Musings from the Chiefio

This is Darned Close to the OS I would build for me

So Lunduke has a very similar set of desired attributes to mine. No SystemD. No capricious RUST. No political crap. No age & ID mandates. Etc.

Where I diverge?

I don’t care about having a “code of ethics” even as benign as his is (the 10 commandments). I just want the damn thing to work.

I don’t need my terminal server to be “a little fun” nor have whimsy. Just needs to look like a terminal.

I don’t mind the regular Brave Browser look & feel, though his choice of the Brave Origin Browser is OK too. But i tend to install multiple browsers, so if he doesn’t include the regular repositories, I’d have to add that.

I prefer ARM chip based systems. I do have a few x86 / AMD 64 bit systems, so I can at least try it out on them. Including my daily driver laptop. I would hope that, fairly soon, an ARM version would come along…

What I like

It is based on Devuan. My present preferred OS. They DO have “community” ARM versions, so a fairly easy path to ARM exists too. Devuan brings with it No SystemD. Yay!!

He is deliberately doing the “poke the bears” on Age and ID mandates. Even if Devuan folds on that (and I do not expect them to fold) he will not fold. As this is a USA based distribution, I think the 1st Amendment protection for Freedom Of Speech (which also prohibits Forced Speech) will eventually prevail, but it could take a decade or two… I do NOT need my personally built and owned computer nagging me about my age every day…

It has a clean, simple and minimalist aesthetic about it. If defaults to LXDE which is my preferred windowing system ;-)

I think that his emphasis is on just making things that reliably work, so expect it will more reliably just work. Which is what I want too.

He is actively going to prevent a bunch of capricious changes and crap from flowing downstream into my grill. Politicized projects. SystemD demented changes. Age & ID crap. Rusting code bugs. etc.

He will have his own code archive and build system, so I suspect i can easily mirror it into my own code archive and build system, and get it 99% of the way to exactly what I want just from that. While I can also get most of the way there off of the Devuan archives – some of their build process documentation and mirror making documentation are a bit less than the clarity I would like…especially on the ARM side. so I expect it will be easier to build off of his.

Being a One Man Show, I expect “new releases” will come along a bit slower, and thus easier for me to keep up with them. As it stands, when I get around to a system release update, I’m usually 2 back and have at least once had to go to the aged out archive to do the system Update that would then let me do the regular system updates…

I’m sure there’s more… but I’m not going to list everything.

The Lunduke Video

Link to his referenced LCOS page:

https://github.com/BryanLunduke/LCOS

In Conclusion

I’m not in a big hurry to try this out. It is a 0.1 release after all. But I’ll eventually get around to trying it. I like the ideas behind it, and I expect it will do a good job of isolating me from a lot of the Unnecessary Drama that has entered the Linux World in the last couple of years.

So mostly this is just a Heads Up that it exists, and that it may matter in the future. I’ll probably get around to trying it out in a few months to a year…

Want to talk through cement buildings?

I found this video about mesh nets, encryption, and local WiFi coms of interest.

I don’t need any of the tech he describes, but I want it ;-)

It is also interesting just from the POV of what is happening in the DIY sector. ISP & Cell Company cutting off connectivity if (when?) things get politically questionable? No Problem… Just fire up this gear, connect your group’s cell phones to it, run Reticulum for encryption if desired, and uplink the whole group to Starlink. Global coms restored…

A bit pricey for one guy to just play with when he has no real use case, but way cheaper than any alternative commercial product for the same capability. For me, my contacts will all either be in visual range, or will be 1/2 an ocean or a whole continent away and on their own.

But I still like the tech and where they are going with it…

FWIW, I’ll be mostly sticking with Satellite link, adding LoRa mesh mostly for fun, and using Cellular Phones until that goes offline in any Aw-Shit. Plus a radio for monitoring & fun (thanks Ossqss!) But I still want to play with this ;-)

As Rust (computer language) True Believers force fit it into projects…

WD-40 removes the Rust…

This is as much a Tech-Political posting and a bit of humor as it is an actual tech posting. For anyone not aware of it, there is a computer language named Rust that has a lot of enforced memory usage discipline marketed as “memory safe”. This is as opposed to C and C++ that pretty much let you do any stupid thing you would like to do (but also let you do a lot of very nice fast efficient things, and do them well and simply, where YOU must know what you are doing with memory).

The advocates for the use of Rust call themselves “Rustaceans”, but I tend to think of them as “Rustians” (for no good reason other than it being shorter and easier). In any case, they have a certain religious zeal in how they advocate for Rust (really too weak a description… more like “proselytize” or “enforce”).

Well, the background on this is that, per Lunduke, the Rustians have discovered A.I. coding tools and are using them to stuff Rust (and thus dependency on the Rust tool chain and libraries) into all sorts of Open Source projects (even into the Linux Kernel…). Whether this is good or not depends on if you are a Rustian or an old school C Curmudgeon.

For me, being a bit of a C Curmudgeon, I see no need, at all, to force dependency on a second tool chain (and all the memory, computes, and dependencies that entails) into a project that already works “as is”. That’s from the “if it ain’t broke, don’t fix it” school of thought (that has served well for decades…).

