militants forced him into working for them as some form of punishment/conversion therapy (obviously political). Sameh is eventually caught smuggling arms into Israel and shot. The IDF traced his dead body and the van he was driving back to Ali in Lebanon, and force Ali to become an informant. Super political! And of course, very personally tragic! Ali and his brother are made pawns and stripped of their agency. When his brother is eventually killed, Ali experiences a mix of emotions—he is devastated, but also grateful because of the shame that Sameh’s sexuality brought upon him. Super horrible! You’ll also note that unlike the other three, Maher’s intro didn’t have a note about what he’s currently up to. That’s because, on the night at the cafe that we witness, Maher is kidnapped and murdered. He is kidnapped and murdered in part for political reasons—he’s a communist—but also for the personal implications that political fact has created—his assailant is the bereaved son of a man who’s factory was destroyed at the hands of a communist revolutionary Maher helped radicalize. George hides his personal struggles, for political reasons. His family lives together, but he reveals to the protagonist on their walk home from the cafe that his parents have been divorced since he was a child. He has struggled with love for his whole life, having grown up in an ice-cold household. But he can never reveal any of this, or stir up any great troubles or emotions, because of his precarious position in the community: a Christian amongst Muslims, a Lebanese amongst Palestinians, a third party caught in a Judeo-Islamic war. Not seeking to anger any dangerous people with guns, he lives a life disconnected from his personal struggles. In this book, and in real life, the personal world has profound political implications and vice versa. And now for the fun part. I made up this word, and its going to live on the screen while we talk about what El-Youssef wants us to learn and believe. Because this book is a commentary on the Palestinian right of return. There’s significant and explicit commentary on it within the book—the protagonist wrote (sort of) a thesis on the subject—including this lovely nod to the audience: “‘It’s a one-way journey!’ he told me,” said Ali, “‘As for those who claim to return to a place where they never were,’ said Bruno, ‘they are simply confusing the symbolic and metaphorical with the possible and actual.’…. [T]here is no such thing as the right of return… I shall write [this] as an essay or a story, which I could call The Illusion of Return. </blockquote> Now, last year, in a wonderful course that I do not really recommend anybody take because it’s quite boring and not that wonderful called HISTORY81B: Making the Modern Middle East</a>, I read a book called The War of Return</a>. Unlike the course, this I highly recommend. It was written by two prominent Israeli leftists and contains an in-depth history that is largely unknown, and records some important oddities related to the right of return and the issue of Palestinian refugees. History in 1948 must be judged within its context, and not against modernity. Much was different back then, and population exchange and territorial partition were viewed as legitimate ways of solving ethnic conflicts in ways that they are perhaps not any longer. We can see this happening all over the world and at a much larger scale than in the Levant, such as in Greece and Turkey or India and Pakistan. So the illegitimacy—or even unique nature—of the circumstances surrounding Israel’s founding and the birth of the Palestinian refugee issue is suspect from that perspective. And beyond that, Palestinian refugees have been treated radically differently from any other class of people. They are in fact not governed by the UN committee that deals with refugees, but by their own UNRWA—which you have likely heard about in the news recently. (As an aside, it’s not incorrect to say “their own” in that sentence, as the vast majority of UNRWA employees since its inception have been Palestinians). They are hereditary where no other refugee group is, creating 5 million refugees from an initial displacement of around 700K. Their status is not surrendered when they gain citizenship to their new host countries, leading to Jordan’s population being majority (~70%) Palestinian refugees. The entire population of the West Bank and Gaza have been considered refugees since 1948, despite living in territories allocated for their own future state. The “refugee camps” that most of them live in—and into which three of our four main characters were born—look far from the camps of the standard Western imagination, but are in fact fully built-out neighborhoods attached to or themselves forming major cities (see: Beirut and Jenin, respectively). Furthermore, attempts to resolve the refugee issue by traditional means—resettlement, economic empowerment, and rehabilitation, as seen elsewhere (including some of the UN’s most successful projects, such as the reconstruction of Korea)—have been intentionally shuttered. Please see these two quotes: The refugee issue, claim Schwartz and Wilf, is cynically manufactured and perpetuated by Arab leaders. </blockquote> And let’s face it… the Arab countries are not the most hospitable places, especially for Palestinians. </blockquote> This first quote is actually from a book report</a> I wrote on The War of Return, and the latter from El-Youssef in The Illusion of Return. The issue of Palestine can be resolved. States can be formed, and given borders. But the right of return—this is crazy important, and something I realized in reading The War of Return—does not refer to the ability of Palestinians to immigrate to the future territory of a Palestinian state (presumably in the contemporary West Bank and Gaza) but rather to the whole territory that once comprised the British Mandate, including the modern State of Israel! This is the issue of Palestinians, not Palestine. And the broader Arab world does not want to resolve it. Because resolving it means closure. It means continuing forward in time. It means that Jews will have sovereignty over a slice of the Middle East. Our protagonist begins many things throughout the book, but never finishes them. He also struggles with time and presence, doubting whether the past really happened at all. His parents left Palestine for Lebanon, and he left Lebanon for London—but try as he might he cannot truly move on. The world around him forces this anti-closure upon him, a relic of his past—Palestinian-Lebanese Ali—literally showing up on his doorstep as he arrives at his anniversary of emigration. But he is wise enough to know that you cannot undo the past. Which! Brings us to: Though not the one you’re familiar with. El-Youssef doesn’t believe in return. The Jewish question I pose to you all now is thus: how can we reconcile this position with the reality of Israel—a return that worked? What, if anything, makes the case of the Jews different? AirPods and Thoughtspace 2024-11-01T16:34:57-07:00 I was walking around the other day, listening to the sounds of the night, when I had a thought about AirPods. This is not the first time this has happened: previously, I have wondered about the percentage of sound I hear that comes filtered through Apple’s transparency software, a number that is likely shockingly high. On this particular occasion, I thought it merited a blog post, mostly because I invented a word and that made me feel like Matt Webb</a>. The thought was: AirPods are bad because the music invades thoughtspace. Not in the way that all noise crowds out thoughtspace—external noise impacts thoughtspace by allocating a relatively larger portion of moment-in-time human compute toward world-processing as opposed to thoughtspace. AirPods, on the other hand, project music and sound into thoughtspace directly. This is because the sound doesn’t interact with the outside world. Instead, it comes through with the same clarity as your thoughts. Perhaps this is a relatively easy fix then? Contra Shipping in Public 2024-10-23T00:00:00+00:00 Mitchell Hashimoto, announcing the release of his new terminal emulator: …I am nostalgic for the days when software was released when it was ready. I like to imagine that if stores like CompUSA still existed, Ghostty 1.0 would be boxed, shrink-wrapped, and on the shelf ready for purchase (but also free and open-source). </blockquote> It’s a delicate balance. I could actually summon a hundred anecdotes on the subject: “ship early and often,” the whole build-in-public movement, the concept of an MVP, etc. All of this is well and good, but it has taken some of the magic out of release day. Shipping before we’re done means that we don’t have a moment to point to where everything changed. No more “iPod, phone, internet communicator.” ★</a> An Insight Tangential to Thorsten Ball's Intention 2024-09-27T00:00:00+00:00 Reading yesterday’s Register Spill post I got hooked on this sentence: Is [the success of frequent and significant rewrites by the founders of Zed] because they have been working on text editors for over a decade and know the domain well and how things should work and how they currently work is just a detail? </blockquote> I love the idea of seeing something for how it should be with its current workings as “just a detail.” ★</a> I Am Amazed That This Exists 2024-09-20T00:00:00+00:00 UNSONG by Scott Alexander. Finished on September 20, 2024. Rating: ★★★★★ The overt meaning of UNSONG is “a book written by pseudonymous rationalist blogger Scott Alexander published as a website in 2017.” The kabbalistic meaning of UNSONG is “the nerdiest, most niche work one can consume, whose physical size as a massive tome may make carrying it around in a backpack across several countries more difficult and contribute to zipper failure.” This we derive from my personal experience. Both of these are discrete from the acronymic meaning of UNSONG, which is of course “United Nations Subcommittee on Names of God,” and thus illustrative of exactly why I enjoyed this book so thoroughly. I had initially dismissed UNSONG as outdated. Surely this was an old and abandoned creation of Scott’s, less important than his contemporary contributions. I was also not hugely into the concept of website-books. It was thus fortuitous that these two concerns were addressed with an updated physical edition</a>. “This was not a coincidence, because nothing is ever a coincidence.” This book exists at the most wonderful intersection of things that I adore. I have already gifted a copy to a good friend, and intend to lend mine out enthusiastically to anybody who will take it. A selection of wonderful quotes: Bizarre surrealist painter Salvador Dalí once said: “I do not do drugs. I am drugs.” He was being silly. He wasn’t drugs. He was Salvador Dalí. </blockquote> “Kissinger, are you sure this was a good idea?” “There is never any surety in politics, only probability.” “But you think it’s probable this was a good idea?” “Yes.” “What about our [expletive deleted] souls, Kissinger?” “I fail to see how they are relevant to geopolitics.” </blockquote> My mother was a Hindu, but Hinduism passes through the paternal line. My father, perhaps if he teaches me Torah then is Jewish. But Judaism passes through the maternal line. I am nothing. </blockquote> Somebody has to and no one else will. </blockquote> I’ve also added King James Programming</a>, referenced occasionally in the epigraphs of UNSONG, to my RSS reader and expect to extract a certain amount of joy from that every day. Some of the Hebrew in UNSONG is not fully intelligible. I have decided to accept this as suboptimal transliteration done in good faith, so that I can move on and love the rest of the book. With all that said, they released a Kindle version</a> shortly after I bought the paperback and would absolutely recommend you get that one instead as it really is quite something to carry around all the time. But do get it. Some of This and Some of That 2024-09-19T00:00:00+00:00 The Great Gatsby by F. Scott Fitzgerald. Finished on July 04, 2024. Rating: ★★★★★ I read this and imagined it contained in a small pocket-sized book, and myself quite the intellectual, amused by quaint fictions. I sought to divine from the tale of hedonism and grandeur exactly the line where one crosses over from glory into excess. A selfish goal, to be sure. I watched the movie</a> shortly after reading the book. I enjoyed the adaptation greatly. I thought it did well in capturing the scale of the parties implied in the novel. I was not a huge fan of the narration. I enjoyed that they threw out any historical pretense and embraced a radically modern soundtrack. It is a young man’s interpretation of this book that defines it as a source of inspiration, but I can’t help myself. I want to throw out my arms and present to an audience the splendor and spectacle of my life and tell them they can’t find anything quite like it anywhere else in the world. Now ideally, I would be able to do as such without Gatsby’s failings. I don’t think a warning about spoilers is required some hundred years after publication: the book doesn’t end well for Gatsby. Is it wrong to say this is due to his refusal to take the past as immutable? His dissatisfaction with only controlling the future? These seem remarkably solvable issues. I am surely missing many others. For now, at least, I will try and fold what showmanship and generosity I can from Gatsby into my own life. To give favors and gifts to those who wash up on my shores. To unabashedly enjoy existence</a>. To fake it, as they say, until you make it. I dove into the world of Gatsby at the recommendation of Jack and his family. My thanks go out to you. Envelope 2024-09-03T00:00:00+00:00 Began: 17 Jun 2024 Concluded: 06 Sep 2024 YCombinator’s job board</a> has high signal to noise for interesting work, but it’s not every day that their internship tab has as compelling a phrase as “SwiftUI Engineer.” I’ve been in love with SwiftUI, Apple’s modern app-building framework, for a long time</a>. Something about its declarative nature</a> just clicks with how I think. I sent in an application, sat down for a live coding interview, and my summer plans were sorted. Envelope</a> is a banking app. It serves as both your primary checking account and your budgeting platform: the two are seamlessly integrated. Every transaction that hits your bank is budgeted by default. Envelope lowers the barrier to entry for your most important personal finance habit: great budgeting. It also helps that Envelope brings all the polish of a fancy Silicon Valley company to banking, a field dominated by the jankiest apps known to man. One bullet point in the job description claimed that I would have code live in production starting from my first day. This was not true. But it was close! In my first week at the company, I built out a new feature shown—to every user, on the front page of the app—when assigning a transaction to an envelope. It is now one of the most used pieces of UI in the app. Over the rest of the summer, I continued building and iterating on practically every facet of the application. By my count we shipped nine long-requested major features and redesigns, touching everything from transactions and envelopes to onboarding and the organization of screens within the app. I took a look on my last day at our improvised Figma kanban board, and was happy and more than a little bit surprised to see the done column overflowing with projects we’d shipped. One particularly fun piece of work that I’d like to immortalize here was totally removed from feature development: we once spent two days debugging the Stytch</a> JWT</a> that our app uses to authenticate users. When we ultimately located the bug—a particularly nasty problem with refreshing the token that rendered it invalid after using the app for longer than five minutes at a time—we were able to fix it with one line. We also got a DM from a co-founder of Stytch out of the whole experience, and sent them some API design pointers that I sincerely hope make it back to their SDK</a> one day. I’m excited to see Envelope grow, and for the additional features that I didn’t have time to put in your hands to see the light of day. From closer integration with external money sources, certain additional banking tools, and some cool UI we couldn’t ship due to dependency issues—I’ll be keeping an eye on the future of the features I built out and the future of the whole enterprise. Josh, the CTO, is brilliant. I was the second engineer at the company—a fun statement for me in its own right, but also formidable by implication. Josh built out the entire backend, external integrations, and SwiftUI application, and had been running them both as a serious product for two years as the only technical person on the team. Josh is deeply opinionated about how software should be developed and maintained, employing a crazy effective philosophy breaking complex problems down into distinct patterns. He is deeply talented, and I spent a good part of my internship just trying to absorb some fraction of his skill. The realities of the job were also completely new to me. Working a 9-5 took some getting used to; at Stanford, I am surrounded by people doing crazy and fun things from the moment I wake up until I pass out and in class for maybe three hours of the day. In school, the day is for play and work is confined to some fraction of the evenings, but at Envelope this was flipped. I think I obviously appreciate the flexibility of how work is done at Stanford over the more traditionally rigid hours. It’s nice to have time to ponder, to be free to meet with people for lunch and tea throughout the day. My work at Envelope also helped me contextualize the sort of programming I want to do in the future. Pure frontend work, devoid of context, is just making a hundred buttons a day—and when you’ve made one button, you’ve really made them all. However, working with data, building out new functionality, reshaping and redesigning the user experience: that is deeply fulfilling. My time at Envelope allowed me to touch on all of these aspects of development and take a holistic approach to my work. My time at Envelope was a blast and a privilege. Every change I made went out to thousands of people using our app on a daily basis. What an incredible opportunity it was to be able to shape something like that so directly. Beeper 2024-09-02T00:00:00+00:00 Began: 05 Jun 2023 Concluded: 01 Sep 2023 As the summer of 2023 rolled around, I found myself opening up a bunch of tabs in Safari to check if any cool companies I had heard of on the orange site</a> were hiring. I remember being shocked and excited when I saw that Beeper</a> had a listing. Their app was built on top of an open, secure, and decentralized protocol called Matrix. I had used Matrix through the Element</a> app with some friends, and was aware of some of the cool projects that took advantage of the network. I was also aware that most of the code that deals with Matrix is open-source and written in Go—two qualities shared by Mabel</a>, the terminal app I had just finished working on. Beeper sat at the intersection of a lot of things I was and am still crazy about. I sat down with Eric Migicovsky</a>, the CEO, and came on as a contractor for the summer. Everything I encountered at Beeper was brand new to me. I had never even been paid to code before, and the team was fully remote. Showing up to the all-hands every Monday on Zoom felt pretty surreal. I started out reverse engineering and documenting Snapchat to bring Beeper to the platform, but was quickly moved to work on our hardware venture, affectionately named (after Blackberry threatened legal action) Beepy. The Beepy</a> was intended to showcase Beeper’s flexibility. What has changed fundamentally about messaging since AIM</a>? Not much. But Beeper, with its foundations on the open Matrix protocol, could take any form imaginable. So to set an example for developers, we brought in the big guns from SQFMI</a> to strap a keyboard and a black-and-white LCD screen together to make something like a modern-day, common-noun beeper</a>! The software I built for the device came in two parts: a setup utility</a> and a chat client</a>. The former got new Beepy users “from zero to messaging,” taking advantage of users’ main computer to gather their chat history and encryption details before copying all the data to the Beepy. The latter was based on the gomuks</a> terminal app and redesigned for the device’s smaller 1-bit display. Both of these programs were wonderful opportunities to leverage the skills I had just learned building Mabel. A huge congratulations to the whole team on your new</a> home</a>! I nursed a pet project while I was at the company that I believe has real potential: Beeper Rewind. How many messages did you send this year? What was your daily average? Who was your most frequent contact? What platforms did you use the most? The least? Who is most active in your groupchats? This metadata is super fun to know, as a user, and totally shareable. I would love to clown on the lurkers in my friend group, and be able to quantify how crucial iMessage is to my daily life. Beeper is the only entity in our digital lives that could provide these insights, and generating a little viral image for your Instagram story could be one way to exploit that position. Hesitating on this goes against some of my core principles, and I should have pursued it further. Speaking of positioning, I don’t think Beeper necessarily optimized its business to be in line with its competitive advantage. We spent a lot of time at the company trying to re-imagine the fundamentals of the messaging experience—it’s probably no surprise that a lot of this involved AI. I think this is all well and good, but ignores the fact that we already had the ability to do something nobody else could, which is to send messages on every platform from one place. I think we should have been focusing on optimizing the mobile app to provide the smoothest messaging experience possible while doing this. Polish, polish, polish. Even beyond this, however, we should have explored the B2B angle. Humane</a> launched their wearable with SMS messaging. How much would their user experience improve if the Pin could talk to your iMessage, WhatsApp, and Signal? How much is that functionality worth? I’ll give you a hint: a lot, and we were the only people who could provide it. I don’t think we ever considered this. Lastly, on a technical note, the concept that underpins Beeper’s ability to send messages across platforms—Matrix’s bridges</a>—is really cool. It is perhaps worth pondering where else it could be applied: bots that observe and replicate behavior across networks, to maximally leverage your efforts. Adversarial interoperability! Vincent Cloutier</a> is doing fun things in this space over on the fediverse. Summarized in a sentence that sounds straight out of Matt Webb</a>’s school of thought: What happens when we ignore the borders between spaces and connect them anyway? In the summer of 2024, I sat down to talk with a mentor of mine—a successful startup founder, executive, and angle investor of some renown—who presented me with the bluntly phrased perspective that people simply don’t want message aggregation. Rather, he believes, distinct messaging apps are actually a feature: they give you separation of concerns</a>. Having sat on this for a little while longer, it occurs to me that the super-app that aggregates your different messaging platforms may already exist, and it’s your operating system. Stratchery on EU Antitrust 2024-08-26T00:00:00+00:00 I would also note that the behavior I am calling for — more innovation and competition, not just from Google’s competitors, but Google itself — is the exact opposite of what the European Union is pushing for</a>, which is product stasis</a>. I think the E.U. is mistaken for the exact same reasons I think Judge Mehta is right. </blockquote> Though I have more complex thoughts as to the actual anti-trust case against Google