Alright, so
131
Replies
0
Boosts
As you may have gathered from some of the posting that Sundog and I have been doing over the last few days, we're kicking the tires on some tech stuff.
Toy production is still ongoing, we're getting ready to order more molds, I met with one of our 3D modelers yesterday about the next models we'll be working up.
You can get Kumiai-Ki, The Mighty Labor Union Mech action figures from @MountainTownToys right now!
But we've solved a bunch of toy production problems, and half solved a couple of others, and a byproduct of that is that we have some extra cycles while we're waiting on parts and tooling to finish solving problems and start making the next batch of toys.
And that means we've been working on some tech stuff.
For me, this mostly means that I've been dusting off old ideas and checking out new projects and checking in on old projects.
I've been writing a lot, about technology, about what I want, about culture, about the ways technology has failed.
Some of these things might eventually become blog posts or a newsletter or a zine.
But mostly, I've been thinking, and trying to fight back in small ways to find or build a new foundation as the forces of capital erode the ground from underneath my feet, technologically speaking.
@ajroach42 So you know: those posts aren't going unnoticed, at least by me. I'm delighted you've had the bandwidth to share them with us, and your approach to wanting back what made "retro" computing useful without falling into unthinking fandom has really helped my response to the same feeling of erosion.
I have, for decades, considered Debian the bedrock of my computer infrastructure.
The thing that I could trust, because it was run by an organization so obsessed with the ideology of software freedom that it routinely rendered itself significantly more cumbersome as a result.
I was there for the IceWeasel campaign! I thought those principles were central to the project.
I was wrong.
So one of the many things I'm trying to do, as I reorient myself in the computer world without Debian as a foundation to build on, is looking for a new operating system.
The Linux kernel project accepts slop contributions, the debian project accepts slop contributions. Where does that leave me?
Put simply, I'm looking for a daily driver operating system that I can trust to be made by people, and not contain the product of an automated plagiarism machine.
@ajroach42 I am trying NetBSD, which you're already considering, but FreeBSD is a bit more polished. what's the LLM sludge factor on FreeBSD though?
Haiku (the BeOS clone) has a fair amount of software for it but it's not quite reliable I'd say, not compared to BSD
@mxchara https://retro.social/deck/@ajroach42/117208626929006880
FreeBSD doesn't have an official AI policy, but so far they've rejected most AI code, but allow it for use in docs and other places.
@ajroach42 OS selection is a hard one, especially once "Intel (compatibles)" is not taken for granted anymore.
That basically leaves Linux (supported by everybody, to some degree, a lot of that in branches that really need an uplift to something current) and NetBSD (less wide coverage, but more focus on keeping things running in a single code base instead of the "every vendor makes their own mess", although that's by "virtue" of fewer vendors actively supporting it, instead of hardware being supported by hobbyist efforts.)
Everything else is rather aspirational in terms of platform support, sadly.
@ajroach42 I'm currently eyeing RedoxOS (strong anti-AI policy), and NetBSD (less strong anti-AI policy1). I've liked microkernels ever since I first played with the HURD, and I like Rust, so RedoxOS is very interesting.
Unfortunately, I have not been able to install it in a VM yet, so that's not an option right now. But something to keep an eye on.
NetBSD on the other hand, works fine so far, and as far as server-side things go, I haven't missed anything yet. Apart from a declarative configuration a'la NixOS, but... that's missing from anything but NixOS & Guix.
NetBSD's policy is that tainted code is banned, except if core gives an OK. They have not done that, and NetBSD has been pretty consistently going their own way, so I have some hope in them not going slop before the bubble bursts (and there'll be no point after). ↩︎
(No, I'm not interested in arguing about the merits of AI, or hearing about how your use case is justified or okay or whatever.
I am opposed to generative AI and LLMs on several levels, but
I don't care what you do personally, and I'm not here to police your decisions.
I don't want AI generated code in my OS, and I don't want to talk about it.)
@ajroach42 I have the same attitude!
Gives me a better chance to focus on the things I actually do care about!
The most complete OSes I've found so far that currently have bans on direct LLM contributions are either https://www.haiku-os.org/ or https://www.redox-os.org/
Guix got closer to a ban than Debian but also failed.
I think technically and license wise https://ironclad-os.org/ is interesting (and has an AI ban), though they don't seem to be anywhere near as complete as haiku or redox.
I wish the pro AI Debian developers would hear about how much end users hate the results of the LLM vote.
