Login
You're viewing the oslo.town public feed.
  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 12:06 PM

    People are like "let's fork Linux", "let's make a new OS", or "let all switch to BoobsBSD" kinda need to chill tho.

    It is a very common misconception of software especially pervasive amongst coders, that you can just whip up any sort of software package at a whim and build something equivalent to the existing, enduring packages.

    Software is a social infrastructure not a product, it is built up over time, it's additive, and accumulative. It slowly accumulates features, edge cases, users, expertise, experts, social purchase, a literature of documentation, help, user/developer's public discussions, and similar.

    Something like an OS or a big app also accumulates addons, apps, drivers, extensions, or whatever else they might be called.

    When you say "let's make a new Linux", it's like saying the metro sucks let's build a new one. Let's use city X's trains to go to our jobs at city Y that we live in. Let's switch to horse drawn carts instead of the metro.

    People would protest if the municipality just decided that they'd trust unpaid teenager interns to drive all public transport vehicles. You wouldn't dream of setting up an entire new transport network because (1) the existing one is YOURS it's a public good, and (2) it's fucking impossible are you insane?

    But when it comes to software it feels like it's this simple and easy trivial thing. It's perhaps the exact same mindset that trusts it to LLMs: it's insubstantial and easy, why worry!? Well cos it's fucking not, that's why.

    💬 7🔄 33⭐ 35

Replies

  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 12:18 PM

    In amongst these knee-jerk reactions, the impulse to fork is the most sensible, as it is akin to an actual protest or strike, but that requires organisation, willpower, commitment, goals, infrastructure, people.

    People seem to think it's easy to fork something like Linux. It's not? It's not a fork when you clone some source code over to another server. GitHub etc have trivialised the word I guess, because their absurd web based patch submission model abuses the word. "Forks" on Github are just wasteful clones.

    An actual fork is an attempt to spin up a new collaborative project. A hostile fork is a form of protest and an attempt at occupying and taking over infrastructure, because SOFTWARE IS INFRASTRUCTURE. Software is not a product. It's not a consumable.

    When you declare (for) a fork you should understand what kind of commitment it is. You should know what you're doing. You should be able to foster or provide the necessary expertise to take over and continue an existing thing that already has momentum and an ecosystem around it.

    Otherwise you're wasting everyone's time.

    💬 1🔄 11⭐ 16
  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 1:03 PM

    I am generally predisposed to giving some benefit of the doubt to people who object a bit too enthusiastically to what I say, but at least try and make it about what I say... especially if you are gonna insult me.

    💬 0🔄 0⭐ 4
  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 1:52 PM

    A pervasive problem amidst tech(ie) people is that they don't understand how the rest of people use computers, and that their ways are exceptional and in the grand schemes of things, unimportant.

    Yes you can make computing work on an incomplete or unsupported platform, I can too. It's fun (or it was, rather...) and somebody has to do it, but that doesn't mean it's the elephant in the room.

    We're no longer in a world where software and computing is an optional extra or an hobby. It's become the main conveyor of bureaucracy and business, it's become an inescapable hydra. I can't book a doctor appt or apply to a job without a functional computer AND phone. I probably just got excluded from a much needed scholarship due to a bug in their MOBILE APP which is the main way of applying to the scholarship and there's no way of handing in a physical application. They're talking of making ATMs refuse cards here and requiring QR codes for withdrawals etc.

    And that matters more than our dotfiles.

    💬 1🔄 1⭐ 3
  • Elda Kingeldaking@weirder.earth
    Oct 14, 2026, 2:20 PM

    @librarysquirrel The entire ideology of "open source" leads in this direction of "just fork it". It is all about the source code (and the license), you need to have the code to change the software, you need permission to change the software... and nothing else.

    What about the governance, the trust, the project members, the documentation, the channels of communication, the partnerships, the third-party support, the technical standards. For such big projects, the staff and the brand and the donors and the conferences. The existing Linux source code is not the bottleneck.

    It treats software as an artifact, a thing that just exists there, you can just have your copy, change it, and it is done. Instead of a project and a piece of infrastructure. You can change the source code! And when the source code changes tomorrow, what do you do. What do you do about the millions of copies of the old source code.

    💬 0🔄 1⭐ 1
  • 💬 1🔄 0⭐ 2
  • d@nny disc@hipsterelectron@circumstances.run
    Oct 14, 2026, 12:52 PM

    @librarysquirrel you are vastly underestimating and in fact fully misreading what it means when people are saying things like "linux is broken". i have read the kernel and found the entire filesystem abstraction to be broken and decided my life's goal should be data integrity after seeing torvalds sign off on a literal crypto backdoor in signature verification in march. i am not going to reproduce fucking linux because it is architected to be difficult to reproduce

    💬 1🔄 0⭐ 0
  • d@nny disc@hipsterelectron@circumstances.run
    Oct 14, 2026, 12:54 PM

    @librarysquirrel literally you're not fucking smart by telling anyone "actually, these people in charge of your open source organizations are have successfully monopolized the work of others like us". do you fucking know who writes linux? it's not fucking linus torvalds.