itself, I do endorse this framing of the European thinking on Big Tech regulation. ★</a> Ideating Tragit 2024-07-30T00:00:00+00:00 There are many, many git</code> forges to pick from. I’m going to make another one—but before I do, I’m going to do a little thinking out loud, and outline the what and why of this side project I’m embarking upon. Decentralization is hugely important for the long-term health of the internet. Indeed, this quality has been fundamental to the network since its inception. This is easy to forget, and hard to see, now. But email was not always Gmail and git</code> was not always GitHub.1</a> A big reason that these protocols have become so centralized is because the companies that offer these services have genuinely innovated on form. Google invented effective, user-friendly webmail. They pioneered email’s contemporary security features and user-facing advancements. They were rewarded for that with users. Nowadays, however, these capabilities no longer require the dedicated engineering of whole teams of Google 10x-ers. With the mature and effective tools and languages available to every developer, the evolution of the open-source ecosystem, and the integration of big-corp features into the open standards of the web, self-hosting does not have to mean settling for less. This has played out most clearly, again, with email. If you wanted to email people, you used to have two choices: assemble the world’s largest house of cards and hope the incantation you crafted from terrifying words that make no sense like dovecot</a> and mailcow</a>—which represent interconnected services that do things you only sort of understand—actually works with any meaningful consistency, or sell out to the Oligarchs. No longer! Want one open-source binary to take care of sending, receiving, security, accounts, redirects, and everything else? It exists, and it’s called maddy</a>, and I’ve been using it to host my email for years now with no effort. git</code>, on the other hand, is very much still in the first stage. You can, of course, use GitHub</a>. It is a colossus, practically unavoidable as a developer, and enabled my start as a programmer. The pull request collaboration paradigm they created has totally supplanted the email-based method</a> that preceded it. But if you want your code to live outside of Microsoft’s proprietary claws, the process is far from simple. My own setup</a> is representative of this problem, and quite janky. It’s a delicate combination of stagit</code></a> and the built-in git-daemon</code></a>, running via Docker Compose, pointed at the home folder of a bespoke git</code> user. I push to the server over ssh</code>, relying on some weird nested home directory hack that makes me very uncomfortable in order to have a nice destination string. To publish a repo I have to create two separate magic files, git-daemon-export-ok</code> and stagit-export-ok</code>. Metadata lives in aptly-named files like description</code> and owner</code>.2</a> I clone and pull with the daemon’s git://</code> protocol. Any changes mean I have to restart stagit</code> so it will generate the relevant pages again. Collaboration is allegedly facilitated over email with a crabmail</a> instance, but it’s been broken since I first attempted to set it up. The whole thing is rather delicate and rough around the edges, which is maybe the worst combination of characteristics for an important service. I think there’s a real opportunity here to build something that could have an outsized impact on the self-hosted web. I don’t want my git setup to evoke the same feelings as the word “modem.” I don’t want to compromise. I want big tech capabilities in a small and efficient package. So let’s leverage ecosystem advancements, think deeply about how we can facilitate collaboration</a>, and make something. Keep an eye out for some form of tangible</a> deliverable</a> in this area sometime (hopefully) soon. I’m probably going to call it Tragit. I have met a profoundly depressing amount of young programmers who cannot distinguish the latter two. ↩</a> </li> The latter of which, of course, is purely aesthetic and has no bearing on security or access rights. ↩</a> </li> </ol> </section> Stanford Quarterly Reflection (Y1Q3) 2024-07-18T00:00:00+00:00 This quarter was a radical departure from the previous two. It was profoundly different and, at times, extremely difficult. Having now come up for air on the other side, and looking at the great joys that I did experience, I feel pretty positively about the whole affair. And the change never stops, as I leave behind the great change of this year for a whole new world: summer. I was in the lowest amount of units possible, a measly 12. These were: COLLEGE 110: The Spirit of Democracy</li> CS 107: Computer Organization and Systems</li> CS 12SI: Spatial Computing Workshop</li> SUSTAIN 132: Sustainable Societies Lab: Exploring Israel’s Innovation Ecosystem in Human & Planetary Health</li> </ol> This significantly reduced courseload was a mistake. I felt unmoored with so little work, attending to my classes by catching assignments just before they fell off my metaphorical plate. I don’t believe this was the most effective means of learning or academic growth. I ended up in this situation because I was afraid to adapt my schedule to change, and I could have avoided it by embracing risk and authoritative decision-making. “Strong opinions, loosely held” feels like a phrase that could be adapted here. I had structured my schedule around a specific class (DESIGN 11) but ended up dropping the class in exasperation after two weeks. I should have found a supplement (likely DESIGN 1 or ARABLANG 21A), and reorganized my schedule to accommodate it. The past being immutable, I will simply acknowledge that I believe this would have been a better choice for me academically and, in certain respects, emotionally as well. Nevertheless, certain aspects of my academic experience during these ten weeks deserve recognition. CS 107 introduced me to genuinely new aspects of computer science—it forced me to engage with topics that I have intentionally avoided in my own work, grapple with them, and come out fully understanding them. These subjects include working in memory-unsafe languages such as C, reading and interpreting x86_64 assembly, certain UNIX system details, and other topics in this orbit. I received a relatively poor grade in this class! This was largely due to my handling of the homework, and one assignment in particular which was significantly late. It was still one of the most fulfilling learning experiences of my engineering career. The highlight of COLLEGE 110 was the opportunity it afforded me to experience Larry Diamond in his element. We exist at a time of great consequence, and his framing of the world, with its challenges and opportunities, was enlightening. I feel very fortunate to have been able to get to know the man. The events I attended with him that paralleled the class were also remarkable. I met Salam Fayyad</a>, and gained a new two-truths-and-a-lie fact from our personal interactions. I also enjoyed the Boiling Moat</a> panel on Taiwanese security, which I attended with my friend Jacob</a> while he was visiting. One panelist absolutely terrified me, and I would like to work for him. The ebook is living on my Kindle, and waits high up on my reading list. CS 12SI and SUSTAIN 132 both represent the unparalleled latitude that Stanford presents its community. A class exclusively diving into the just-released Vision Pro, with real expert access and attention to each student project, is an unbelievable offering. Like CS 40, CS 12SI was new a little bumpy, but I can see it growing into something extremely impressive and startup-accelerator-y, especially under Gabriel’s</a> excellent guidance. SUSTAIN 132 lives in such a cool niche of academic study I am grateful that it exists at all—and it helped that the teaching staff was cool on top of it all. Social engagement was without a doubt where my focus lay this past quarter. My sleep schedule lurched around the clock, Oura</a> be damned. And pouring all of my energy into a complex web of human relationships occasionally reminded me of the gossip trap</a> theory of prehistory, which was less than pleasant. I should spare a moment to appreciate and bid a very depressing goodbye to the seniors who made this year special for me, and will not be there when I return. Losing you all really took me by surprise, and I will not be the same without you. Emily, you’ve been there for me like no one else and I cannot acknowledge that enough. I’m rooting for (and counting on, honestly) your return. Ben, you changed the trajectory of my career. I was honored to receive some of your reflections in my inbox, and do hope to see that continue, if only for my own enrichment. Joven, we met for the first time in earnest over spring break, but by the time we were the last two people in London, enjoying blueberry muffins and sipping on lattes on our way to the airport, we had hit our stride. No one else will debate the finer points of café society</a> with me, and that is a deficiency on their part. And of course, Sophia—I was perpetually amazed at our relationship. You’re the best, and I have faith that there is little that can stop you from continuing to be just that. In a somewhat surprising turn of events, this quarter I joined the fraternity Phi Kappa Psi. This unique aspect of American university life played no role in my parents’ experience, and thus I was raised without a great desire to participate in it when I went off to Stanford. Nevertheless, as I met more and more Phi Psis, I was startled to realize that I really enjoyed hanging out with them. I no longer find this nearly as startling, having been to the ends of the Earth and back with them (or at least, military bases, casinos, and watery watchtowers). To the brothers that convinced me to join—Odin, Reisner, Lichu, Deveen, Deitrich, Gareth, Farman, Santi, Milo—I owe you a great thanks, and perhaps an apology as well for being so stubborn. I’ve wanted to be a rapper for as long as I can remember, and you brought that out of me. I’m extremely grateful. And to the fresh set of brothers—Bradley, Panos, Arjin, Tiao—that I am a part of, you make me excited for the future. I have not forgotten about the London gang, nor my promise to write about the trip in this reflection. For those readers planning an outing to the kingdom on which the sun never sets, I have only three important notes—the white saffron hot chocolate at Dark Sugars</a> will change your life, the chai at Dishoom</a> is bottomless, and if you’re ever having too much fun at an interactive performance of Guys and Dolls</a> the dude to your left will berate you. My deep thanks to the Danielpours for playing host to us all, and welcoming us into your home for a lovely shabbat. Everyone on this trip changed my life this year, and I love you all so very much. In London and at Stanford, to have such friends is life’s greatest joy. Many of you have made your way into this post already, but I will descend into lists again for those that have not. Stella, you built a home on campus in the Warehaus that did not have to exist, but was for me a heaven of frosted flakes and late night antics. Zach, I’m coming for your PRs and your creatine water bottle. You undoubtedly fill out the commemorative London shirt the best of all of us. Daniel, you inspire me consistently by moving to action when I’m standing still, afraid. Name a more iconic duo</a> than us. And Kelly—you are so wonderful. But these are not the only people at Stanford who are important to me. Vivek, I cannot wait to live with you next year. Nate, you are simply way better at tennis than I am. Trun, we are living the lives we have always wanted. Jack, if there is anyone in my life who possesses the magnetic people powers of Gatsby (which I say as someone who, at present, admires the man with no qualms) it is you. Sam—to meet your family was a great honor and I’m sorry they saw me give you a piggyback ride right before. Vedant and Riya: just because we’re no longer all in the great dorm of Crothers doesn’t mean tea at 3 in the morning has to stop, and indeed I hope it doesn’t. Ryan, you’ve once again ended up a key player in a crazy highlight-of-the-quarter evening and I can only hope that this continues. Longhair, it didn’t work out this quarter, but come fall you bet we will hit the Pyramid of Giza. And Naama: you are the only person with whom I have a weekly scheduled meal. I miss you all. It occurs to me that turning the personal section of a reflection into basically a list of thank-yous is perhaps not particularly reflective. It’s somewhat hard to keep a journal in public. Perhaps this is just the nature of things. Perhaps I have just once again waited too long to finish writing and publish this post. Perhaps I shouldn’t be married to the format I’ve used in previous posts. We shall see. I remain the person that I was when I began Stanford, but I happily believe that have become an improved and refined version of that person. I have aquired two new greenfield skills, in Arabic and assembly language. I had no knowledge of the two before this year, and now I do. These allow me to tangibly do things I could not do before. The majority of my development, however, has been less tangible—but all the more profound. Being at Stanford has placed me at the center of the most incredible environment I could imagine; remarkable people surround me at all hours of the day, and their effusive brilliance cannot help but be impactful. As the summer continues, and the community that comprises this small Eden is dispersed, I have a hard time imagining that this year happened at all. My time at Stanford was a short, recent burst. I’m now back in San Francisco, and though I’ve maintained as large a sprinkling of Stanford in my life as I can, it feels almost as if I have just awoken from the most vivid, marvelous dream. This summer will be exciting and hopefully rejuvenating as well. It will also comprise its own Q4 reflection! But I’m not embarrassed to say that I’m looking forward to returning to the Farm in the fall. I enjoyed dreaming, and I have a feeling that it’s not over yet. Running Out of Time Wallpaper 2024-07-05T00:00:00+00:00 A minimalist macOS wallpaper to remind you that time is ticking. My little sister was listening to the Hamilton soundtrack</a> yesterday, and as I was listening I got hung up on a particular line. How do you write like you’re running out of time? </blockquote> I remember my older sister once told me a story that Steve Jobs had a calendar on the wall with a circle for every week in the average lifespan. At the end of every week, she said, Jobs would fill in a circle. One down, a discrete and finite amount left to go. The sands of time</a>. Memento</a> mori</a>. You are running out of time. The wallpaper was made in Affinity Photo. It responds dynamically to light and dark mode thanks to Equinox</a>. The font smells of sea foam</a>. Download it</a> for yourself! Caret 2024-05-22T00:00:00+00:00 Caret</a> is a command line tool for browsing Lobsters in your terminal. It brings the Hacker News alternative just a few keystrokes away in the environment that you’re probably already using. Caret is heavily inspired by the excellent circumflex</code></a>. It attempts to emulate much of the same functionality with a much simpler implementation, using the Charm</a> open-source libraries. Caret relies on Lobsters’ unofficial JSON API, and will adjust to changes in this potentially volatile source of data as needed. Caret is debatably a tangible deliverable</a>. I developed it before I announced this intention, but the announcement has been long delayed, so who knows. Many programmers—myself included—practically live in the terminal. It’s a fast-paced, information-dense environment. Caret brings Lobsters to meet its readers where they are, and enables readers to navigate the site quickly and effectively in a manner familiar to them. Compile Caret from the source code on my forge</a> and give it a go</a> (no pun intended) in your terminal. CSS System Colors Dashboard 2024-05-20T00:00:00+00:00 CSS System Colors Dashboard</a> is a website that gives developers a visual preview of CSS’ underused system-color</code> data type. These colors are context-dependent, changing between operating systems and browsers; the dashboard provides effective visualization of the colors across environments. I discovered native colors in CSS through Jim Nielsen</a>. These colors are extremely useful for developing “along the grain” of a platform. I now use them all the time! This new dashboard</a> lists the available system colors and displays their value in context. The dashboard is a tangible deliverable</a>. It’s a simple, cute tool with a lot of utility. This project was developed in part as an experiment in learning Rust. I used axum</a> and askama</a> in building the site, selected for their (aspirationally) intuitive interfaces and strong embrace of the type system. However, I found the development experience in the language overcomplicated and unergonomic. I still hope to enter the Rust community—I am enthusiastic about the ideals of the language and its potential—but will seek out other ways to do so. Check out the live website</a> in your browser to explore your system colors and view the source code on my forge</a>. Stanford Quarterly Reflection (Y1Q2) 2024-04-20T00:00:00+00:00 In this quarter, Stanford became my default. As such, my memory of my time at Stanford has begun to take the blurry and general form I associate with “real life.” My workload increased this past quarter, as I left the one-unit courses behind in favor of some required humanities courses and entertaining medium-unit classes. I received my first A+, and also my first A-, in my Stanford career. On balance, this lowered my GPA by 0.01, as the unit counts were skewed in favor of the A-. I took five courses: ARABLANG 2A: Accelerated First-Year Arabic, Part II</li> CHEM 29N: Chemistry in the Kitchen</li> COLLEGE 102: Citizenship in the 21st Century</li> CS 40: Cloud Infrastructure and Scalable Application Deployment</li> PWR 1KAA: Writing & Rhetoric 1: Forward Momentum: Writing About Movement(s)</li> </ol> Arabic continued to be a positive staple of my week. It also expanded its impact, when I took a late-night excursion into San Francisco with a friend visiting from UChicago and we tried out a pizza place recommended on the authority of Khaled. It was lovely! The proprietor remembered Khaled from his days at USF. We also went on later that night to discover a crepe place near where we went to high school, which was surprising as we had not encountered it before and because restaurants open past midnight are rare gems in San Francisco. I got back to Stanford very late. CHEM 29N was a total treat, if not particularly rigorous in either its chemistry or its cooking. It is stuck in the classic funding limbo, wherein the program is given little money due to its low output but cannot increase its effectiveness until it receives more money. Still, I got a lovely addition to my fledgling apron collection. COLLEGE was aggressively mediocre. There is great value in the liberal arts and indeed they are fundamental to Stanford as an institution. You will, however, not find this value in COLLEGE 102. The class is a compromise: primarily between those who want to require a liberal arts core and those opposed as well as between those who want to overhaul the Western canon and those who do not.1</a> The end result is boring. CS 40 was a course super relevant to my day-to-day coding work and a lovely introduction to the exciting world of 3-unit courses, but as it was its first quarter running there were still some kinks to work out. “Figure out programmatic declaration of self-hosted services” has been on my todo list for quite a long time, so I was excited to both learn how to do this and get school credit for it; however, instead, I just spent a few weekends spray-and-praying AWS CDK gibberish at the screen. I understand that they’re going to switch to Terraform or Ansible next year, and I wish them luck with that (and myself luck with self-study of the same). Ideally, I would have come away from the class with the skills required to write cloud-agnostic declarative infrastructure, but I have not. Unless you’re looking to hire me, in which case I absolutely have. I am unable to imagine a more painful academic experience than PWR 1KAA. To gather the world’s best and brightest—driven, talented youth—at great expense in money, time, and effort, and then force them to divert significant effort into this aimless toil feels criminal. It is astounding that this has been allowed to occur. To absorb the material of this course would be a detriment to your writing ability. And though I suspect there would have been no PWR 1 courses that I would have loved, it did not have to be a complete failure—for that, the instructor is responsible. As explanation, and a means of self-restraint, I will simply deliver the following anecdote: A very good friend of mine, in a different PWR course, had scheduled a session with a tutor to work on his final essay. His tutor canceled, and he was assigned my PWR instructor in their stead. He left the meeting astounded, with no actionable advice and significantly more confused. He had some choice words, and told me genuinely that he felt sorry for me. I pray, sincerely, that I never again encounter anything like this course during my time here. I fear that I have waited too long, and have become too engrossed in the day-to-day of this new quarter, to do this section as I would have liked. But I will appreciate people nevertheless, and devour my camera role repeatedly as I seek to return to the proper mental state. I want to appreciate Nika and her illegal bunny. I want to appreciate Daniel and Ryan for a raucous night of festivities; I admire Daniel greatly for his ability to commit and his fearlessness in social situations, and I am grateful to Ryan for his non-stop encouragement. Huge love to Jack and Sam, who tried the Vision Pro with me and got many a late-night Zareen’s. Kelly, for an even later Zareen’s, failed attempts at glasses shopping, and lovely brownies besides—you are a light. Vivek for climbing on the Shangri-La scaffolding. Nate for hitting 195 before me. The JSA Retreat stands out as a highlight of my experience this winter. The chances of finding such a space as we created in that house are less than one-in-a-million, and I am in awe that it exists here at Stanford. I would not be the same without it, and I don’t think it an exaggeration to say that it is the defining part of my Stanford experience. Diego’s cooking was not half bad, and I love I Love London.2</a> I was warned that winter would be dreary and horrible. I’m just not sure what people are on about; this was my favorite quarter yet.3</a> Let the record show that, on these fronts, my positions currently stand as follows: I am opposed a required core at Stanford, and broadly aligned with the idea of updating the Western core. To the former, I think the increasing number of required courses is bad; the WAYS system, in which you are required to study specific fields but have a great deal of choice in how you do so, is significantly more aligned with the spirit and culture of Stanford. If you want a core, and the specific impacts that having such structure brings, look elsewhere—aforementioned UChicago friend is thoroughly enjoying his time on the school’s intense and rigid path. We should not attempt to backport this to Stanford. To the latter issue, I am hopeful that we have advanced as a species since much of the classics were written. I do not hold these texts sacred. I do, however, hold excellence sacred—and don’t think updating this corpus will be as easy as cutting a few pieces and throwing in a few modern works from traditionally marginalized voices. This process needs to be done well and by actual experts, which I do not believe has happened yet in a curriculum