And I really wish the GNU project would make up it's mind before trusting emacs, guile, or hurd for the future.
@alienghic I've landed on netBSD for the timebeing, I think.
Of course, all of this is made all the more complicated by the fact that, along side most major software projects allowing AI code contributions, a bunch of software projects are ever more obviously being operated by Nazis.
And I don't want bigots writing my code either.
So I'm exploring.
I didn't think I would need to find an operating system that wasn't Debian or a Debian derivative, so I haven't really paid attention to the larger OS world.
From what I can see, the operating systems that have taken the most strong stance against the kinds of generative AI utilization that I'm concerned about here are NetBSD, FreeDOS, and ... illumos (or at least some Illumos distributions such as Tribblix)
I've also seen some general "we don't want AI here" conversation from Haiku.
@ajroach42 haiku is super neat and has some cool window management stuff, but last I looked they didn't have a good daily driver browser which makes it a non-viable daily driver OS to me
If I weren't going through a very slight tech burnout I'd probably be shifting to a BSD distro, but right now I'm just enjoying working on my model trains outside of work
@trainguyrom I spent the last three years out of the industry, and the last two years entirely embedded in making television programs and toys.
And, as a result, I have the mental willpower to start working on some of this stuff again.
It took 3 years.
@ajroach42 I've been poking at this too. No idea how functional it is, but I believe that Redox might also be anti-AI and anti-Nazi. I think there's some public statements suggesting that's not the case by some project contributors, and a fairly sternly worded response from the project originator, but I'm not sure how firm it is in the community at large. Might be worth a look (I am considering a VM). I believe Haiku is also in a similar camp.
It's a frustrating pickle.
@Unlikelylass I haven't heard of redox before today, but a couple of folks have mentioned it.
I'll take a look, but I suspect NetBSD is where I'm going to land for at least the short term.
But hardware support among these alternative operating systems is inconsistent (with NetBSD sitting at the top of the pile, but well under Linux and Windows.)
And, as with all things, there are degrees to all this. I recognize that I will probably end up having to make compromises in some of the things that I use in order to continue to do much of the work that I want to do.
In much the same way that I begrudgingly keep a windows 10 box around so that I can interface with some windows only hardware, I will allow Linux on devices that support Linux, or for activities that I can't perform in other environments unless/until I can afford to replace them.
But, if I'm going to be exploring alternative OSs and reexamining the foundational layer of how I approach computing, I might as well question it all.
I am looking at new ways of doing things, and new software, and new ideas about how software should work, and new ideas about what is *good* about the software community, up and down the stack.
I'm not just questioning my OS, I'm questioning everything about modern computers (and not for the first time! I was already banging this drum nearly a decade ago https://ajroach42.com/observations-on-modern-computing-the-last-10-years-were-a-misstep/ )
@Cyberneticist See three posts up in the thread: https://retro.social/@ajroach42/117208626929006880
@ajroach42 whoops, apologies, I missed that!
So yes, I'm going to set up a NetBSD laptop and a NetBSD server and start figuring out how much of my core life I can move onto them.
I'm going to migrate things off of my yunohost server and into the NetBSD server. I'm going to try to use Linux only when I don't have another choice.
I think these things are pretty obviously values aligned with the position that I've taken.
But I'm also going to start doing some other things.
@ajroach42
I've been using NetBSD for over 20 years. I'm willing to help if you have questions.
@ND3JR I'm just looking at the edges right now, but I appreciate the offer.
@ajroach42 it might be interesting to look at using NetBSD's pkgsrc with Linux on systems were you'd rely on Linux' hardware support. I have very good experiences with that combination.
Generally I am quite attached to GNU, of which NetBSD also includes a lot, but I'm also curious to play with the BSD userland at some point.
Another interesting candidate might be HyperbolaBSD.
@jschwart Thanks for the tip.
@ajroach42 Well that's 2 people in almost as many minutes mentioning NetBSD on a laptop. I'll be interested to hear how your journey goes, if you wouldn't mind sharing.
@tehstu I'm sure I'll be talking about it.
I'm starting with a (currently semifunctional) pinebook pro, so step one is to make it work again.
@ajroach42 @tehstu
You guys definitely know more about computers than I do, but is it worth considering #OpenBSD ? As a noob, the only thing I haven't been able to do is reliably mount a USB drive.
@doboprobodyne @tehstu whats their current AI policy?
This bit might get a little controversial, and I want to make sure that I say what I mean to say, and that means I'm going to take my time to write it.