Well… aside from just the political drama and Tech Food Fights injecting Rust into the Linux Kernel has caused, there’s the technical point that in order for Rust to actually add any memory safety, you MUST use calls to Rust libraries, that are not used by non-Rust code. So as soon as Rust has to talk to some C program or library, you generally lose any memory safety “enhancement”. So from my POV: a load of resource consumption, complexity, and drama inserted into a project for not much to zero gain.

Basically Rust and C are incompatible in their philosophy at the library level, and that incompatibility requires one of the two programming schools to adapt to the other. Rustians resent that they have to adapt and thus lose their only feature (“memory safe”). While the established C project resents being asked (forced?) to conform to an alien invader. Thus the Drama of Rust-ing. Neither side wants to re-write their libraries to accommodate the other (if that is even possible).

The Rustians attitude is basically that the whole world ought to convert to Rust anyway so might as well start with it, and that will help the Troglodyte C programmers get motivated to learn and use The One True Language!

Needless to say, I’m a Troglodyte… Frankly, over the last, what? 45 years now? 50? I’ve seen the One True Language burst on the scene, flower, fart and fade so many times I’ve lost track. I’ve also (as of about 25? years ago) decided to just wait a decade before hopping on the next One True Language bandwagon.

I’ve been through the Pascal fad. The industry driven PL-1 push. The military ADA demand (and was sent to Germany to look over an ADA compiler). Perl. Ruby. Ruby on Rails. That whole Object Oriented movement that includes C++ and C# in the Microsoft world. Then the Python push from Academia into industry (still ongoing) and even my recent exploration of Julia as a parallel processing friendly language.

I’ll skip over the Shell Wars… From sh, to csh, to ksh (Korn Shell), and on to bash, etc.

Along the way I learned FORTH (a fascinating and very useful language especially for small hardware) and used it a bit. COBOL that is still hanging on in very old code all over the business world. RPG (don’t ask…) and ALGOL (popular in Europe in the ’80s & ’90s). I’ve also watched FORTRAN evolve into Fortran that looks like a bastardized ALGOL / Pascal hybrid. Wrote one APL program once, and the next morning could not remember how it worked… (sometimes called a “write only language”).

And likely a dozen other languages that came and went from my workspace as various maintenance projects and contracts required. (Also leaving out the “4th generation non-procedural languages” I used for various databases and other database systems – about 1/2 dozen I think…)

The point?

I don’t come to this from a position of Language Advocacy. I use whatever I have to to get the job done. Often that meant picking up a “new language” in a few days so that I could fix something written in that language. To the extent I have a bias, it is just this: Don’t complicate things with multiple odd languages if you can keep them simple. A kind of KISS principle… Simplify and polish beats The New Language Du Jour.

Well, the Rustians have come for GIT. Lunduke covers it nicely. They’ve shoved some Rust into the Git project. Why? Was it broken? Nope. Just to spread, to proselytize, the One Pure Religion Language of Rust. Because “it is good for you”…like it or not.

OK…

So someone didn’t like it, and came out with a fork. The developer of Xlibre produced the “Git 2.55.0 WD-40 flavor” (because WD-40 removes rust…).

In a reply to Lunduke, he points out that A.I. tools can remove Rust as effectively (or more so) than they can insert it ;-)

So A.I. has come to the Rust Wars. Competing robotic tools molesting open source code. Reminds me a lot of the “block structures” vs “go to” wars of decades back… (I was an advocate of the ALGOL / Pascal like block structure and another programmer was using non-structure goto like style. We were both doing maintenance on the same code base. I’d “clean up” a section for clarity, then patch it. He’d undo my “clean up” and reapply my patch and his changes. I eventually automated my reversions of his reversions and he gave up…)

The human behaviour does not change, only the coding style of the Food Fight… and sometimes the Language du jour…

Here’s Lunduke with The Rest Of The Story:

FWIW, to get a sense of just how many computer languages have come, and gone sort of, and why one tires of the Language du Jour wars: take a look at the Rosetta Code list:

https://rosettacode.org/wiki/Category:Programming_Languages

After a while one becomes jaded about the need for Yet Another Programming Language that this time for sure! will solve the problems of human bad coding practices with Yet Another Straight Jacket… or that promisses vast improvement in coding efficiency via channeling the programmer into a limited set of pre-packaged behaviours.

Originally Library Functions were going to make programming vastly more efficient. And, to this day, FORTRAN libraries are some of THE most efficient math methods on the planet.

Then Object Oriented programming was going to do it. BUT you had to learn their whole catalog of functions that was a huge workload in itself.

And on and on…

Sidebar on Me: I’m not against Rust. You do you. I’m against forcing others to change when they do not want to. (I’ve learned from my block structured war…) I’m against Missionaries Proselytizing to an audience who does not wish to hear their “message”. (Be it Christianity, Islam, Object Oriented OO, or Rust…) I’m for just leaving people be and doing your thing as a living example… Unfortunately, when dealing with an existing project code base “doing your thing” by injecting Yet Another Language & Tool Chain is proselytizing.