I have encountered. It has certainly not happened in the COLLEGE department, where courses are taught by brand-new “teaching fellows” who agree upon a core curriculum but all refuse to teach it and instead occasionally do and say things ranging from the abhorrent (see: Ameer Loggins) to the casually wrong (my teacher suggested, and doubled down on, a claim that China and the Iranian Revolutionary Guard are more accountable to the citizenry of their nations than the United States government). ↩</a> </li> Which, while we are on the subject: I’m saving my spring break trip for my Q3 post. ↩</a> </li> Small sample size, but still. ↩</a> </li> </ol> </section> A Lovely Read 2024-04-20T00:00:00+00:00 The Seaside Hotel by Michael Holt. Finished on April 17, 2024. Rating: ★★★★☆ I had the great privilege of studying under Michael Holt for two years. When, two weeks ago, a friend forwarded me and a fellow student of Holt’s the link to his new book</a>, the three of us immediately purchased copies and instituted a haphazard book club. I think it is a very beautiful thing that I am able to read a physical work written by a friend. Holt has done an incredible thing through the creation of this tangible relic of his thought and conciousness. I find it very inspiring. It makes me consider writing in a way more serious than I ever have before. The form of the book is independently notable. I am a proponent of compact, purpose-built objects. I love my Kindle</a> because its existence is limited to precisely what is necessary and nothing more. Any object</a> can be similarly refined. The Seaside Hotel is such an object. Its unique dimensions are subtly unsettling, subverting your assumptions to break you into a concious focus. Less poetically: it is small and almost square, light and fun in the hand. I read the book sitting on a sunlit porch, and then tucked away stretching out across the length of a light-brown couch. Though I am concerned this says something about my current mental state, The Seaside Hotel brought me a feeling of contentment. I recommend it to all. Though it almost certainly goes without saying, I am excited to see where Holt goes in the coming years and will follow his works with great interest. I Suppose This Is Modern Fiction And It's Pretty Good 2024-04-12T00:00:00+00:00 The Seven Moons of Maali Almeida by Shehan Karunatilaka. Finished on April 01, 2024. Rating: ★★★★☆ When I started this book, I thought “ahh yes a book about the ghost of a dead gay Sri Lankan photojournalist.” By the end, I was thinking “ahh yes a book about the ghost of a dead gay Sri Lankan photojournalist!” This paragraph makes more sense out loud, where the emphasis is clearer, but the crux of my point is that at first I was bewildered by its subject matter but by the end of the book I was thoroughly appreciative of the work and in fact enjoyed it very much. I was looking for a book to get me out of my reading slump. I had started several, and was making little headway through any of them. When I began on this book I would not have thought it would be any different. Yet it was! The world is fascinating and its mythology completely original. Maali is complicated and, as such, makes me want to keep unraveling his story. I learned a lot about Sri Lanka through reading this. This has had immediate payoff as I have been making niche jokes and references to my Sri Lankan friend and really catching him off guard. I should probably back this up with some actual historical research given the book’s general lack of commitment to the nature of reality. It was good to read some surrealist fiction. I haven’t done much of this since last spring, when I was in a class on the subject. There are actually cool things to come on that specific front, which I will try and remember to link here when they come to fruition. If you are looking for something to intrigue you and sort of make your brain cells feel fuzzy, this might be the book for you. Toward Tangible Deliverables 2024-02-05T00:00:00+00:00 I would like to become the kind of person who consistently produces discrete works. I am a big believer in compounding success. You can get used to thriving, get used to a mindset of novel thought, get used to creating rather than consuming. You can make it a habit. We can game our psychology here. However productive you may be in reality, it is hard to feel the progress being made without distinct milestones. Chipping away at a project of amorphous and twisting scope is quicksand for your time. Dividing up large goals is key to creating a feeling of control and—critically—doing real work. I admire the programmers and designers who throw off project after project into the public eye. My Mastodon timeline is chock-full of mockups and demos, people actually shipping all the time. So, I’m working on it. This next part is for internal use, but browse at your leisure. In: blog posts, fun web projects, “mini apps” for Apple platforms, microblogging but with images/gifs, emailing other internet people, /now</code> and /using</code>, videos, livestreaming, art, bamboo (any usage). Out: project multitasking, not publishing, unclear distribution, false productivity. Links: Matt Webb</a>, Julian Lehr</a>, Neil Sardesai</a>, Vinoth Ragunathan</a>, Gavin Nelson</a>, Daring Fireball</a>, Casey Neistat</a>, Marty of Poolsuite</a>, todepond</a>, Robin Rendle</a>. Stanford Quarterly Reflection (Y1Q1) 2024-01-09T00:00:00+00:00 Life changes slowly and then all at once. I started at Stanford University on September 19th, 2023, an “all at once” day that began the rest of my life. It is no small thing to leave the home you’ve known all your life for a place as strange as Stanford. I am profoundly grateful for the opportunity to do so and for the people who have made that possible. It turns out to be a wonderful kind of strange. This past quarter felt like being thrown onto a bucking horse, breaking it, and immediately turning it toward the racetrack and joining a heat that was already underway.1</a> It was immediately fun. People are supremely important. Having people you care about to eat with when the evening comes—this one thing can make anywhere feel like home. This is not a profound realization, but it is one that I have come to through my time here. I have been extremely lucky to be joined at Stanford by one of my best friends from high school. It’s completely cheating: I am certain that the transition would have been far more difficult on my own. To leave home and be totally alone is an experience I still have not had: when I moved to Stanford I had Vivek. I have been further showered with folks who I am grateful for, new friends that—as anyone who has gone off to college will tell you—I already feel like I’ve known all my life. What I most want to say is this: when I went off to college, it was my most sincere hope that I would meet remarkable people. I have found this sort of person at Stanford in abundance, and nowhere more than with you, my friends. Thank you Jack, for welcoming me into your home—and giving me a potato for my shelf. Thank you Josh, for debating with me the linguistic intricacies of Hebrew curse words. Thank you Daniel, Riya, and Amalia, for hanging out with me in far-flung cities; it’s remarkable to land somewhere strange and have friends there to welcome you. And thank you Kelly, for introducing me to Corner</a>, and more importantly being everything for everyone in difficult times. And to all those with whom I shared food, ran around in the cold, stayed up into the wee hours of the night, and woke up for in the early morning to go lift; it’s been a lifetime these few short months and I am amazed I have been able to spend it with you. I took six classes, which I found after a confused two-week sprint where I picked up and quickly dropped, among others, a three-hour long comparative literature class.2</a> My schedule eventually settled into the following: CS 106B: Programming Abstractions</li> ARABLANG 1A: Accelerated First Year Arabic</li> HISTORY 81B: Contemporary Middle East</li> CS 7: Personal Finance for Engineers</li> PHYSICS 59: Frontiers of Physics Research</li> ECON 3: Big Ideas Lecture Series</li> </ol> I loved this schedule, and am quite lucky to have stumbled into it. Sean, my 106B professor, is wonderful and an endless well of millennial, self-deprecating positive energy. I hope nothing ever stays that spirit. My Arabic professor Khaled has an incredible tendency to launch off on half-hour long tangents in the middle of a lesson. It was in his class that I saw my name for the first time: We had already learned the alphabet when Khaled used my name as an example. I looked up at the chalkboard in amazement.3</a> When did I first see my name in English or Hebrew? This time, I snapped a picture. HISTORY 81B fell into the trap of most Middle Eastern history courses: it taught the same old Arab-centric narrative, depriving all characters of agency and giving the state of the region an air of inevitability. I took it to hang out with Vedant and Hassan—which I did! Our post-class coffee routine caused me to be late to Arabic on many an occasion. I also enjoyed our final project, which I published here</a> as stylistic practice for an eventual submission to the ACX book review contest.4</a> The final three classes were all speaker series. Whenever I tell this to upperclassmen, they universally respond with a shoulder pat and a wistful gaze as they reflect back on when they were young and naive and took speaker series. I hope I never come to adopt this position: I think they’re the greatest. In PHYSICS 59 I heard the sound of a single atom moving. CS 7 filled a huge gap in my knowledge and helped me contextualize my finances.5</a> And ECON 3 is one of those classes that can only exist with the unique resources found here. The observation of change is a difficult thing. I find it almost impossible to embody the version of myself that was once living so differently. The blank slate of college is a canvas for personal change. I’m trying to use it to live as intentionally as possible.6</a> When you are in control of all the facets of your life, why not design it? The amount of distinct events that occur each day in my life on this campus is incredible. Perhaps one of the things I appreciate most about Stanford is this compression of time and space. There are so many people doing so many things here, that each day doesn’t truly feel complete until something unique and memorable has occurred. Last night, for example, I spontaneously dropped in on a class in the law school a friend of mine is taking that just so happens to be taught by the father of a high school friend.7</a> How glorious! The second quarter is already upon us. This time around there will be a lot more writing of papers. And yet, I am sure it will be—as this last quarter has been—genuinely exceptional. עם ישראל חי I have never done anything remotely like this. It is a metaphor. Your horse-wrangling mileage may vary. ↩</a> </li> COMPLIT 214: Shipwrecks and Backlands, mostly inspired by an excellent course on surreal Latin American literature in high school; thalassic Iberian literature seemed like a good next step. Ultimately, I thought the discussion format combined with the extended duration made it not a good fit for what I would like to do at Stanford. I was also briefly in SYMSYS 1: Minds and Machines, but was told that I would receive several zeroes on the first week’s work and my grade would suffer. I believe that’s not how the shopping period is supposed to work, so I dropped out in search of classes excited to have me. ↩</a> </li> I know, right? Chalkboard! I’d never really seen one of those before. Amusing that in the middle of Silicon Valley, the chalkboard is alive and well in the depths of Main Quad. ↩</a> </li> War of Return was not assigned specifically, we were just tasked with writing a book review. The review I will eventually submit to the ACX contest will likely have decreased usage of “Ibid.” ↩</a> </li> Let it also be known that the lecturer, Adam Nash, is extremely cool, has nice hair, and writes a blog of his own</a>. ↩</a> </li> Sleep well, be jacked. ↩</a> </li> Because, it seems, taking law school classes as an undergraduate freshman is something you can do. If you’re insane. ↩</a> </li> </ol> </section> Epic Acid Science Fiction 2023-12-26T00:00:00+00:00 Dune by Frank Herbert. Finished on December 25, 2023. Rating: ★★★★☆ Really on a “very good, but not my favorite of all time” streak here. Also, I am now setting the precedent that not all reviews will be extremely in-depth. In some ways, these are less reviews and more a reflective log with occasional critique. Regardless: Dune rocks. I first encountered this book as a wee lad when my older sister read the series. I am happy to have finished the first book all these years later. It’s such a cool world! I’m a sucker for the desert, and I think this book does it very well. I would love to see Arrakis. There are many different languages in Dune. The Atreides have a battle language. Other families have their own secret languages. Fremen have their own tongue, which incorporate words from another, third language. I already speak a somewhat esoteric language that covers me in most American contexts, but I would like to have my own secret language to communicate in true privacy. Super Humongous Spoilers</summary> The final battle nearly gave me a heart attack. Very stressful. Then it ended with no George R. R. Martin-esque twist, and instead Paul basically totally won against the entire universe. Perhaps intellectually disappointing, but very satisfying. </details> Dune is long as hell. I took advantage of a cross-country flight to knock out most of it. This will factor into my prioritization of finishing the series. Then again, I’m flying back across the country at the end of the week… A Foray into Theology 2023-12-20T00:00:00+00:00 The Lonely Man of Faith by Joseph B. Soloveitchik. Finished on December 17, 2023. Rating: ★★★★☆ The Lonely Man of Faith was a birthday present from a dear friend. It is written from a perspective that challenges my own, running quite contrary in some aspects from the way in which I have chosen to live. Indeed, this is what made it a fascinating book to read: the interaction of differing opinions and views is where there is perhaps the most room for growth. Furthermore, The Lonely Man of Faith is written at quite an abstract level, and as such there are many things within its pages that Rabbi Soloveitchik and I agree on and indeed within which we can find shared meaning. Now, assorted quotes and commentary: The Halakhah believes that there is only one world—not divisible into secular and hallowed sectors… Page 79 </blockquote> This aligns with how I think of Jewishness. I am a Jew in every context, not just those that may be considered “religious” by those on the outside. The dual-nature image of man proposed by Rabbi Soloveitchik is this book’s central focus. At times I find this description becomes fairly romantic: However, [the discovery of a companion], since it is part of the redemptive gesture, must also be sacrificial. The medium of attaining full redemption is, again, defeat. This new companionship is not attained through conquest, but through surrender and retreat. “And the eternal God caused an overpowering sleep to fall upon the man.” Adam was overpowered and defeated—and in defeat he found his companion. Page 38 </blockquote> The emphasis above is mine. There is something I enjoy about the framing of deep human relationships being formed through mutual surrender. If the Rabbi were to appraise my life, I am certain he would say that I spend far too much time in the conquesting mode of his “Adam the first,” and it may be beneficial for me to explore other ways of existing. Perhaps there is value in this. I am certainly sympathetic to, and involved myself with, the individual search for meaning. However: I do think there is a vast chasm between that and not touching light switches on Shabbat. The Lonely Man of Faith is a book of philosophy—it is difficult to draw from it hard conclusions. Much of the worth of these things comes from the reading itself, and clashing with the ideas on the page. The journey, not the destination, as it is said. The book is not long. I enjoyed it. Quotes from My Camera Roll 2023-12-19T00:00:00+00:00 I have been collecting fun little writings in Photos with the intent of distributing them here. Now I can delete them. This will lift a great weight from my shoulders. It will all be gone soon enough and you know it, so look good and close while the close lookin’s good. </blockquote> “Close lookin.” Musk, via email, would only say: “I have never lost money for those who invest in me and I am not starting now.” </blockquote> From an Axios article, in response to a question about X/Twitter. Seems ill-fated, but an impressive display of confidence. Argentina, back in its familiar position of being a fiscal and economic disaster zone… </blockquote> Funny, for me, far away from Argentina. This put him in position to try the Hitler Slingshot Maneuver… </blockquote> Hell of a maneuver. The old mantra’s gotta go, even as it still rings in our ears and echoes down deep caverns of the Web. </blockquote> A metaphor that makes the Web a physical place is good in my book. We’re imagination bottlenecked. There’s low-hanging fruit for the next decade. </blockquote> A fun but frustrating concept. How to come up with radical ideas… The only other thing I will add is that I don’t think anyone needs to justify asking for support through “hosting costs money” or the like. Art is worth supporting, so even if the money I send goes to paying for your truffle brownie habit, you do you. </blockquote> From Manu’s lovely People and Blogs, the fourth edition</a>. A philosophy I agree with, even if I am not huge on truffle brownies myself. Don’t avoid places with beautiful women in them, ignore who’s already there and take a beautiful woman with you, or at least one who looks beautiful in bistro lighting. Trust me, the food tastes better. </blockquote> From one of the most bizarre and beautiful screeds</a> I’ve ever encountered. Let the coffee be hot, & the liquors chosen with special care. </blockquote> A local restaurant menu. A Review That I Should Probably Have Saved For Another, Less Contentious Time But Nevertheless Did and Ultimately Enjoyed 2023-12-16T00:00:00+00:00 The War of Return by Adi Schwartz and Einat Wilf. Finished on December 03, 2023. Rating: ★★★★☆ The War of Return is a book written by two self-described “prominent Israeli leftists” that makes a bold claim in its subtitle: “Western Indulgence of the Palestinian Dream Has Obstructed the Path to Peace.”1</a> The book was originally published, in Hebrew, in 2018. The English translation debuted two years after in 2020. Three years later, and three months after the October 7th attacks, those who hope for peace seem to have more reason for despair than ever. I read War of Return hoping for a fresh perspective. So together, let’s turn to co-authors Adi Schwartz—an Israeli journalist formerly at Haaretz, the country’s largest left-wing newspaper—and Einat Wilf—a former Labor party politician—to discuss “the single largest obstacle to lasting peace” and how we might go about solving this seemingly intractable conflict capturing the attention of the world.2</a> War of Return</h2> The Arab-Israeli conflict, stretching now for over seven decades and continuing to impact the lives of millions of people throughout the Middle East, has had numerous near misses with peace. This is because of a fundamentally incorrect assumption often held by Israeli and Western negotiators that “this [is] a territorial conflict that [can] be solved by partitioning the land into two states, that the Palestinians only [want] a state of their own in the territories, and that the Israeli occupation and the settlements [are] the primary obstacle preventing peace.”3</a>4</a> According to Schwartz and Wilf, they have misunderstood the problem. In reality, our favorite co-authors say, this is an ideological conflict over the very existence of Israel. It is an absolute rejection of minority sovereignty in a region dominated by the Arabs, and it is not new: “…the belief that Zionism was an outrageous injustice predated the war and caused the Arabs to violently oppose the Jewish national liberation movement many decades earlier.”5</a> The territorial framing of the conflict is thus not conducive to a solution—no state, with any borders, will appease a movement that feels like partition itself is an injustice. …this was not a conflict between two national movements, each seeking first and foremost its own independence, but rather about one group (the Arabs) seeking first and foremost to foil the independence of another (the Jews).6</a> </blockquote> Palestinians’ claimed right of return, in the minds of Schwartz and Wilf, is a deceptively named aspiration to negate Jewish self-determination. There can be no democratic Jewish sovereignty where Jews are a minority.7</a> It is thus that the Palestinian right of return is “not merely about moving ten or twenty miles to homes left behind, but primarily about returning to the time before the terrible defeat of the Nakba and the establishment of the state of Israel,” by making Jews a minority in their own homeland.8</a> In essence, it is about rewinding history and undoing the creation of the Jewish state. I myself misunderstood what was meant by “right of return” before reading this book. As the phrase is used, the right of return is the right of Palestinians to return not to a future Palestinian state in the West Bank and Gaza—as I had believed, and supported—but in fact, to the sovereign state of Israel. Return “could only be realized in the territory of the state of Israel atop the ruins of the Jewish right of self-determination.”9</a> To mandate a right of return to Israeli territory is to reject “the principle of territorial partition” and a two-state solution.10</a> War of Return attributes the creation of the refugee issue to this rejection of the two-state solution by the Arabs. They argue that if the Arabs had accepted partition in 1948 and established a state of their own, no one would have been displaced:11</a> …it is a fact that the departure of the Arabs was a result of the war and only of the war. Before the Arabs waged war against partition, they did not leave their homes. The Arab flight and the refugees from the war were neither inevitable nor necessary nor inherent in Zionism.12</a> </blockquote> The blame for the displacement of the refugees, they claim, is unflinchingly on the shoulders of the Arabs, not the Jews. No one is entitled to the status quo ante: Those who wage war to eliminate another people, and to prevent their achieving independence, cannot legitimately complain that “they suffered an exceptional injustice” when they lose and flee the land.13</a> </blockquote> Furthermore, according to Schwartz and Wilf, the legal right of return simply does not exist: “No legal obligation or treaty existed that… obliged Israel to let [Palestinians] return to its territory” in the aftermath of the 1948 war.14</a> Flight and expulsions occurred throughout the 20th century, including during this seminal war. Indeed, unlike in Israel where an Arab minority remained after the war, “not a single Jew remained in the areas conquered by Arab forces.”15</a> How can it be that the Palestinian refugee problem still exists 75 years later, and at a greater scale than its start? “The answer to why the Palestinian refugee problem still exists lies neither in the conditions of its birth nor in its scale nor in the number of victims: nothing here is unique. The answer must lie elsewhere.”16</a> That elsewhere, War of Return posits, is in the refusal of the Arabs “to solve the [conflict] by creating a new status quo in the Middle East” in which Jews and Arabs could have exercised self-determination side-by-side—which was accomplished through the political manipulation and exacerbation of the Palestinian refugee