I haven't stopped, I'm just being careful.
Alright, I also made a sandwich and grabbed some chips. I should be slightly more able to speak correctly.
Everything that's happening in computers today is evidence to me that the free software movement failed.
It's not so much that the free software movement has failed. It's like saying the antifascists failed in fascist Italy. Like duh, right? But is the reason that they opposed fascism? Is the reason that they were designed to fail, or that they were never a good idea? No the reason is they got shot and killed in the streets.
Similarly, when it looks like the free software movement is failing, adjust your scope a bit broader. Software is failing, because society is failing. Once society stops failing, all the AI bros making it fail will be pushing up daisies, and free software movements will work just fine. We have to form a better society, not shit on open source and talk about what a loser that fat bearded man is.
You're selling toys. That does more to bring light back into the darkness of society than any fork of Linux ever will. Both are good ideas IMO, but keep your sights on what really matters. Fucking over the exploiters, and supporting each other.
The Free Software Movement is a lot like Libertarianism. It sounds great as long as you don't look at the details.
@rrwo I haven't said that part in this thread, but I've said it out loud a couple of times this week.
The opt-out libertarians of the whole earth catalog became the free software movement.
And they had some great ideas, but they were also wrong.
Rugged individualism is a path to destruction. We need communities.
Yes, but what is a "community"?
A lot of software projects fall out of fashion, their maintainers get older or burn out or move on to other projects.
Most people are unpaid.
Ecosystems don't have a good way to pass ownership to a new generation, yet all sorts of unsavoury entities can use the software.
Keep saying it's not that they can use open source software. It's that they can force us to use their closed source, secret, proprietary software. That's the problem. We can't solve it by continuing to let them do that, and just try to be super secret ourselves.
@rrwo this gets out of tech and into community building and organizing, which is a much more interesting topic that has been written about for centuries.
For a somewhat recent text, I suggest emergent strategy, post scarcity anarchism, anarchy works, or mutual aid. But I tend to subscribe to anarchist schools of thought in community organization, and that's a whole big complicated can of worms that I don't feel like fighting about today.
@ajroach42 yeah 100%. I think it goes to show that property rights over information is only enforceable retroactively due to the fact that there’s literally no difference between transmitting and copying and that’s a big tube of toothpaste to not be able to get back into.
We relied on property law that was designed to protect people with more property. It’s all a lot of words on paper when your code is used to hurt people.
Or, at least, depending on which vision you have in your head of Free Software, it has either entirely failed as a movement, or it has ended up where it was always headed.
It’s amazing to me how we’re watching the tech world fall apart at the feet of monetization, enshitifiction, AI, and whatever other Bullshit trend they’re chasing this week.
The world we're living in, where privacy disappears more each day and everything gets worse, is basically the consequences that the Free Software folks have been warning about and talking about for decades!
But, by and large, the tools that are being used to implement these consequences *are* so called "Free" software.
Even before we let AI eat everything, our so called alternatives were compromised.
Canonical partnered with Amazon to send them your keystrokes, Firefox was almost entirely funded by Google, Duck Duck Go has always been Bing in a trench coat.
And, even outside of AI, one of the most widely used Free image editing programs is named after an ableist slur, and is actively hostile to anyone who suggests the name should change.
Free Software alone was never going to save us.
@ajroach42
To be fair, the slur is one that is mostly only used in the US (we have our own ablest slur here instead that people in other countries throw around not realising what it means here). Over here, the association of that word is with the sex thing. Either way, though, not a particularly good look for serious software...
@ddlyh indeed, neither is acceptable.
@ajroach42 yeah the name of that graphics package is always bothered me. Great program, tone deaf branding, at the most charitable.
@phaedral Actively and intentionally offensive branding.
It was intended to be a silly joke, but when the devs figured out that it upset people they doubled down again and again.
@ajroach42 The problem being that free software costs money to build, and any attempts to try and recoup those costs for full time development are met with backlash. So the only real options are for development to move glacially slow, or rely on a constant churn of developers.
Like, as far as I can tell, only the Open Home Foundation behind Home Assistant has had a good run with this. Partly because building hardware just works well into the business model of Home Assistant, and there's a desire by end users for compatibility for it for them to have an ability to demand royalties for its "works with" program.
That said, Home Assistant is in a niche. A lot of other projects aren't that lucky to have that kind of star power.
@soulfire
Does it?
Software generally gets written to scratch your own itch.