issue.17</a> Schwartz and Wilf argue that the critical issue is the Palestinians’ demand for return to Israel. Progress toward peace is made on all fronts but return, the “one article that Israel [can] absolutely not agree to, as it [entails] its very suicide.”18</a> Return is instead silently propped up by Arab support—all but guaranteeing a continued, violent existence for Palestinians and all others in the region. Promising initiatives to resettle hundreds of thousands of refugees in the immediate aftermath of the 1948 war, in the Jordan Valley and the Sinai, went nowhere. The “biggest rehabilitation project of the 1950s for Palestinian refugees,” a farm run by Musa Alami (a prominent Palestinian nationalist), employed thousands of Palestinians in the Jericho area—with a specific focus on orphans of the war—and grew orchards and productive crops over thousands of dunams, with export contracts to Saudi Arabia, fifty wells, and a school.19</a> It was leveled in 1955 by Palestinians believing its existence would help “enable the resolution of outstanding political disputes between the sides.”20</a> The refugee issue, claim Schwartz and Wilf, is cynically manufactured and perpetuated by Arab leaders. The long-term adoption of the position that “improving the living conditions of a few hundreds of thousands of refugees [is] less important than their war with Zionism” has led the Arab world to the creation of a Palestinian refugee-hood that is completely divorced from the experience of every other refugee group.21</a> Indeed, Palestinian refugees are not governed by UNHCR, the UN agency for refugees, but by their own temporary commission—UNRWA—the regular extension of which “has become a quasi-automatic annual tradition” at the United Nations.22</a> War of Return devotes a great amount of effort to describing how UNRWA “was transformed from being a failed agency for refugee rehabilitation to a very successful organization for” halting progress in the Middle East.23</a> “For decades,” Schwartz and Wilf say, “UNRWA has sustained a parallel world of policy and executive decisions that serve the Palestinian narrative alone,” and leave the Middle East in a radicalizing limbo that actively works against peace.24</a> The list of UNRWA oddities is very, very long. Unlike all other groups, UNRWA’s “Palestine refugee” status is hereditary—resulting in a registered population of over 5 million people from an initial group of approximately 700,000 displaced Palestinians (see chart below). Unlike all other groups, refugee status is not surrendered when additional citizenship is achieved; indeed, 2.2 million UNRWA-registered refugees are citizens of Jordan, but they retain their “refugee” status. Astoundingly, these refugees make up 70% of Jordan’s population: “It is difficult, bordering on impossible, to get a consistent answer from Jordanian officials to the question of how the Jordanian state sees its own citizens.”25</a> Another 2.2 million UNRWA-registered refugees live in the West Bank and Gaza, territories allocated for the future Palestinian state, making them refugees within their own future state. An additional million are officially split between Syria and Lebanon, territories where “most of them do not even reside… anymore.”26</a> Since the 1960s, “most of the [Palestinian] refugee camps were neighborhoods of the Arab towns next to which they were built,” with housing markets and daily realities entirely different from the Western image of vast, impoverished tent cities.27</a> Within these refugee-camps-that-are-cities, Western-funded UNRWA-run schools teach students “a narrative of victimhood, based on a singular, striking injustice,” which have resulted over time, Schwartz and Wilf believe, in a direct connection between “the perpetuation of UNRWA for political reasons to the emergence of” Palestinian terrorism.28</a> Thus, according to Schwartz and Wilf, the purpose of the continued use and expansion of refugee status in this situation is to perpetuate and reinforce the Arab claim toward the right of return and its inherent goal of eliminating Israel.29</a> The War of Return is worth reading, but is difficult to synthesize. The book is disorganized: it has a message it wishes to impress upon you, but is not sufficiently clear and driven in doing so. It is without a doubt the most thoroughly cited book I have ever encountered. A full third of its page count is dedicated to footnotes and bibliography alone.30</a> In an issue swimming in contentious Instagram infographics, Schwartz and Wilf have brought the receipts.31</a> In doing so, however, they interweave theory, history, and proscribed solutions in a manner that leaves the reader with a significantly improved understanding of the conflict but great difficulty summing up this new knowledge. The book desperately needs a more linear structure. Additionally, assertions about Palestinian thought are found throughout the book and can be difficult to prove true or false. How would one go about assessing the claim that “the Palestinians’ commitment to the idea that they are still refugees and also possess a right of return to the state of Israel is deeply embedded in the Palestinian identity and its collective ethos?”32</a> Schwartz and Wilf proffer that it “is an issue on which no Palestinian political opposition or dissent exists,” which is perhaps as good a proxy as you will find.33</a> I don’t necessarily doubt that it is correct that there is a cultural narrative of “perpetual injustice” in the Palestinian camp, but I am cognizant of the fact that it is difficult to prove definitively. The book’s arguments are made weaker by their occasional reliance on alleged Palestinian beliefs, as opposed to evidence of action. Furthermore, the book’s critique of the West—which we are led to believe by the subtitle will be severe—is, in essence, that it has failed to sufficiently counter anti-Israel extremism in the Arab world. The book makes a compelling argument that this is the case, and that “geostrategic interests” (read: oil) have muddled what would otherwise be clear opposition to an ideology that seeks to eliminate a UN member state.34</a> Still, this strikes me as a somewhat confusing target for criticism in this case when it may be more appropriate to condemn the Arab extremists themselves. With that said: this book managed to significantly change my thinking on the conflict. As someone who thinks about this a fair amount, I would consider that on its face to be a significant endorsement. If that’s not enough of an endorsement, here’s another one: you should probably read this book. I now realize that I did not at all understand the Palestinian refugee issue before reading this book, and in its aftermath feel confident and prepared in its discussion. Schwartz and Wilf did not challenge my fundamental view of the conflict, but they gave me a much clearer understanding of the refugee issue, which has profound implications. The War of Return has been a timely read ever since its publication. Indeed, in the ensuing years, very few of the foundational facts and conditions it addresses have experienced any shift; the Arab-Israeli conflict had reached somewhat of a standstill. It is the “fervent hope [of Schwartz and Wilf] that in writing this book [they] contribute in a meaningful way to real and lasting peace.” As such, their proposals need adjustment.35</a> With the outbreak of the Israel-Hamas War in the aftermath of October 7th, there is potential for a serious shift in the history of the conflict. After the war is over, there may be an opportunity—for the first time in a long time—for meaningful change. The parties must move quickly to final-status negotiations, to bring a conclusion to the violence that has plagued our peoples for decades. We can no longer think about slow change. Two states, for two peoples, as originally envisioned by the United Nations in 1947 as “lawful, moral, and legitimate” solution.36</a> In order for that to happen we must be guided, in part, by the book’s total refusal of the right of return: When Palestinians complain that recognizing a Jewish state means relinquishing the right of return, the response should be, “Yes, that is exactly what it means.”37</a> </blockquote> If indeed we stand, surrounded by violence, on the precipice of peace—the storm before the calm, if you will—then this book stands to be more relevant than ever. In the interest of full transparency, I will admit here that the subtitle begins with the word “How,” which I have not included in the quotation for the purpose of sentence flow. ↩</a> </li> Literal, “The War of Return”</a>. ↩</a> </li> Schwartz and Wilf, The War of Return, 55 ↩</a> </li> As you read this review, you may find yourself confused, thinking: “These citations seem to imply that War of Return has thousands of pages. That seems unlikely.” You would be correct. I read War of Return on a Kindle, and thus have had great difficulty finding stable page numbers. Instead, I have provided a “location.” You may ask yourself: “How do I use a location?” To which I respond, “This is a question for Amazon.” For now, I will simply apologize in advance. ↩</a> </li> Schwartz and Wilf, The War of Return, 131. ↩</a> </li> Id. at 148. ↩</a> </li> This is just definitional—you’d get outvoted. You can also look at historical examples: there was no self-determination by Jews in Arab countries, or the United States, or anywhere else that Jews lived. One needs a majority. ↩</a> </li> Schwartz and Wilf, The War of Return, 761. ↩</a> </li> Id. at 2448. ↩</a> </li> Id. at 2961. ↩</a> </li> It is left unaddressed how fledgling Israel would have handled a much larger Arab minority. In the modern day, Israeli Arabs make up around 20% of the population. ↩</a> </li> Schwartz and Wilf, The War of Return, 387. ↩</a> </li> Id. at 436. ↩</a> </li> Id. at 907. ↩</a> </li> Id. at 416. ↩</a> </li> Id. at 496. ↩</a> </li> Id. at 1398. ↩</a> </li> Id. at 2961. ↩</a> </li> Id. at 1631. ↩</a> </li> Id. at 1329. ↩</a> </li> Id. at 1474. ↩</a> </li> Id. at 1953. ↩</a> </li> Id. at 2248. ↩</a> </li> Id. at 3559. ↩</a> </li> Id. at 3785. ↩</a> </li> Id. at 3516. ↩</a> </li> Id. at 2065. ↩</a> </li> Id. at 2496. ↩</a> </li> I would like to note, at this point, that there is a lot more to say about UNRWA. In fact, there are probably several books worth of things to be said about UNRWA. If you want to read one such book, you should definitely read War of Return! There are comparisons of budget details and staffing numbers between UNRWA and UNHCR, analysis of the success of other major UN revitalization agencies like UNKRA and why that didn’t happen with UNRWA, and more. But, for the purpose of this review, we have to move on. Apologies. ↩</a> </li> I love footnotes. I especially like when they are humorous, instead of just page citations, which I realize this review—much like War of Return itself—lacks. So here is one in compensation. ↩</a> </li> Including but not limited to archives from the UN, Israel, US, UK, and Al Jareeza’s Palestine Papers, interviews with high-ranking Israeli politicians and military figures, and a wealth of books, articles, reports, and position papers from throughout the conflict’s long history. ↩</a> </li> Schwartz and Wilf, The War of Return, 3246 ↩</a> </li> Ibid. ↩</a> </li> Id. at 1309. ↩</a> </li> One such adjustment, if Schwartz and Wilf are taking suggestions, would be to address the rise of the new Israeli right. Netanyahu’s current government contains a minister with a conviction for terrorism; their book was published before this latest example of extremism from the Jewish side, and I would hope an updated version would address this. ↩</a> </li> Schwartz and Wilf, The War of Return, 1006. ↩</a> </li> Id. at 3408. ↩</a> </li> </ol> </section> James Hoffmann x Bellwether Coffee 2023-08-18T00:00:00+00:00 Today, I’m going to talk about this video</a>. A short time ago, I found myself on the Stanford University campus. I grabbed a quick bite to eat at Tootsies</a>. Against the wall, I noticed a large black-and-white machine. It had an almost Apple-esque minimalism, with brass details accentuating its design. I was fascinated, and appreciatively took note of its Bellwether Coffee</a> insignia. I figured it likely originated with one of the area’s indigenous startup fauna. Not long after, I saw a video pop up in my YouTube subscriptions by one James Hoffmann</a>. I’m not terribly engaged in the world of coffee, but I do quite like James. Lo and behold, the video began and James Hoffmann was sitting atop a Bellwether Roaster! Glorious. I thoroughly enjoyed the video, drafted up a blog post in appreciation of the gleeful coincidence, and here I am half a year later publishing it. If you haven’t seen the video yet, I beg you give it a watch.</a> Here’s to serendipity. Spies of No Country 2023-07-28T00:00:00+00:00 Spies of No Country by Matti Friedman. Finished on July 28, 2023. Rating: ★★★★☆ The Thousand Crimes of Ming Tsu 2023-03-09T00:00:00+00:00 The Thousand Crimes of Ming Tsu by Tom Lin. Finished on March 09, 2023. Rating: ★★★★★ Movies I Saw Other People Watching on the Plane 2023-01-21T00:00:00+00:00 Everybody has to entertain themselves somehow on fifteen hour flights. Some people watch movies. I watch people watch movies and make a list of them to post on my blog. Bullet Train</summary> I like this movie, good on you whoever you are. </details> Don't Worry Darling</summary> Been on the periphery of my watchlist for a minute now. From my limited perspective watching without sound on a tiny plane screen made even smaller by the fact that it's someone else's sitting down the aisle, it has compelling visuals. Going to move it up on my list. </details> Mud, Sweat, and Beards</summary> Uhhh... </details> Lightyear</summary> I want to give this a try, but I'm fairly certain that I'm going to stop before I finish it because it'll feel like a waste of my life. </details> Stanley Tucci Searching for Italy </summary> This might actually be titled Umbria. Seems rather pleasant in any case. </details> Where The Crawdads Sing</summary> I feel like I've heard of this but I have no idea what it actually is. </details> The Movie Where Idris Elba is a Sex Genie for the Older White Woman </summary> This movie is called Three Thousand Years of Longing. Tip of my tongue, I swear! </details> The Movie Where Liam Neeson (I Think?) Makes the Classic Meme Phone Threat </summary> IGN says this is called Taken. </details> The <details></code> element is really fun and I should use it more. Though it does lead to a scary amount of HTML</code> in my Markdown. There are four hours left on this return flight and I’ve expended all the blog post ideas that I came up with on the flight over. Going to have to find something else to do now. Naftali Bennett's Inspirations 2023-01-20T00:00:00+00:00 I recently had the opportunity to meet former Israeli prime minister Naftali Bennett</a>. This was very fun and rather unexpected! During my time with Bennett, he was asked about his “top three inspirations as a leader.” A bit template-y of a question, but I found his answer to be an interesting list—so I’ve recorded it here for posterity: David Ben-Gurion</li> Anwar Sadat</li> Lee Kuan Yew</li> </ol> Perhaps I should go into journalism. Neuromancer 2022-11-29T00:00:00+00:00 Neuromancer by William Gibson. Finished on November 29, 2022. Rating: ★★★★★ Who by Fire 2022-11-22T00:00:00+00:00 Who by Fire by Matti Friedman. Finished on November 22, 2022. Rating: ★★★★★ Moonwalking with Einstein 2022-11-20T00:00:00+00:00 Moonwalking with Einstein by Joshua Foer. Finished on November 20, 2022. Rating: ★★★☆☆ The Desert and the Sea 2022-10-15T00:00:00+00:00 The Desert and the Sea by Michael Scott Moore. Finished on October 15, 2022. Rating: ★★★★☆ Pomegranate Seeds 2022-09-29T00:00:00+00:00 Just before the start of Rosh Hashanah, I found myself standing over a kitchen sink. To my left, two pomegranates sat next to a large stainless steel knife atop a cutting board. Ahead of me, resting in the sink, was a bowl brimming with water. To harvest the seeds of a pomegranate, you first cut the fruit into quarters. These wedges are then submerged in water, where you basically fidget with the fruit until the segments detach. The red, semi-translucent seeds sink to the bottom of the bowl while the white-brown excess floats to the top to be conveniently discarded. On this particular afternoon, I was in a bit of a rush: I needed seeds quickly. I threw a couple wedges into the bowl, and started working on one of them. I finished the wedge and tossed the skin aside. Without really looking around, I reached out, unfocused, into the waters of the bowl. After a few seconds, I paused and stared down at the pomegranate in my hands. I hadn’t put any thought into the motion. I reached out my hand, and must have been subconsciously tracking the movement of the wedge. Completely on instinct, I had found a second wedge. The experience was so seamless it left me mildly startled, observing it after the fact. Some keyboard-based programs give the feeling of interfacing directly with technology, a near echo of the instinctive feeling of working with your hands. VR and AR aren’t there yet, but the applications are fairly self-evident. Startlingly seamless is a good experience to aim for as developers. Mabel 2022-09-29T00:00:00+00:00 Mabel</a> is a fancy BitTorrent client for the terminal. Programmed in performant and memory-safe Go</a>, the TUI is the torrenting state-of-the-art on macOS and Linux (and anywhere UNIX-y). Keyboard-centric, terminal-based workflows provide developers with a sense of focus and control unparalleled by the GUI. In my personal use of such programs (like NeoVim</a>, Amfora</a>, and many others), I noticed a distinct lack of functional torrent clients: existing options were dated, spewed logs to stdout</code>, and proved generally disappointing. Mabel modernizes torrenting from the command line. Mabel builds on the excellent work of many others in the open-source community, like Charm</a> and anacrolix</a>. It’s the second public release of SMMR Software</a>, a collaboration between FIGBERT</a> and Jacob Neplokh</a>. Mabel is available through a package manager near you. For more, explore the source code on GitHub</a>. Aggregation Theory, Virtuous Cycles, and Nitter 2022-06-27T00:00:00+00:00 In 2015, Ben Thompson first proposed Aggregation Theory</a> in an article by the same name. He argues, in short, that the internet “has fundamentally changed the plane of competition,” in that in our era, “the most important factor determining success is the user experience.” Why is it then, that YouTube and Reddit</a> and Twitter</a> all… suck? The answer, I believe, lies just a sentence later in the article: … the best distributors/aggregators/market-makers win by providing the best experience, which earns them the most consumers/users, which attracts the most suppliers, which enhances the user experience in a virtuous cycle. </blockquote> When YouTube first launched in 2005, it provided users with a vastly better experience than anything else available at the time. No longer did you have to download videos via BitTorrent</code> or ftp</code> (or however else one procured elephant videos</a> back in the day): now you could stream them on the web. 17 years have passed since then, and the world is a much different place. YouTube is the king of video on the internet. And yet, your browser’s default <video></code> tag provides a way better experience than the slow and clunky YouTube player. The story is much the same for similar giants. Reddit was founded alongside YouTube in 2005, Twitter a year later. Both websites are slow, practically unusable on mobile, and heavily limited for those without an account. One would think that if success was truly predicated on a better user experience, such glaring flaws would lead to other platforms “winning” by simply providing a better UX. We can see glimpses of what this might look like with alternative front-ends, like teddit</a> and Nitter</a>, which wrap around the existing platforms while massively cutting down on bloat – providing an objectively better UX. Nevertheless, YouTube and friends remain on top. The pivotal realization here lies in the second half of Thompson’s paragraph quoted above: the idea that having a better UX grows the initial userbase of an aggregator, and that having those users enhances the experience for future users (who further improve the experience, attracting more users, and so on and so on). YouTube and Reddit and Twitter had a better UX than their competitors in the early 2000s, and attracted millions of users in the ensuing decades. Now the moat is too big for alternative platforms to overcome – no new startups have yet built a user experience more attractive than YouTube’s 2.5 billion users. d3c3nt 2022-02-19T00:00:00+00:00 d3c3nt</a> is a simple, clean, and flexible theme for Zola</a>. The theme evolved from use on my site, made with vanilla CSS and HTML – and completely sans Javascript. All in all, it’s fairly… decent. d3c3nt styles common HTML elements like headers, paragraphs, and lists; plus semantic elements like blockquotes, tables, abbreviations, and more</a>. It also provides custom shortcodes</a> to incorporate modern web elements in your markdown. To see the theme in action, poke around the demo instance</a> – and explore the source code on my forge</a>. Privacy Redirect for Safari 2021-08-14T00:00:00+00:00 Privacy Redirect for Safari</a> is a web extension that automatically redirects sites to alternative front-ends like Nitter</a> and Teddit</a>. These front-ends provide users with ways to access the content of popular platforms, without exploiting your private information. Until now, extensions that could redirect to these front-ends were limited to users of Chrome and Firefox. Privacy Redirect for Safari expands this functionality to Apple’s native browser. Unlike other browsers, for many years Safari did not encourage the development of extensions. However, recent updates from Apple have signaled a change in direction, and useful extensions for Safari can finally be developed. Privacy Redirect for Safari is one of the first of this new wave of extensions. Privacy Redirect for Safari is the first step in creating a wealth of high-quality extensions for Safari. It’s built with the latest and greatest in Apple technologies, like SwiftUI and Safari Web Extensions</a>. Privacy Redirect for Safari is a fork of Simon Brazell’s Chrome and Firefox extension</a>, porting the extension’s functionality to Safari. It solves one of the most frustrating problems with front-ends like Nitter</a> – the need to manually edit links – by having the browser edit them for you. Privacy Redirect for Safari is the first public release of SMMR Software</a>, a collaboration between FIGBERT</a> and Jacob Neplokh</a>. It’s currently available for Mac, iPhone, and iPad. Download the app on the App Store</a>, and explore the source code on GitHub</a>. SimplyTranslate 2021-07-02T00:00:00+00:00 SimplyTranslate</a> is a collection of “translation scrapers with interfaces for gemini and the web,” made by metalune</a>. In other words, it’s a privacy-respecting front end for Google Translate, similar to Twitter’s Nitter</a>. Over the course of a couple days – and one synchronous hacking session – I containerized the program to make it easier to deploy on servers worldwide. When I first found the program, I immediately wanted to add it to my collection of self-hosted front ends. I had only one problem: I run my services with Docker, and SimplyTranslate had no Dockerfile. I tried slapping something together quickly, but curling from the container only threw an error: curl: (52) Empty reply from server</code></pre> I hopped onto the #simple-web</code> IRC channel on Libera.Chat</a> and started chatting with the devs. A couple days of back-and-forth later, we sat down to debug together. After scouring the internet, we determined that the application was binding to the wrong interface: 127.0.0.1</code>, which is only accessible from the local machine, instead of 0.0.0.0</code>, which is public. I cooked up a pair of working Dockerfiles – the first running the app directly and the second through uvicorn</a>. Metalune made a repo for the Dockerfiles</a>, and I sent in a patch via email</a>. It was committed</a> within the hour. This contribution was particularly interesting! We collaborated over IRC and across timezones, and I ended up returning to git’s roots by submitting the patch via git send-email</code>. Compared to the GitHub workflow, this feels very natural! You can use SimplyTranslate over on metalune’s hosted instance</a>, and browse the source code on Codeberg</a>. An Update on Bad Backups and Internet Celebrity 2021-06-27T00:00:00+00:00 My recent post on the Great Alpine Migration Tragedy</a> of 2021 garnered some serious attention! I got a number of emails, a whole host of feedback on Lobste.rs</a>, and over 100 points on Hacker News</a>! As I understand it, that means I’m now a top internet celebrity (and eligible for the 10 KB Club</a>). Still, I must remain humble even as I catapult toward digital infamy and thus I am here to respond to feedback, answer questions, and talk about what I’ve done to prevent something like this from happening again. That’s a good point: I’ve been advised to update my methods. Several folks pointed me towards strace</code></a>. Sending the SIGUSR1</code> or SIGINFO</code> signal to Tarsnap would also work. The -v</code> flag could have been helpful as well. A number of folks were confused as to why I didn’t keep the old machine</a> around during the transition. Others detailed their own migration strategies</a>. In general, it seems that folks agree you should keep around your old machine until you’re totally certain everything is up and running in your new setup. As I mention briefly in the article, I’ve actually done this in the past myself! This time, however, there was no second server. Everything was done in-place. I cannot change this, as it occurred in the past. In the future I will avoid this kind of migration. A number of people took issue with my take-aways from the experience. See: Trying to address a lack of confidence by increasing the backup frequency doesn’t make sense. The backup frequency is the most trivial thing to adjust and doesn’t address deeper issues, like the fact that you need to dump/restore databases properly and shouldn’t copy files from a live database. by klodolph</a> </blockquote> I’ll start by saying that that increasing backup frequency from zero to anything greater than zero does, in fact, address a significant issue. I now have backups where there were none before. Still, the point stands. Without quality backups, and quality restore strategies, the backups are not as useful as they could be. There was some fun memeing in the Hacker News comments in response to me being a teenager. Specifically, wildmanx said that they wouldn’t hire me for their “company IT or devops or whatnot.”</a>. Several people have a sense of humour: Good idea to let him graduate from highschool before hiring him! by nick__m</a> </blockquote> I’ll add for further memery that in addition to the stress of my entire server infrastructure burning, I had to deal with passing the behind-the-wheel driver’s test. I’ve now been a licensed driver for about a week and a half. Coincidentally, I got my server back running the same day I passed the test. I moved all the data in my self-hosting setup into a single directory. The files were previously all lying about my home directory, which is what led to the tragic “forgetting about .env</code> amongst all my dotfiles” incident. I set up a cronjob that runs the backup script acts</code></a> daily, set to back up the self hosting directory. It currently keeps a buffer of 31 daily and 12 monthly backups, and keeps around yearly backups forever. It’s been running for a couple days now, and it’s all been smooth sailing so far. I’d also like to say thanks to a number of people who were especially kind, and made this a fun experience: Graham Percival, an employee at Tarsnap Backup Inc., reached out to me directly – we had a fun and interesting email exchange.</li> hannu</a> was the first to write the later much-echoed sentiment that reflecting on one’s mistakes is a useful and laudable thing to do (see also abraae</a> and argomo</a>).</li> </ol> I’m going to continue iterating my self-hosting setup. I’ve still got a couple new services</a> I might want to throw up there. I also can’t forget about that dedicated page</a> I promised… The Wrong Way to Switch Operating Systems on Your Server 2021-06-17T00:00:00+00:00 After moving my server to Hetzner</a>, I built up a large collection of self-hosted services I use on a daily basis: from fun things like an RSS reader</a> and an IRC bouncer</a>, to critical services like my email</a>. I ran them all with docker-compose</code> from a Debian</a> VPS. For the last couple months, however, I’ve been meaning to move away from Debian and towards something more minimal and clean. Over this last weekend, I decided to move to Alpine Linux</a>. The transition was supposed to be quick and dirty: Shut down all the services running on my VPS</li> Make a backup of relevant files with Tarsnap</a></li> Mount Alpine Virtual ISO image and setup the OS</li> Restore files from Tarsnap backup</li> Bring everything back up</li> </ol> In a previous move between two servers, I simply rsync</code>ed the relevant files over to the new VPS. Here, where I’m just switching operating systems on a single server, I figured I could make a backup with Tarsnap, and be done within the day. However, backups are much more complex than simply transferring files from one server to another. My haphazard strategy resulted in three days of stress and frustration as I clambered to restore a self-hosting empire that I myself had reduced to ash. I began my work on the transition full of optimism, if a bit stressed. I had read through the Tarsnap online documentation a number of times, and was ready to make my first attempt. I loaded my Tarsnap account up with USD$10 and ran: $ sudo tarsnap -c -f backup-name docker-compose.yml ...</code></pre> My terminal sat empty for hours. There were no changes – the process was running, but there was no feedback. I was nervous. What if it failed silently? How can I check? What should I do? </blockquote> I pressed <Ctrl-C></code>. To my horror, stats printed to the screen: the backup had been 90% complete, and I had stopped it. Convinced I had ruined the backup completely, I deleted the partial backup from Tarsnap and started again from scratch. This was my first, but not last, moment close to tears. I went to sleep and let the backup run overnight. Day Two began well: I woke and the backup was finished! I wiped the VPS, installed Alpine, and brought it up to spec. I created a regular user, configured SSH, and decided to use doas</code> instead of sudo</code> for a change. Alpine, so far, feels great to use. None of the cruft that bothered me when using Debian. With Alpine set up, I started to restore the backup: $ doas tarsnap -x -f backup-name</code></pre> Once again, after running all day it had not finished. I opened up a new tmux</code> window and poked around the filesystem. All my files seemed like they were already there… What if it failed silently? How can I check? What should I do? </blockquote> I pressed <Ctrl-C></code>, cutting off the download, and tried to bring everything back online: $ doas docker-compose up -d</code></pre> It errored out. All my environment variables were undefined. Then it hit me: I forgot to back up the .env</code> file. My eyes welled up. Still, I was determined. I worked to reconstruct the .env</code> file from secrets I had stored in Bitwarden</a> (my offline copy, because my vault is self-hosted and was thus down). I ran it again: $ doas docker-compose up -d</code></pre> One of my services was missing a Dockerfile to build. I shouldn’t have pressed <Ctrl-C></code>! I was a total moron. I put on a sad song</a>. I was close to tears once again. I gathered what was left of my resolve and trudged onwards. I searched tarsnap</code>’s manpages looking for something to speed up my download. I found a number of flags that could have helped me make a backup better the next time around, but nothing that would help me restore the backup any faster. With nothing in the manpages, I went to look at the helper scripts</a>. That’s when I found it: redsnapper</a>. A Ruby script that runs multiple tarsnap clients at once to extract archives fast. Fucking precisely. I wiped out the incomplete files I had restored, downloaded Ruby and started restoring from the backup once again: $ doas redsnapper backup-name</code></pre> I changed the song</a>, and watched the files fly by on my screen. I went to sleep, confident I would wake to good news. The download had failed trying to download a large .mkv</code> file. I restarted redsnapper</code>, explicitly excluding the .mkv</code> it had failed to download, and let it run until it came on another movie and crashed again (an hour or so later). I excluded the second movie file and sent it to run again. This was a long, boring process. It sucked. Then I realized something. redsnapper</code> kept crashing when it hit movies I had stored in Jellyfin</a>. I don’t need Jellyfin at all. I’ve never watched a movie more than once. The movies take up massive storage on disk, and keep causing tarsnap to crash. They don’t compress well either, so they take up a fuckton of space in the archives. I can always download the movies again if I want to give them another go. Why the fuck am I forcing myself to deal with this shit? </blockquote> I stopped the download in the middle - the day’s third, after two earlier attempts that ended after encountering movie files – and changed the command slightly before rerunning. After a number of errors I couldn’t explain, I realized my account was negative and topped it up with another USD$25 before running: $ doas redsnapper backup-name -- --exclude='*/jellyfin/*'</code></pre> I returned to my computer a couple hours later. redsnapper</code> had stopped, with a whole lot of files extracted and a couple errors at the bottom about symlinks. I figured, this time, it had probably done everything properly but couldn’t create the symlinks (probably a flag missing somewhere). I manually went through my files creating the symlinks, and then brought everything up with docker-compose. I checked the containers. All up. I checked the logs – no immediate errors visible. I opened figbert.com</a> on my laptop. It appeared. Service was restored. Hallelujah. I made a lot of them. Here are a few: After shutting down my containers, I backed up my entire setup. This included a number of “live” databases</a>, .git</code> folders, and other data that I either did not need or could reconstruct once the move had been completed.</li> I didn’t back up the .env</code> file I use to store secrets for use in docker-compose.yml</code>. I was luckily able to reconstruct it from individual secrets I stored in my password manager.</li> A thorough read of the manpages before I started (rather than just the online guides) would have revealed several helpful flags: -v</code> to see what files tarsnap</code> is operating on, --aggressive-networking</code> to take advantage of the datacenter internet speeds, and --recover</code> to resume interrupted backups, to name a few.</li> We already talked about Jellyfin. Even with very little content in Jellyfin, the collection took up huge amounts of space on disk and in the backup (especially because video files don’t compress well), and sat entirely unused. It is now gone. Good riddance.</li> </ol> What did I learn? Well, I’m still devising a plan to prevent things like this from happening in the future. Here’s the plan currently: Back up everything every day. I’ll build a buffer of three “rolling” backups, where backups collect up to a max of three and then, as new backups are created, the older backups are removed. The backup script will shut down the services, dump the databases (i.e. convert as much content to plain-text, easily-compressible formats as possible) and make a time-stamped backup (currently only with Tarsnap, but perhaps in the future with a number of other services). Simply having high-quality backups to restore will already be a huge leap forward. I’m also definitely going to continue using redsnapper</a>: the speed gains it gives on large backups are crucial. I really should read all the documentation before I try something new. I’ll write further about my self-hosting setup as it evolves, and publish the backup script once its finished. I’ll also maintain a dedicated page on my site describing my self-hosting setup as it changes. Also, I’m sure there are people more knowledgeable about Tarsnap than I. That’s basically the point of this article. If you are one of these people, please don’t hesitate to email me</a> if you’ve got corrections, advice, or just want to flex that you know how to do backups better than I do. Tofu 2021-04-07T00:00:00+00:00 Tofu</a> is an easy-to-use, open-source, two-factor authentication app designed specifically for iOS, made by Calle Luks</a>. I adopted it after migrating from Authy, and ended up overhauling its icon system. It’s a really well designed app: When I migrated to Tofu, it had a limited number of hand-designed icons in one big Sketch file</a>. Each icon was created and exported manually. I set out to simplify</a> the process, and Calle proposed</a> we create a shell script to generate icons on-demand from a folder of pngs. I got my shell hat on and a week later, my pull request</a> was commited to upstream with a brand new bash script. The script</a> iterates over a given directory using sed</code> and regex</a> to grab the names of each png inside and add them to Xcode’s assets folder. It uses a heredoc</a> as a template for the new imageset’s JSON, and the sips</code> cli to generate properly sized icons for use in the app. Contributing to open source projects is incredibly fun. If you’ve got a free weekend, take some time to hack on your favorite codebase. Big thanks to Calle</a> for creating such an awesome app – I look forward to collaborating more in the future. Download the app on the App Store</a>, and explore the source code on GitHub</a>. A Package in the Bush 2021-03-19T00:00:00+00:00 A little more than a month ago, I noticed that my computer charger had slightly yellowed at the ends. Though I’m unsure how, it seemed like it had sustained some sort of heat damage. I went over to Wirecutter</a> to pick one out, ordered it off Amazon, and… it never came. Checking the app, the package was marked as delivered. Checking my front door said differently. I got a refund and ordered another, which I’ve been using ever since (pretty good, only downside is that it’s only 1m long, but it makes up for that in speed). A couple days later – or maybe weeks, I’m not sure – I noticed an Amazon package tossed midway into a bush by my front door. It wasn’t the full box kind, made of cardboard. More like an envelope, I suppose. Huh, that’s neat. There’s an Amazon package in the bush. Wonder what it’s doing there. </blockquote> After several weeks of thinking this every time I exited my front door (a fairly rare occurance, these days), I opened it. Wet from the rain, I figured it was time to bring it inside. It would appear as if the person who received my first charger order delivered it to me, but couldn’t quite make it to the door. Now I’ve got two chargers, for the price of one. One from Amazon. One from the bush. Some quality shitposting 2021-02-10T00:00:00+00:00 I was doing my daily HN and Lobster trawling, when I stumbled on a post</a> by boringcactus</a> which sounded somewhat familiar. About half way through I realized it was a truly epic shitpost and genuinely burst out laughing. The story begins on February 2, 2021, on Drew Devault’s blog, with a post entitled I’m tired of this anti-Wayland horseshit</a>. It’s a good read – Drew’s a smart guy (he makes Sourcehut</a>). In it, he disparages detractors of Wayland and says that much of the people who criticize it haven’t given it a close enough look and should think about what they’re saying before they go shitting on maintainers of an all-together pretty dope project. Seven days later he wrote Rust: “Move fast and break things” as a moral imperative</a>, in which he criticizes the Rust Foundation’s attitude toward stability and platform support. A fatal mistake. In one fell swoop, and one s/Wayland/Rust</code>, boringcactus ended this man’s career. Check it out: I’m tired of this anti-Rust horseshit</a> Quite the reMarkable Device 2021-01-22T00:00:00+00:00 Lately, there’s been renewed interest in clean, simple technology built to help us focus. Protocols like Gemini</a> strip away the chaos of the web. Hardware hackers fit screens in mirrors</a> and build beautiful minimalist displays to read the news</a>, display data neatly in a picture frame</a>, or provide a daily summary</a>. Hidden amongst these many awesome projects is the reMarkable 2</a>. I’ve been using the reMarkable for the past month or so and it is a seriously solid device. I use it mainly to take notes, at which it excels. The e-ink display is the best I have ever seen, barely ever refreshing the entire screen and almost entirely without the temporary artifacts that plague similar devices. The pencil is comfortable to use – slightly fuzzy – and, interestingly, entirely passive: it never needs to charge. Given that it’s meant to replace your paper, it shouldn’t be too surprising that the tablet’s also incredibly thin. In one of my classes, I’m required to take Cornell notes. On what may be a related note, it’s a terrible class. Regardless, I need to get an A and so notes I shall take. The reMarkable comes with a built-in Cornell notes template. It’s slightly different than the format I’m used to, but it fits the bill well enough. It’s hard to describe the experience of using the device. I would say that it feels shockingly natural. Using it to take notes feels like writing on a clipboard with none of the usual annoyances – the paper sliding, having finite pages; and all the benefits of a digital device – sending files via email, OCR, a select-and-drag tool, etc. It’s convenient, feels nice, and performs well. Most importantly, it makes me excited to use it every time I turn it on. In another class, the instructor distributes worksheets every once in a while to complete during class. Now that we’re in distance learning due to COVID, these are pdf</code>s – perfect for use with the reMarkable. I download them onto my laptop, upload them through the app, and voila. Look Mom, no scratch paper! I can write directly on the worksheet. I’ve only started doing it this week, and it’s amazing. Sure, there are programs on the computer that allow you to write on a pdf</code>, but doing math with a trackpad sounds like torture. With the reMarkable, it’s enjoyable. That’s the device’s biggest impact. I used to hate writing by hand. I would beg my teachers to let me type assignments so I didn’t have to use a pencil – what am I, a caveman? Now, my paper has superpowers. It’s also a suprisingly good reading device, with native support for epub</code>s and pdf</code>s. It doesn’t have a backlight, but to be fair neither do actual books so I’m not too bothered. The default font size is quite large, and the UI is really minimal which makes for peaceful, undistracted reading. I own two Kindles already, but I’ve taken to using the reMarkable instead for a number of reasons: There’s no ads or tracking. On Kindle there’s an ad on the bottom of the homescreen, which expands to the whole screen when you turn it off – plus it sends every move you make to live forever with Big Papa Bezos. Not so with the reMarkable.</li> The screen on the reMarkable is physically larger. Though sometimes the compact size of my Kindle comes in handy, like when traveling, having a nice big display is definitely an advantage. My Kindle is closer to a large phone, and the reMarkable is definitely a tablet.</li> The reMarkable is a fairly open device. The Kindle, on the other hand, is locked down and dripping with DRM (fairly easy to break but still an encumberance).</li> </ol> The developers have also made a browser extension for Chromium-based browsers called Read on reMarkable</a>, which I would love to see the insides of but is unfortunately closed-source. Basically, it takes any webpage, turns it into an epub</code>, and sends it off to your device. I love it and use it near constantly: whenever I run into anything on HN that is either really long, or I just want to save for later, I hit a button and it sends it to the tablet. Again, just incredibly convenient. EDIT: Dario Vladović</a> has pointed out to me that extensions are unpackable, and thus though it’s not open source it is, by nature, source available. This section is titled “jailbreak,” which is actually a bit of misnomer because the reMarkable runs Linux and you can ssh</code> into it with ease. It’s also not too fitting because I’m not just going to talk about modifying the device, but also about official accessories. My device came with a Book Folio</a> and Marker</a> in the box, which is pretty good value. I did find out while writing this that they offer a “Marker Plus” that comes with a built-in eraser. It is, unfortunately, out of stock currently but I plan on buying it once it returns in “January 2021,” so supposedly sometime in the next week or so. Ok now time for the jailbreak-y stuff. There’s an active community on Freenode, an unofficial wiki</a>, and an Awesome list</a>. All of these are really great resources for cool stuff you can do with your device, and I’m planning to begin experimenting with them in the coming weeks. My first goal is to get KOReader</a> running so I can sync my extensive Calibre library to the device. I’m going to avoid adding any games</a> so as not to add additional complexity to a device that aims to remove distractions, and proceed slowly so I don’t brick my fancy new toy. Stick around to see how that goes. The recently published fairly viral Ditherpunk article</a> by Surma</a> gives a really interesting overview of image dithering. After reading the article, I noticed that the reMarkable itself uses dithering to “fade off” your writing and make it look like pencil. Neat. Also, comparing the reMarkable to the Kindle makes me want to jailbreak that too – I’ll find some time to look into that soon. EDIT: I ended up jailbreaking one of my Kindles, which I now use daily as my primary book-reading device (as opposed to the reMarkable, which I think of as an article-reading device). I should write some stuff about that. My First RegEx 2021-01-21T00:00:00+00:00 (?<=\.\/IssuerIcons\/).*(?=\.png)</code> Let’s break it apart piece by piece. It’s meant to return the filename from a given path in a specific format. When given ./IssuerIcons/Example.png</code>, the script should return Example</code>. The first section is a positive lookbehind: (?<=\.\/IssuerIcons\/)</code>. This asserts that the selected portion must by preceeded by ./IssuerIcons/</code>. The section section is a pattern match: .*</code>. This matches all characters, excluding newlines, of any length. The third and final section is a positive lookforward: (?=\.png)</code>. This asserts that the selected portion must be followed by .png</code>. Yesterday, I wrote my first regular expression. I’ve always regarded regex as an arcane art of true shell wizards – and for the most part, I still do. Now though, I’ve gotten a glimpse of their world. I wrote this expression for Tofu</a>, an open-source 2FA app for iOS. I had been thinking of making more “issuer icons” for the app when I noticed that the icons were all designed in one Sketch file. This could become a problem if two people were to edit the file at one time, so I suggested a solution</a>. The author got back to me with another idea: ditch Sketch altogether, and replace it with a shell script that generates icons from a directory of png</code>s. I wrote it</a>. I’ve since refined it, with help from ThinkChaos</a>, to make it independent of any preceding directories using sed</code>: sed -E 's:.