Free/open software as a concept/culture is actually what makes it easier to share your efforts with the world.
Putting any form of monetising onto it brings quite a bit of paperwork and the taxman into it. (Literally the shareware deal nowadays in many jurisdictions triggers that.)
@ajroach42 I honestly thought gimp was a sex fetish slang thing, not an ableism thing. I looked it up, and 🥴. 55 and learning something like that.
@nicksalt It's either or both, depending on which part of the english speaking world you're in.
@ajroach42 @nicksalt I wish they would climb down off their high horse. A rebrand to Imp wouldn't be all that hard.
@phil_stevens @nicksalt At this point it's intentionally malicious.
@ajroach42 @nicksalt It sucks because I categorically Do Not Want to learn another graphics application.
But.
Here we are. Any suggestions?
The free software movement did some good! I don't want to dismiss it. I am deeply grateful for the existence of the free software movement, and computers would be worse and less safe without it.
But it was insufficient.
@ajroach42 I agree. Free software was/is insufficient... but now what? I'd like to see more discussion on "post free software" models. How can we create a safe space again?
I jotted down some thoughts on this here:
https://iooioio.xyz/convivial-software.html
What are your thoughts? Back to closed-source shareware?
@iooioio This is a long thread, still ongoing. You've replied to a post in the middle.
Consider clicking through.
@ajroach42 Yeah, this is the part that I wanted to respond to.
I see more posts are popping up. Look forward to seeing where this goes.
@iooioio Two years ago, in this blogpost https://ajroach42.com/community-software-building-the-future-inside-the-rotting-husk-of-the-past/ , I started talking about an idea I was calling Community Software.
I've written a fair amount about it since then, but I haven't published most of that because I needed to take a break from computers for a while, and focus on doing things that made me less of an angry person.
Ultimately, I think the path forward will be a social one, rather than a technical or legal one.
What comes next, I think, is a movement for Human Scale, Human Centered software.
From the very beginning, it was also about power, about wielding power, about being protected.
Many of the people associated with the free software movement got a free pass to be horrible, bigoted, or abusive, because they also wrote good code. (or, even sometimes because they wrote code at all, regardless of quality.)
This is playing out now with DHH, but it's been there from the very beginning, with far too many people to list.
For years, I was of the opinion that it didn't matter who wrote the code as long as the code accomplished my goals.
I no longer believe that.
Ideology influences decisions, controls the possibility space.
This isn't an absolute, and it doesn't wholly dictate what software I will use (I am, after all, still an internet user in spite of it's DARPA origins), but peoples values influence their choices and you gotta remember that when you're looking at any given piece of software.
But that's neither here nor there because the worst damage the free software movement did was in expectations.
The Free (libre) software movement seeded the idea that software would also be Free (gratis) and that anyone charging for a piece of software must be punished.
And, in the process, broke the only mechanism for funding software development that ever really worked, and replaced it with telemetry and advertising and surveillance and subscriptions.
@ajroach42
> The Free (libre) software movement seeded the idea that software would also be Free (gratis) and that anyone charging for a piece of software must be punished
I really don't think that's fair comment. I started learning about Open Source in the late 1990s, which led on to learning about software freedom in the early 2000s. I was downloading gratis software for Windows before I learned about either, as were most people I know.
(1/2)
Where the software freedom movement fell short was not actively researching, developing and promoting internet-enabled revenue models for funding Free Software development. Imagine if projects like OpenCollective had started in the late 1990s, or Free Code apps released as shareware/ nagware (ie compile it for yourself or pay for the full version).
I was one of the few people talking about this stuff in the late 1990s/ early 2000s.
(2/2)
@ajroach42 "that anyone charging for a piece of software must be punished", I've seen people accuse the free software movement of this, but I never saw that idea being promoted by the free software movement. On the contrary the idea is that one gets paid for writing code (instead of receiving money over and over for the same code).
The majority (95%) of the software written has only a single purpose and user why such model works for most programmers.
@jschwart It certainly wasn't promoted from within the developer core (most devs want to get paid) but many of the most enthusiastic free software evangelists (at least, when I was coming up in the FOSS world) were former WAREZ dealers for whom the idea of paid (l)OSS was sacrilegious.
@ajroach42 One half-formed idea that has been rattling in my brain is that the idea of free software ended up working exclusively for infrastructure.
Because collective software doesn't have to be gratis, but it is not a good product. But infrastructure is inherently collective regardless, even if it is funded by the big corpos.