+/(.+)\.png:\1:'</code>. It was a really fun experience – I’ve never written bash before! It reminded me a lot of Drew DeVault’s post Become shell literate</a>. I work primarily from my terminal, and am slowly working my way up to true mastery – I’m nowhere near close, but this was a step in the right direction. SASS and Light Mode 2021-01-02T00:00:00+00:00 I’m on a roll! A second blog post in less than a month! Crazy. Anyway, as you may have noticed – depending on whether or not you read this via RSS or on the main site – I changed the site styles. I got rid of dark mode, added styles for code and keyboard blocks, and changed the look of links to a new cool design. Update: I brought back dark mode. Light mode gave me anxiety. It all started the other day when I was cruising through cyberspace and noticed a rather unique looking link style. Instead of the traditional underline styling</a>, hyperlinks were surrounded in square brackets (like Markdown). I thought it was a really cool change, and made the site stand out – so I stole the idea and implemented it here! Yay for stealing. More on this in a future post…</del> However, when I went to add the new style to my site I noticed that my global.css</code> file had gotten rather lengthy and disorganized. Given that Zola</a> comes with built-in support</a> for Sass</a>, I thought I might try it out. I found it easy enough to pick up given that I already know CSS. I decided to use Sass’ indented syntax</a>, which required more work to convert from my existing CSS but – in my humble opinion – looks much cleaner. Given that I was already tinkering with site styles, I decided to implement a couple other changes. I abandoned the site’s dark mode, as though I myself am an avid dark mode user, I couldn’t find a way to implement it in a satisfying way. If I figure out a Sass-y way to do so, I may add it back. I added a border around inline code</code> blocks because I thought I already had done that</del>. Lastly, I added StackExchange’s amazing kbd</kbd> styling so I could write ⌥</kbd> + ⌘</kbd> + N</kbd> in this article</a>. I’m pretty happy with the new styles, but I have no doubt I’ll continue to tweak it as I find cool designs around the web. Thanks for dropping by: see you next time! FIGBERT PS: It’s interesting that Sass supports both of the two modern syntax paradigms, both the indents and linebreaks camp and the brackets and semicolons camp. I don’t really prefer one over the other, but it’s an interesting divide. How to Mirror Your iDevice to your Mac 2020-12-31T00:00:00+00:00 I recently found myself debugging a mobile game I’m working on (sneak peak</a>) with a friend over Jitsi</a>. I had no trouble sharing my Godot</a> window, Xcode console, or IRC bouncer</a>. But how were they going to view the output on my phone? Well turns out it’s really easy, and requires no non-default software: Open Quicktime</li> Press File</kbd> -> New Movie Recording</kbd> or ⌥</kbd> + ⌘</kbd> + N</kbd></li> Plug in your iDevice (I’ve only tested with an iPhone) via USB</li> Click the down caret by the red record button</li> Select your device as the Camera</li> </ol> That’s it! I didn’t even need to record anything, as I just used Jitsi’s built-in Window Sharing feature to share the screen live. If you want to record your phone with something more powerful than Quicktime, you can use OBS</a> and select your phone as a Video Capture Device. Hope you’ve enjoyed this post! I know it’s a bit shorter than normal, but I’m hoping to post more frequently starting soon as a bunch of interesting</a> things</a> are happening in and abouts my life. A Very Convincing Review of Terraria 2020-11-16T00:00:00+00:00 If I wasn’t certain I wanted this game before, I am now. Moving To Hetzner Cloud from DigitalOcean 2020-11-01T00:00:00+00:00 Since I began working on self-hosting, I’ve been using a DigitalOcean</a> (affiliate link) VPS running Debian 10</a>. However, after investigating the alternatives in the space, I’ll be moving to Hetzner Cloud</a> when my DigitalOcean credits run out in six months. Here’s why. I went with DigitalOcean as my hosting platform primarily because of the USD$50 credit included with Github Student</a>. I’ve used USD$20 of that credit as of today, and the remaining money should run out around April 2020, giving me almost a year of VPS for free. Not a bad deal. With that said, as my credits hit the half-way point, I’ve begun to think about what’s to come. Namely, the fact that my current server is running at a constant 80% memory usage – severely limiting what I can self-host. I currently use the cheapest DigitalOcean VPS available. I have 1 vCPU, 1GB of RAM, 25GB of SSD, and 1TB of data transfer. All that power is packed in a clean USD$5 per month. This is, as far as cloud hosts go, pretty good bang for your buck. Hetzner blows that out of the water. Take a look at this table: USD/Month</th> vCPU</th> Memory</th> Storage</th> Traffic</th> Provider</th></tr></thead> $3.41</td> 1</td> 2GB</td> 20GB</td> 20TB</td> Hetzner</td></tr> $4.77</td> 2</td> 2GB</td> 40GB</td> 20TB</td> Hetzner</td></tr> $5.00</td> 1</td> 1GB</td> 25GB</td> 1TB</td> DigitalOcean</td></tr> $6.70</td> 2</td> 4GB</td> 40GB</td> 20TB</td> Hetzner</td></tr> $10.00</td> 1</td> 2GB</td> 25GB</td> 2TB</td> DigitalOcean</td></tr> </tbody></table> Check out the math on that, seriously. Hetzner’s base tier offers 1 vCPU, 2GB of ram, 20GB of SSD, and 20TB of data transfer for USD$3.41 per month. That’s double the ram and 20 times the transfer for almost half the price of my current DigitalOcean droplet. The fourth down in the table above, the one I plan on purchasing, costs USD$1.70 more than DigitalOcean and has quadruple the performace of my current VPS. That’s more than even Docker could need. Ok that might be a slight exaggeration, but it’s still a lot. Hetzner is a green hosting provider</a>, as certified by the Green Web Foundation</a>. DigitalOcean is not. Obviously, climate change is important. If we look at it objectively, it’s one of the only extinction-level threats to human existence at the moment. All that is to say that switching to a green hosting provider is the only thing I need to do</a> according to this online website carbon calculator</a>, and you know I can’t resist online website metrics. Anyways, that’s a pretty good summary of my thoughts on the move. I’m probably going to use the move to make a few other changes to my hosting setup (i.e. docker-compose</code> to k3s</code></a>, umami</code></a> to plausible</code></a>). I’ll keep you all abreast of the latest updates to my hosting in future posts. Until then, this is FIGBERT signing off. I should make a more consistent outro. MacOS Control Center is Better 2020-10-07T00:00:00+00:00 I’ve been using the macOS Big Sur Beta for a couple of months now. So far it’s pretty fantastic, and a surprisingly smooth experience over all. One brand new feature in this update is the introduction of the Control Center. A close relation of its iOS counterpart, they differ in one key way: Control Center on macOS actually turns off services. Your browser doesn't support the video tag, which I use in place of .gifs, and/or the video formats in use here – sorry! </video> Did you catch that? It’s pretty subtle. When you try and turn off Bluetooth or Wifi on iOS, they aren’t actually turned off: they’re just temporarily disabled. The interface still broadcasts, it just doesn’t accept new connections. It’s one of my biggest pet peeves, and I’m glad Apple got it right with macOS. Now if they could just port it to iOS… Going Full Static with Zola 2020-08-22T00:00:00+00:00 Those of you who read my last “I Wrote This” post</a> will know that I was having some trouble with my website. My site was coded using Sapper</a>, a Svelte</a>-based web-app framework I had been using for some time. I had chosen to use Sapper because it allowed me to stay as close to the web-metal as possible, while still letting me do some fancy things like use components, scoped CSS, and server routes</a>. However, after diving deeper into website tests and statistics, I started noticing that my “static” site had a lot more moving parts than I thought. The HTML was crammed full of inline scripts and blob://</code>s, tanking performance, wreaking havoc on my CSP</a>, and breaking the site for people with scripts disabled. I decided to move the site to Zola</a>, a ludicrously simple static site generator made in Rust. Feel free to check out the source code here</a>. I decided I was going to rewrite my site because I have a bad habit of rewriting everything all the time</del> largely because of Sapper’s underwhelming response to this Github issue</a>, which proposes a “strict export” for Sapper sites to remove inline scripts and use of eval()</code>. I think this is a great idea, but it unfortunately has not received much attention (though it appears that as I’m writing this, it has been added to a “Roadmap Triage” project board). I started a new branch and began working to translate my site to Sapper’s main competitor, Routify</a>. Sapper and Routify are not the same thing, but for me they both would serve well enough. After around two days, I had a working MVP of my site in Routify. EDIT: Sapper is being retired in favor of SvelteKit</a>, which has the same issue</a>. smh. Then disaster struck: I got a bunch of emails from Github. A series of high priority security vulnerabilities had been found in dependencies used by basically all of my web projects. I spent a day force-updating all the dependencies of my web projects – a bit of a pain because npm refuses to natively upgrade breaking changes – and decided to stay as far away from the Javascript ecosystem as possible. I hate that when I install a JS framework, a fundamental tool of modern web development, I install a million other dependencies that could, and often do, have critical security vulnerabilities. I’m thankful that Dependabot caught these ones, but it really killed my enthusiasm for using any JS framework on my site – which means Routify was out of the picture. I decided to use a static site generator</a>. I’d heard of many of the big boys in the past, like Hugo</a>, Jekyll</a>, and Eleventy</a>, but they all had their own problems when I looked at them in the past. Hugo has god-awful templating syntax, Jekyll is Ruby-based and I don’t know Ruby, and Eleventy isn’t even an escape from Javascript! So I decided to use Zola</a>, a “one-stop static site engine.” Zola is made in Rust, so it’s super fast, and it’s designed to be dead simple. Seriously: the CLI has only five commands, everything is configured from one .toml</code> file, and your content is all written in “Augmented</a> Markdown</a>.” The interesting thing is that there’s honestly not much more to the story because of how easy and simple Zola is to use. All of my posts and projects go into the content</code> directory, my CSS, favicon, and miscellaneous files (non-content related stuff like emojis and public keys) go in the static</code> directory, and templates and shortcodes go into the templates</code> directory. If I was using a theme, it’s files would go into a theme</code> directory. My slow Python script to convert Markdown posts to Svelte (which was perfect at first but I then packed full excess tests and sandboxing) is gone. Zola handles that automatically.</li> I got rid of TailwindCSS, and replaced it with custom styles. It’s actually pretty fun to write simple custom CSS, especially with modern tools like variables.</li> Writing new posts is ludicrously easy now. I write a post in Markdown, throw any images or videos used in the same directory</a>, and publish.</li> Zola comes with a whole bunch of features built-in that I didn’t have before, like syntax highlighting and anchor links (the latter of which I have yet to set up). Other things are just handled automatically, like feed generation or i18n.</li> Build times are much faster. Exporting with Sapper wasn’t slow, but it didn’t feel instant. Zola does.</li> </ul> You sacrifice a certain amount of control by using a static site generator, like link</a> properties</a>. You could solve this with shortcodes</a>, or by contributing to the project (which I plan to do).</li> I mean that’s really it to be honest.</li> </ul> EDIT: You can now add attributes like noreferrer to links automatically with Zola… So I guess there’s no more drawbacks? Yeah that feels right. I’m really happy with using Zola, and I look forward to continuing to work with it in the future. I want to publish my blog’s styles and templates as a Zola theme</a>, but I have to iron out a few kinks (like anchor links, which are still a bit finicky on my end) before that. EDIT: I have since moved on from this theme on my own site, but you can still use the theme as d3c3nt</a>, a simple, clean, and flexible theme for personal sites. I also have yet to re-implement a bunch of the indie-web features and GoatCounter</a> analytics of my old site into this version. Overall though, I think it’s been a really fun and productive experiment using Zola, and I’d highly recommend using it for anybody looking for a great, no-nonsense static site generator. EDIT: I fixed anchor links, and I decided to forgo analytics altogether. My site’s better off without JS. Until next time, FIGBERT. Problem Exists Between Keyboard and Chair: How I Spent 2 Days Chasing a Bug that Didn't Exist 2020-07-28T00:00:00+00:00 Post-WWDC2020</a>, I decided to rewrite the backend of txtodo in SwiftUI using the new App and Scene structure</a>. Rebuilding the app from scratch may have not been the best choice, but during that process I have massively simplified the app’s data structure, despaghettified some messy UI code, and spent two full days trying to solve a problem that didn’t exist. This is the story of that last bit. The new app, so far, was mostly the same as the old version but without the AppDelegate.swift</code> or SceneDelegate.swift</code> files (using the new XCode 12 multiplatform app template). I also combined the Core Data FloatingTask</code> and DailyTask</code> entities into one Task</code> entity. By this point, everything was running well enough so I started to migrate more code into the new codebase starting with the fetch request: @FetchRequest( entity: Task.entity(), sortDescriptors: [ NSSortDescriptor(keyPath: \Task.completed, ascending: true), NSSortDescriptor(keyPath: \Task.priority, ascending: false), NSSortDescriptor(keyPath: \Task.name, ascending: true) ] ) var tasks: FetchedResults<Task></code></pre> Tasks are displayed as TaskView</code>s in a ForEach</code> loop on the homescreen, which is simple enough. The TaskView</code> struct, however, is relatively complicated. The purpose of TaskView</code> is to represent and manipulate a single Task</code>. In the previous version of the app (I’m going to call the original version 2.0 and the rewrite 3.0 from now on), this involved passing a number of attributes individually to be manipulated as the view’s @State</code>. When migrating the view, I reduced this to a single @ObservedObject</code>. I also removed some of the text styling, which I planned to port over after I got the UI functional. I ran the app on my device, and this happened: Your browser doesn't support the video tag and/or the video formats in use here – sorry! </video> Well that was unexpected. Instead of checking off the tasks I selected, tasks were checked off starting from the bottom and ascending – obviously not the intended behavior! My first thought was that it was caused by the use of @ObservedObject</code> to declare the view’s task property – I haven’t seen it used to manipulate a Core Data entity before, but it’s worked fine so far in txtodo – so I rewrote the variables to match version 2.0. // VERSION 3.0 struct TaskView: View { @Environment(\.managedObjectContext) var managedObjectContext @ObservedObject var task: Task @State var priority: Int @State private var config = TaskConfig() // UI... } // VERSION 2.0 struct floatingTaskView: View { @Environment(\.managedObjectContext) var managedObjectContext @ObservedObject var task: FloatingTask @State var completed: Bool @State var name: String @State var priority: Int @State var deleted: Bool = false @State private var editingText: Bool = false @State private var editingPriority: Bool = false @State private var viewingNotes: Bool = false @State private var confirmingDelete: Bool = false // UI... }</code></pre> Still no change. It was getting pretty late at this point, but I decided to stick it out for just a bit longer. I rewrote the TaskView</code> struct from scratch two more times to no avail. Something was wrong, but I had no idea where it was and there was no way I was going to figure it out at two in the morning by coding it again the exact same way. The next morning, I took a look at the code again. If the problem wasn’t in TaskView</code>, where was it? The only other thing in the UI was the button to make a new task, which looked something like this: Button(action: { let newTask = Task(context: self.managedObjectContext) newTask.name = "test" newTask.priority = 3 newTask.notes = [String]() as NSObject newTask.id = UUID() newTask.date = Date.init() newTask.daily = true do { try self.managedObjectContext.save() } catch { print(error.localizedDescription) } }) { Text("Add") }</code></pre> Some of you may have figured it out by this point. At the time, I was still confused – this was the exact method I was using in my previous app, but with preset values – how could it be broken? I modified the generation slightly so I could tell the difference between tasks, and hopefully get to the bottom of the issue: let newTask = Task(context: self.managedObjectContext) newTask.name = String(UUID().uuidString.prefix(Int.random(in: 5..<9))) newTask.priority = Int16.random(in: 1..<4) newTask.notes = [String]() as NSObject newTask.id = UUID() newTask.date = Date.init() newTask.daily = Bool.random()</code></pre> I ran the app again and saw this: Your browser doesn't support the video tag and/or the video formats in use here – sorry! </video> The tasks weren’t being marked off in ascending order. They were being moved to the bottom automatically when marked as complete, which I couldn’t see because a) all the tasks were identical and b) there were no animations to indicate that was happening. They were sorted by the FetchRequest</code> with a NSSortDescriptor</code>, to make sure that the unfinished tasks are the first thing the user sees. The “glitch” I had spent two days chasing down was entirely by design, and I had just forgotten. There were two main things I learned from this experience. First, it’s incredibly important to be able to take breaks. The difference between spending two days trying to fix a non-existent glitch and realizing it’s a feature you implemented could be as simple as a nap – it was for me. Secondly, your test and placeholder data is more significant than you might think: garbage in, garbage out</a> definitely applies here. If all your test data is the same, your tests are not good tests. To make the sorting more clear, I randomized the tasks’ priority, name, and category (as seen above) and added an animation with .animation(.easeIn(duration: 0.25))</code>. The current prototype looks something like this: Your browser doesn't support the video tag and/or the video formats in use here – sorry! </video> This has been a really fun blog post to write! A got a big laugh out of this bug chase, and I hope you’ve enjoyed reading it. Till next time, FIGBERT I Wrote This #3 2020-07-04T00:00:00+00:00 It appears I’m not that great at keeping a weekly schedule. Right after I published my last post, I started a class on text adventures and have been living and breathing in Inform7</a> ever since. I’ve also spent some time working on this site – though I’ve got even bigger changes coming in the future – learning a few new languages, and listening to a whole lot of music. I think something happened in cybersec</a> too? Oh also I learned how to drive. These past few weeks, my main work has been on interactive fiction as part of a summer class for my school. The first week was spent learning the history of text adventures (as I will refer to them for the rest of the article, becuase calling them “interactive fiction” makes me sound like a dick</del> a snob). I gathered a collection of games to play in my free time (if I ever have any more of that), from the very first of the genre – Adventure – to modern ones like Lost Pig. In the second week, we all made our own text adventures. It was a really interesting experience – drastically different than any other programming I’ve done. The main thing I struggled with was the semi-NLP style of Inform7</a>. Rather than telling the computer what to do, I felt like I was making suggestions. I also struggled with the editor a bit. Inform7 isn’t open source, which means there’s only one IDE you can use to program in the language. That would be fine if the IDE was high quality, but I often found it really bad and draining</del> mediocre. This was made a little better after I changed the editor colors to be much more vibrant so I wouldn’t fall asleep</del>, but I still felt like it needed some work. Regardless of any struggles along the way, the payoff after the game was finished was huge. Watching my little sister play a text adventure for the first time, one I had made, was awesome. Without further ado, I present to you: One Angry Wizard, or the Tentacular Adventures of our Brave Hero through a World Most Strange and Foreign. You can download the .gblorb here</a>. I made some changes that you guys can see, and some other ones behind the scenes. I first encountered webrings on Hacker News</a>, where I find most of my interesting internet things. From there, I decided to join both the geekring</a> and the Hotline Webring</a> – you can find links to the webrings in the nav on every page. Joining the Hotline Webring was designed to be supremely easy, but the geekring was a little more complicated. I’ve never been a big IRC person (though not for lack of trying), and I ended up joining via the HTML form and getting my key via email. I may or may not have confused my number with my private-ish key, and had to use a little bit of bfg</a> magic to fix that.</del> EDIT: I would now very much consider myself a big IRC person. My site was previously hosted using a continuous deploy on Netlify</a>. This was a pretty great system for me – all I had to do was commit my code to the GitHub repo, which I was already doing, and the they would take care of everything and serve a great site – here’s the kicker – for free. This was working really well for me for a while, until I decided to add security headers to the site. I tried doing this in two different ways: using Helmet</a> as described in the Sapper Docs</a> and using Netlify _headers</a>. The Helmet middleware didn’t work with Netlify, and I’m honestly not sure why – possibly because of how they host the file server (I’m not sure what/how they serve the site), or maybe I just f*cked up somewhere idk JavaScript is hard. Using the Netlify config didn’t work either because I’m not a paying subscriber, which wasn’t too clear at first. This would probably be solved if I just gave them money, but I’m not going to do that any time soon. I decided to move to DigitalOcean</a> and host the site myself (or I guess not totally myself – I don’t feel comfortable running it on a Pi from my house just yet) with Caddy</a>. I went with DigitalOcean partially because of their dev-focused business model, but mostly because they’re cheap (my droplet is USD$5/month and I get $50 credit with them from the GitHub Student Developer Pack</a> – my first year of hosting for only $10). Unfortunately, I didn’t quite run the transition too well. It definitely wasn’t zero downtime – it was probably more like a-few-hours downtime. I doubt this really affected anyone because of my small audience, but I was still up late working on it. If it had happened a few days later, it might have gotten mixed up with the Cloudflare downtime</a> (I don’t use Cloudflare). My current setup uses a custom CaddyServer configuration to serve my site, which I am absolutely loving so far: it is so much easier to run advanced setups with Caddy than Apache or Nginx. I wanted to move away from Javascript analytics (via GoatCounter</a>) to log-based analytics (via GoAccess</a>) but found that Caddy’s structured logs are not supported natively. I’m hoping they’ll add it</a> soon though! Also, I stumbled on this article</a> while writing this, and may check that out. I’ve got two major todos for my site in the near future. The first thing I want to do is Dockerize. Using Docker will significantly clean up my current setup, and allow me to host more fun things on the same machine (for example, my own analytics</a> and maybe git</a>). With that said, I probably should get a little better at Docker first… For beginners, Flavio Copes’</a> has a few pretty great posts on the subject. The second thing I want to do is a bit of a larger project, and one I’m not yet sure how I’m going to solve. I’ve run into some problems with the frameworks I use to develop figbert.com</a>. Sapper generates inline scripts and blobs, which tanks the security of my CSP. Their proposed solution</a> is to inject nonces with JS middleware, but this doesn’t work with CaddyServer. I would much rather Sapper avoid inline scripts</a> altogether, but this doesn’t seem likely. I really like the freedom that writing static sites in Svelte provides me versus other static site generators, but this could be a dealbreaker. I might talk a look at Routify</a> and see if that’s any better. I’ve also been reconsidering using TailwindCSS, however useful it is, after reading these</a> three1</a> articles</a>. I attempted to replicate the current look of figbert.com</a> without TailwindCSS, and failed. So I’m probably going to rewrite the site again, though I’m not sure how. Through iteration, we will arrive at a stable version – this one, unfortunately, is not yet it. If I do leave Svelte, which is probably a 50/50 chance at this point, I would probably go either back to Hugo</a> (with some variation of the incredible archie</a> theme) or to Zola</a>. So, some pretty crazy things have happened. For one, Twitter got hacked</a> by some people from OGUsers who got access to some internal managment tools. This hack, though not necessarily the most technical, was super high profile. There were a few really good articles</a> written about the hack, but I also noticed a few people acting in ways that I thought were not appropriate. More specifically, Brian Krebs</a>, who reacted to the hack by repeatedly</a> doxxing</a> the hackers, many of whom are still teenagers. He’s received relatively little criticism</a> for doing this, though it appears to be a pattern of behavior</a>. There was also a brief mention of a new macOS malware</a>, but judging by the lack of further publicity it’s probably not that common in the wild. That, or people don’t care because it’s spread mainly through torrenting sites and they don’t want to protect people they perceived to be morally beneath them – which is wrong, obviously. I learned how to drive! Or rather, I got kind-of licensed to drive a car in the US. I’ve been driving around a lot lately as something fun I can do to pass time in quarantine, and last week I finished my online driver’s ed course and got my learner’s permit. It was a surprisingly simple and COVID-friendly process! The most complicated/least COVID-friendly part was going to the DMV to take the written test, but I wore a mask and gloves and glared at anybody who got remotely close to me. The next steps are much more dangerous to do during a pandemic, but we have a year to complete them so we’re going to delay them (some in-person driver’s training and the driver’s test) for a bit. For now though, I have a piece of paper that says I can learn to drive! I’ve also taken some time to focus on decentralization. I’ve had a Mastodon account</a> for a while, but I rarely used it. I didn’t have a mobile client, and I almost never use social media on my computers. To help change my behaviour I downloaded Mast</a>, a paid Mastodon client for the Apple ecosystem. I went with Mast mostly because it’s absolutely gorgeous – and open source</a> – which is something I value in the apps I use. Apps that look really nice are fantastic inspiration for my own apps’ designs, and this one is no exception. I’ve also started using three other apps much more heavily recently: Feedly</a>, Octal</a>, and Element</a>. Feedly is a fantastic RSS reader, which helps me keep track of all the interesting blogs that I find online (I read somewhere about self-hosting an RSS reader, which sounds pretty cool, but I forget what it was called). Octal is a HN client for iOS, which is really handy for keeping up with my favorite tech news when I’m away from my computer. Element (previously Riot), is the first-party Matrix</a> client that makes using a decentralized E2EE chat service feel better than Discord. I love it. In an ideal world where I could choose the tools I use to communicate with people, I would only use Signal and Element/Matrix (for replacing iMessage and Discord/Slack, respectively). Oh my god there’s so much new music. Jonny by Bar Tsabari</a></li> Kukuriku by Eden Ben Zaken and Omer Adam</a></li> Ta’azri et Atzmech by Dekel Vaknin</a></li> Sivuvim by Eden Hason</a></li> Lecha Dodi by Moshe Peretz and others</a></li> Haravot BaPita (Album) by Peled</a></li> Magevet BaAvir by Noroz</a></li> Nadav_15 by Shekel</a></li> Moshe by Narkis and Miri Mesikah</a></li> Tik Tok by Mohamed Ramadan and Super Sako</a></li> Aliyato VeNiflato Shel Shem Tov Hevi by Tamir Bar</a></li> Falafel Pop (Album) by Quarter to Africa</a></li> Lama Kacha Atzuva by Idan Raichel and Stav Beger</a></li> Kapara by Kevin Robin and Rotem Cohen</a></li> Ehad HaAm by Aya Zahavi Fayglin</a></li> </ul> I hope you enjoyed reading this update! I know it’s much longer than usual, and I’ll try and keep them shorter than this from now on. I plan on writing more short, topic-focused articles rather than just updates, but those will come soon. For now though, I need to focus on knocking out some of the larger projects that I’m working on. I keep finding cool and interesting things to do, and I need to work on finishing the ones that I’ve already started! Salamat, FIGBERT This previously linked to https://edvinleander.com/2020/07/15/stop-the-overuse-copy-pasting-and-unecessary-libraries/</code>, but the site has been down for years now and I neglected to archive it. Such is the nature of the internet. ↩</a> </li> </ol> </section> How to Replace Keybase in 3 Easy Steps 2020-07-03T00:00:00+00:00 Ever since Keybase was acquired by Zoom</a>, a company</a> with</a> a</a> very</a> bad</a> history</a> with</a> security</a>/privacy</a>, people wanted an alternative</a>. There have been a few different alternatives proposed: this is the best</del> mine. Before we talk about replacing Keybase</a>, we should have a good idea of what Keybase actually is. It’s main features are as follows (ordered as on the website): E2EE chats and messaging (people and teams).</li> Cryptographic identity verification from around the net.</li> KBFS</a> (Public signed file hosting, private E2EE file storage w/ sharing, Static site hosting??</a>)</li> Git repositories? Crypto? An alternative to PGP?</a></li> </ul> I’m not the first person to try this, obviously. Some brave folks have tried to build Keybase alternatives, such as keys.pub</a> and the brand-new Keyoxide</a>. I’ve tried both, but found that though they both are good in their own right, they are not the solutions that I am looking for. There are a few great pre-existing options for encrypted messaging: Signal</a>, ProtonMail</a> if you want to go full email, Telegram</a>, and WhatsApp</a>. However, they all have their problems (though I use the first two on a daily basis). Signal requires a phone number, and is more of an iMessage/text replacement than a Slack-style chat app. Protonmail is literally not chat – it’s email. Telegram is (debatably) not</a> secure</a>. If you use WhatsApp for security you might be crazy – I only use it because it’s the way to communicate with people in the Middle East and Africa. Instead, I would recommend you use Matrix</a>. Matrix is an “open network for secure, decentralized communication,” and it’s the perfect replacement for Keybase’s chat and I would argue most other chat apps too</del>. It utilizes E2E encrypted messaging, and can be self-hosted as well or if you’re cheap like me just get your friend to host</del>. In addition to a Matrix server, you also need a client. For this, I recommend Element</a> – though Nio</a>, once stable, will almost surely be my go-to. Element is a beautiful Matrix client with a bunch of awesome features, including Slack-like integrations, and apps for pretty much every major platform (Linux, MacOS, Windows, iOS, Android, and a web client). Plus it looks a lot like Discord.</del> Replacing Keybase’s original function</a> is probably the most difficult part of this tutorial: cryptographically verified identity proofs is a great and innovative idea. I would swap this out with an IndieWeb</a> profile – one part</a> of the larger microformats</a> HTML structure. There are some pretty great tutorials out there (I would recommend this one</a> by the fantastic Kev Quirk</a> and this one</a> by Brian Wisti</a>), so I won’t go into too much detail about exactly how to do that. However, it’s important to note that though Kev recommends hiding your h-card with the display: none;</code> property: don’t do that</a>. I just merged my about and contact pages onto my homepage, and added the microformats classes to my existing markup. </picture> Replacing KBFS is easy to do, but hard to get right. Swapping to Google Drive</a> is probably the move that most people would make, but that abandons the entire security/encryption aspect of Keybase. There’s also Dropbox</a>, but that has the same problems as above. ProtonDrive</a> has potential, but it’s not out yet. EDIT: ProtonDrive is now in beta for paying subscribers, but I am shifting away from the Proton ecosystem for other reasons and do not recommend it. More in a future blog post. Enter Syncthing</a>. Nikita Tonsky</a> wrote one of my favorite posts of all time about Syncthing – go read it. One reason Syncthing is so great is that it’s not the same thing as KBFS or any of the other “Drive” solutions. Instead of being a file hosting system, it’s a “continuous file synchronization program” – aka p2p. You have no data limits other than your storage and no third-party to worry about. Plus, sharing folders is also incredibly easy. Just read the article. It would be a shame to talk about text chat, or really any form of communication, in this new pandemic age without talking about video chat. After all, the whole reason I’m writing this article is because the new videocalling giant Zoom</a>. So, how have I replaced Zoom and how does that relate to replacing Keybase? Well, Matrix happens to have a fantastic Jitsi Meet</a> integration</a>. Plus, the folks over at Jitsi are working on E2E encryption for their calls</a>. I’ve integrated Jitsi Meet into my self-hosted instance of Matrix, and now all my videocalls are just that – mine! Swapped chat to Matrix and Riot.</li> Swapped identity verification to Indieweb.</li> Swapped file storage/sync to Syncthing.</li> Added videocalling to chat program via Jitsi.</li> </ul> Keybase is a great service, and the people who work there should be really proud of what they’ve built. However, given Zoom’s aquisition of the company, the stability and security of the product have been called into question. So, ever one to hop on a hype train, I jumped ship. I’m really happy with my solution, and I’d love to hear your thoughts as well. Keybase Has an Onion Address 2020-07-03T00:00:00+00:00 Little known fact about Keybase</a>: it has two onion addresses. The short fncuwbiisyh6ak3i.onion</code> and the longer keybase5wmilwokqirssclfnsqrjdsi7jdir5wy7y7iu3tanwmtp6oid.onion</code>. Huh. I Wrote This #2 2020-06-27T00:00:00+00:00 Hi! So, I just started this series and I already missed a week. Oh well. This week was a crazy week for Apple and their operating systems, as they announced iOS 14 and macOS 11 Big Sur at this year’s WWDC</a>. I’ve also started working on a new game project, another little side project, and finding a good alternative to Keybase</a> now that it’s owned by Zoom</a>. When I was a kid – or at least, a younger child than I am now – I had a friend who jailbroke his phone. I thought it was super cool, and wanted to do the same. At around the same time I was watching TechSource</a>, I got deep into those Top 10 Amazing Jailbreak Tweaks</a> compilations. However, I wasn’t allowed to jailbreak my phone because my parents said it wasn’t allowed, and that was the end of that. Until last week, when I decided to jailbreak my iPhone 7 running iOS 13.5.1 using the checkra1n</a> hardware exploit. To do this, I couldn’t use my USB-C to Lightning cable</a> due to a known issue</a> and instead swapped to a USB-A to Lightning cable with a USB-A to USB-C adapter. After jailbreaking the phone, I began to install a host of tweaks. One of the most interesting things I learned about the jailbreaking community is that charging upfront for apps is quite common for premium and well-known tweaks, which stands in high contrast to the iOS App Store’s ubiquitous freemium payment model (a free download followed by a monthly subscription) especially in the light of the Apple-Hey.com controversy</a> of recent weeks. All in all, I spent around USD$21 on tweaks, which is infinitely more than I have spent on the App Store – so maybe charging upfront does work. Huh. Your browser doesn't support the video tag, which I use in place of .gifs, and/or the video formats in use here – sorry! </video> Soon after completing the jailbreak, WWDC</a> happened and new developer betas were released. Never one to miss jumping on a hype train, I updated both of my main devices to the new software</a>. These new betas come with a whole host of bugs, as is to be expected, which I plan to write about in a blog post coming soon. So far, however, I’m liking these updates – the PiP on iOS is incredible, and the new design on macOS is growing on me. Some interesting resources: An interview of Craig Federighi by MKBHD</a></li> These</a> three</a> issues</a> on GitLab for iTerm2 users</li> These</a> two</a> posts for Firefox</li> This Github issue</a> for Homebrew users. so basically everybody.</del></li> </ul> Interestingly, I was having a bunch of issues with brew (curl</code> was broken?) after updating to Big Sur but now it seems to be working flawlessly. Very nice. For my next medium-large project, I’m going to need to make some 3D art and models. One problem: I don’t know how to do that, or rather, I didn’t. Thanks internet! I powered through the incredible Blender Beginner Tutorial Series</a> by the Blender Guru</a>. Here’s my final animation: Your browser doesn't support the video tag, which I use in place of .gifs, and/or the video formats in use here – sorry! </video> Now that Keybase is compromised</a>, I thought it would be good to seek out an alternative. Initially, I thought that keys.pub</a> would be perfect – I thought it was a slightly-tweaked fork. However, this does not appear to be the case. Instead, it seems kinda like garbo – it doesn’t have the budget Keybase had, and as a result, is majorly lacking in features. However, I’ve replaced Keybase completely, and am happier with this setup than I ever was with Keybase. My new setup uses Matrix</a> and Riot</a> for messaging, Jitsi</a> for video chat, Syncthing</a> for file storage, and an Indieweb</a> h-card for identity verification. I plan on writing more about this at a later date, but I’m really pleased with the results so far. I’ve got a few interesting projects in the pipeline at the moment, not counting the few articles that I’ve already promised to write: Glitches in macOS Big Sur and How to Fix Them and How to Replace Keybase in 4 Easy Steps. My next major project is a dreidel game/simulator for mobile devices – I plan to code it with Godot</a>, make the models with Blender</a>, and do any music/sound stuff in FLStudio</a>. Naturally, I’ll keep you all updated on any progress there. I’m also working on a much smaller project that should take around a day or two to complete. The working title is “howmanyisraels.com” – essentially, it’s a website that takes the size of the State of Israel, and compares it to the size of a given location and tells you how many times Israel could fit inside it. All I need to do is find a suitable mapping API (preferably not by Google) that will tell me the size of different locations and I can whip it into a static site! I’m also planning on revamping the txtodo macOS app with Mac Catalyst, which I feel I haven’t properly explored. I’ve also discovered a bug in the app that causes issues when tasks are modified by CloudKit instead of the user, so I’m going to have to work on that as well. Shouldn’t be hard! </picture> I’m not sure what tense to write my headers in – I currently write them in present continuous</a> tense but I am considering writing them like I do Git commits. Huh. I’m also committing to writing more non-update blog posts (i.e. the two I mentioned above), but we’ll see how that goes. Writing is hard, and you shouldn’t make false promises. I’m also working on implementing webmentions to the site, which is a little difficult with a static site but I’m going to give it a try. Lastly, I really need to learn how to use vim but it’s so goddamn hard. I’m writing this blog post with it but I’m definitely not using it to its full potential at all. Wow, these have been a great two weeks for the Israeli music scene. In no particular order, here are some good new songs (basically in the order they appear on the trending tab): Habib Albi – Static and Ben El</a></li> Rak Banot – Itay Levi and Stephane Legar</a></li> Shvuaim – Eden Hason</a></li> Yom Huledet – Eliad</a></li> Hatzi Madleket – Dudu Faruk</a></li> Hayde – Peled</a></li> Ze Mi She Ani – Peled</a></li> </ul> I know music isn’t tech, but I like it and now it’s a recurring section. So, that’s been this week’s update. I’m really enjoying writing updates on this blog, and I hope you stick around. Until then, farewell! – FIGBERT I Wrote This #1 2020-06-13T00:00:00+00:00 So, I’m FIGBERT – I wrote this. nailed that intro.</del> I’m a highschooler working in the tech world, currently stuck at home like pretty much everybody else who’s not trying to die. I plan on using this blog to writing mainly on my experiences in the tech world, either in focused articles centered around one topic or more summary-style roundups like this, the “I Wrote This” series. I’ve always found statements like that kind of funny in their generalness, and was thinking the other day that if I ever made a game studio I would name it “I Made This Game.” Then, when I was trying to think of what to name this series (The Weekly Fig? Fig Talks?), I thought it would be funny to do something similar. I plan to release these once a week on Saturdays, if all goes to plan. Without further ado, here’s a brief summary of interesting things that happened this week. Some people reading this may be familiar with my most recent project, txtodo</a> – you can read a full writeup of the project here</a>. The point of txtodo was to create a todo list application without any feature bloat I found in other competing solutions – every task deletes at midnight. This week I launched the MacOS companion app for txtodo, completing my long-term “cross-platform” goal for the app. I went from not knowing SwiftUI at all to having two published apps that talk to each other! Making computers talk to each other is, like, the hardest thing you can do with computers! Ok, technically a lot of it is just Apple APIs but still, let me have this.</del> I think there is probably one more major change I could make to the app (moving the tasks from an @State</code> variable to an @EnvironmentObject</code>), but after that it’s pretty much bugfixes and… the bank? I don’t know. I while back I saw that when I editing my project README.md using the GitHub web app, it got a little green “verified” badge next to the commit message. I looked into it for a bit today, and found this GitHub help article</a>. One GPG install and Protonmail-key-download later, now all my commits have a green badge! Neat! I made the switch from nano to vim a few months ago, but up until a few days ago I was using the barebones default macvim</code>. Not anymore! I DuckDuckGo-ed “vim configuration,” and downloaded what is supposedly “The ultimate Vim configuration.” from amix</a>. It’s a pretty good configuration out-of-the-box: I’m using the Awesome version with the fantastic Jetbrains Mono</a> font. I’m still not that good at actually using vim to its fullest potential but I’m confident that I’ll improve with a little practice and ascend to elite vim status in no time. </picture> Speaking of elite programmers: Rust! I’ve been thinking about learning the language for a while, and now that I have no main project (other than maintain my finished apps and sites), it’s as good a time as any to learn a new language. It’s idiomatic syntax combined with low-level abilities makes Rust seem a little bit like Python+ to me. It also helps that it’s skyrocketing in popularity – maybe I’ll do the cliche Rust developer thing and post about how much I love the language on the orange site</a> once I learn it. ok so I was planning on linking a bunch of posts like that here, but when I looked for them on HN I actually couldn’t find any, but I distinctly remember reading them so I stand by this statement.</del> I made the switch from Brave</a> to Firefox Developer Edition</a> this week, after accumulating around 50 USD in BAT. I started using Brave almost when it came out and loved it – the chromium feel with a powerful built in adblocker? It sounded too good to be true – and perhaps it was: over time the browsers performace began to slow down, they hijacked links</a>, launched a weird Zoom competitor</a> with no fanfare or announcement, and refuse to get rid of their required KYC</a>. Anyways, I dropped it. Now my dock is all blue! </picture> On a non-tech-related note, a bunch of good new music came out recently. In no particular order, here are some good new songs (largely out of the Middle East): Coronavirus by Mohamed Ramadan</a></li> Zot Ani – Ela Li Lahv</a></li> Million Dollar – Noah Kirel and Shachar Saul</a></li> Naughty – Tzagay Boi</a></li> TROLLZ – 6ix9ine & Nicki Minaj</a></li> </ul> So, that’s been this week’s update. I’ve got some exciting ideas for my next posts, so I hope you stick around for what’s to come. Until then, farewell! – FIGBERT txtodo 2020-05-20T00:00:00+00:00 txtodo</a> is a minimalist open-source todo list app inspired by Jeff Huang’s One File to Rule Them All</a> and Mac-assed Mac apps</a> like IINA</a>, Sketch</a>, and Safari. It’s available for download on iOS and macOS via the App Store. txtodo lists your immediate, short-term tasks to help you get things done without overthinking it. At midnight, all tasks are discarded so you can start fresh tomorrow. You can also create up to three long-term tasks that “float” with you from day-to-day. On 15 February 2020, I deleted my account with Trello. A solo developer, I know my projects inside and out. I know what features need work, and which graphics need a refresh. I don’t need a massive crudware</a> suite to keep track of every little move. I need something small. A dead-simple UX and midnight expiration date prevents the overwhelming buildup of tasks we delay over and over again, and helps you deal only with tasks you can get done that day. Download the app on the App Store</a>, and explore the source code on my forge</a>.