So FOSS projects get a solid OS for servers and embedded in appliances, lots of network libraries, web browsers as a platform to deploy services, specialized tooling for big industries. And some hobbyists doing the odd passion projects.
But "regular people" don't ever pay for FOSS. They might use it for free, but even that is often a concession they make.
@eldaking Open Source software certainly thrived in the infrastructure space. I'd have to think more about this.
I was born in 1990.
I grew up through the last days of shareware and the early days of modern free software.
I got to watch the shareware model die, I got to watch the Open Source revolution.
If you weren't there for the shareware days, what I'm about to say might sound impossible, but small software projects released as limited demos with paid upgrades used to handily pay a *good* wage to solo developers or small teams.
There was a lot about the shareware era that wasn't great, don't get me wrong, and there's a lot about the modern free software era that is much better.
I am a huge fan of not paying for my operating system, and *many* of the things I do are only possible because there is a good, community made, gratis and libre software package out there to make it possible.
But I'd gladly pay ten bucks to keep the kdenlive or [insert project here] developers fed.
One time payment, single transaction, let me mail some cash or a check to a PO Box.
My favorite kind of shareware was the "donationware" or "nagware" model.
Give me the full application, and every time I launch it, give me a splash screen that says who I can pay to make the splash screen go away.
It was great! I paid for stuff that I ended up using.
(I've talked about my history with Shareware in the past. I'm not going to get into it again, but I suspect there's a fedithread or a blogpost that you could find about the flea markets I went to as a kid.)
Here: https://retro.social/@ajroach42/113481632282311122
That was two years ago. I haven't re-read it. I wonder how much I agree with.
Two years ago was around the time I noped out of computers more or less completely to start making toys.
To be clear, I don't think the fault for the "I'm not paying for that" mentality lies solely with the Free Software movement.
I think it mostly belongs to The Internet in general, and as much as major corporations pretended to hate it, I think it largely served their interests.
Generosity is harder to find, when hardly anyone is making a living wage.
That right there is probably THE single biggest factor holding us back. Capitalism's social engineering, constantly pushing everyone towards precarity and away from the security it takes to share freely.
I think I disagree with almost every part of that.
Many of the most generous people I've ever known have been struggling to make ends meet, or actively failing to do so.
When I was working in tech, my well off work friends would send me venmo requests for "my portion" of the appetizer that we "shared." These were men making $250k/year or more, many of them worth multiple millions of dollars.
When I was working retail and not making ends meet at all, my other co-workers who also didn't have enough to go around *always* had something to share.
The thing that makes the modern capitalism so insidious is the way it steals your time. The time spent working, the time spent exhausted, the shadow work of self checkout and building your own ikea furniture, the time spent navigating systems designed to trick us.
Simply being in a precarious position doesn't, as a rule, make someone behave selfishly (and simply being in a stable position doesn't make someone behave generously.)
But also! I'm not talking about Generosity, or about people who can't afford to spend money. The time period in question, as discussed in the post you replied to and *at length* over the subsequent half dozen posts, was a period of financial stability in which the only people who could participate in the systems I am discussed had the disposable income to have a high speed internet connection at a time when that was a geographic and economic luxury.
People stopped paying for things because it was significantly easier not to pay for them. Doing the "right" thing was significantly harder than doing the "wrong" thing and, in the vast majority of cases, neither had any material impact on anyone.
( I should unpack that)
Modern computers are very fast.
Modern network connections are very fast.
Sometime in the early 00s, the scales tipped, and suddenly it was easier and faster to download a song off the internet than to go to the store and buy a CD and rip the CD.
Network connections weren't gigabit, and hard drive space was limited, but it's easy to understand why piracy was so commonplace, even without the cost aspect. Even on a naked DSL line, I could download an album (or a podcast, or a mixtape) faster than I could go to the store.
Corporations took a while to catch up, and when they did, they mostly embraced the convenience, which is what most people wanted anyway (hence Netflix and Spotify) and when that convenience starts to disappear (Netflix today) piracy takes an upswing.
Now, the beauty of the Free Software movement is that, if your primary developers decide to do something stupid, you can pack your things and go.
We saw this recently with Audacity and Tenacity.
That gets a lot harder in the Shareware model (but this is not an impossible hurdle to navigate! The important parts of free software can live alongside the important parts of the shareware model.)
@ajroach42 Fucking yes. We seem to be thinking in similar directions.
https://lyk.so/enclaves.html
https://lyk.so/collective.html
https://careful.party