FIGBERT

Google Refreshes Brand Icons to Make Them The Worst

Like Someone Who Actually Just Wants to Talk About the Settlements

ME163*

Coconut Water

Stanford Quarterly Reflection (Y3Q1)

</svg> </a> BIO81: Introduction to Ecology</h3> Three good things came from this class:</p> It satisfied the first third of my Domain Focus for the Design major.</li> I learned that palm trees are, in fact, a grass.</li>

Aspirational Surfing

Recurse Center

The Fall of a Great Product

Projectile Launcher

I Think This Might Have Been Epistolary

Developments on the Git Forge Horizon

Vault

How to Format a micoSD Card for the Nintendo Switch

Two Pockets

Text Conversation, Transcribed

On the Recommendation of a Friend

Cloudlab

Stickies

Cordially

Rapid Riter

Night

Bad Book Good Ideas

Blind Date A Book

It's Time to Build New Hardware

Chronological Sort Considered Positive

Stanford Quarterly Reflection (Y2Q1)

</svg> </a> Academics</h3> This quarter is what Stanford is supposed to be. I learned. A ton. I was enrolled in:</p> MATH 21: Calculus</li> COMPLIT 37Q: Zionism and the Novel</li> SUSTAIN 345: Sustainable Societies Lab</li> DESIGN 11: Visual Thinking</li>

Stanford Quarterly Reflection (Y1Q4)

Biting the Hand That Feeds You

A Screed of the Old Style

Roll Call

More from Tarpenning

What I've Learned from Stella Levi (So Far)

Baymax on Memory Lane

Waking the Public to Waking Lions

Marc Tarpenning on Innovation

Daring Fireball on Love

The Novel of Palestinian Disillusionment

AirPods and Thoughtspace

Contra Shipping in Public

An Insight Tangential to Thorsten Ball's Intention

I Am Amazed That This Exists

Some of This and Some of That

Envelope

Beeper

Stratchery on EU Antitrust

Ideating Tragit

Stanford Quarterly Reflection (Y1Q3)

</svg> </a> Academics</h3> I was in the lowest amount of units possible, a measly 12. These were:</p> COLLEGE 110: The Spirit of Democracy</li> CS 107: Computer Organization and Systems</li> CS 12SI: Spatial Computing Workshop</li>

Running Out of Time Wallpaper

Caret

CSS System Colors Dashboard

Stanford Quarterly Reflection (Y1Q2)

A Lovely Read

I Suppose This Is Modern Fiction And It's Pretty Good

Toward Tangible Deliverables

Stanford Quarterly Reflection (Y1Q1)

Epic Acid Science Fiction

A Foray into Theology

Quotes from My Camera Roll

A Review That I Should Probably Have Saved For Another, Less Contentious Time But Nevertheless Did and Ultimately Enjoyed

James Hoffmann x Bellwether Coffee

Spies of No Country

The Thousand Crimes of Ming Tsu

Movies I Saw Other People Watching on the Plane

Naftali Bennett's Inspirations

Neuromancer

Who by Fire

Moonwalking with Einstein

The Desert and the Sea

Pomegranate Seeds

Mabel

Aggregation Theory, Virtuous Cycles, and Nitter

d3c3nt

Privacy Redirect for Safari

SimplyTranslate

An Update on Bad Backups and Internet Celebrity

</svg> </a> To Infinity and Beyond</h2> I’d also like to say thanks to a number of people who were especially kind, and made this a fun experience:</p> Graham Percival, an employee at Tarsnap Backup Inc., reached out to me directly – we had a fun and interesting email exchange.</li>