    💬 1🔄 0⭐ 0
  • d@nny disc@hipsterelectron@circumstances.run
    Oct 14, 2026, 12:55 PM

    @librarysquirrel and go to hell after calling linux the metro system. i'm a former government employee—we do work that fucking lasts the test of time. we build bridges that don't break. do you know what the file cache is in linux and its role in copy dot fail? it's a technology unchanged from UNIX. i'm obsoleting it

    💬 1🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 1
  • Max Lee :ablobcatbongoangry:the_moep@mastodon.de
    Oct 14, 2026, 1:07 PM

    @librarysquirrel
    The key difference beeing thet Linux is not public infrastructure. It's a private project managed in an undemocratic way.

    What we need is actual public software infrastructure which is managed by publically elected bodies, not by some random dudes and some capitalist companies for profit.

    💬 1🔄 0⭐ 2
  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 1:13 PM

    @the_moep When I say public infrastructure I don't mean the current ownership model but the functional, social nature of the thing.

    💬 1🔄 0⭐ 1
  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 1:15 PM

    @the_moep I.e. it's in the public interest despite pervasive private ownership and it's difficult and potentially wasteful to create a parallel one. Maybe a better comparison is the agricultural sector which is thoroughly enclosed

    💬 0🔄 0⭐ 1
  • Max Lee :ablobcatbongoangry:the_moep@mastodon.de
    Oct 14, 2026, 1:17 PM

    @librarysquirrel I believe that the social nature that is desireable stems from the ownership model. The kind of nature that publically owned infrastructure has cannot be replicated by privately managed projects.

    Linux even is the best example for that kind of non-functional social structure because a regime, no matter how benevolent, is still authoritarian.

    💬 0🔄 1⭐ 0
  • MHC Levinmhcat@neuromatch.social
    Oct 14, 2026, 1:20 PM

    @librarysquirrel

    I think that's a straw man.

    The people who say those things are not the people who do those things.

    I celebrate the genuine efforts to step up and solve this new problem via forks and freezes, and I include efforts to make new OSes, often depending on prior art cherry-picked from good projects. They take a lot of time and effort and will not be consumer grade (or "social infrastructure" grade) any time soon if ever. But many of us who care about this new problem will be happy with less complex, less complete software long before it runs the phones and the TVs and the websites, as we were when we became enthusiasts in the 80s and 90s and beyond.

    I see this new problem (slop etc) as a catalyst. The re-shaping of tech as a tool of hegemony has always been bad, but this new problem has radicalized a new generation of people. Some will never make the connection, but others already see this as an extension, albeit a fairly huge one, of capital attempting to discipline us and assert ever more decisive hegemony. It's always good to oppose that, and to work to create alternatives. Dual power is the only way any of this ends well - no mass movement can successfully overturn the status quo using imagination alone.

    So yes, many noisy fools underestimate the work, or attempt to rabble rouse based on essentially nothing of substance. But others do the work, and yet more pull it into recognizable, usable shapes. Eventually enough normal people demonstrate that you can actually live that way, and for everything valuable you lose, you also gain something you lacked previously.

    💬 0🔄 1⭐ 0
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 1:25 PM

    @librarysquirrel There's a fair amount of Linux use-case BSD doesn't cover, but there's also a lot of overlap. These kinds of shifts happen when people start to move what they can, play with the "new" tech, etc.
    The tech rags used breathlessly insist that Novel Netware wasn't going anywhere, and no one is going to switch away from it in favor of other things. Things change.
    So far I have not been telling people "yes, move to BSD", only trying to support them where they are already experimenting with it or thinking about it, but pretending like there aren't isn't already a stable, well supported *BSD ecosystem with most of the software most people want, that I take exception to. BSD was a stable, complete, free Unix-like OS before Linux even existed.

    💬 1🔄 0⭐ 0
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 1:29 PM

    @librarysquirrel Yes, these are *social* institutions. That's a good and valid point. The collective and individual BSD social institutions appreciate support of all kinds.
    Most likely what will happen is the BSD ecosystem (there's more cooperation and sharing than competition between FreeBSD, NetBSD, OpenBSD, DragonflyBSD, etc) will just start to get more support and more users, as Microsoft and AWS continue to assimilate Linux. It'll be slow, over time, as people want to, and are able. Painting this as a binary that everyone does or doesn't do isn't good framing.

    💬 1🔄 1⭐ 0
  • 💬 0🔄 0⭐ 0
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 1:56 PM

    @librarysquirrel

    > People are like ... "let all switch to BoobsBSD" kinda need to chill tho.

    Oh?

    > whip up any sort of software package at a whim

    Packages already there.

    > Something like an OS or a big app also accumulates addons, apps, drivers, extensions

    Apps are there. Device drivers are there for the most part but more is better. BSD has certainly been accumulating fixes for a long time.

    > like saying the metro sucks let's build a new one

    People who *want* to use BSD instead might be more like continuing to use the metro after private equity bought the bus system.

    > protest if the municipality just decided that they'd trust unpaid teenager interns to drive all public transport vehicles.

    There are a lot of teenagers I'd trust to drive the metro (furries, mostly). Way more so than LLMs. Every time I try to walk in SF, I almost get killed by a self-driving car, Waymo or otherwise.

    > But when it comes to software it feels like it's this simple and easy trivial thing.

    When BSD has the package, yes, it takes time to move data over and re-install all of your packages and set everything up again how you like it. It's not trivial. But when the packages and drivers are there, it isn't intractable either.

    > perhaps the exact same mindset that trusts it to LLMs: it's insubstantial and easy, why worry!?

    I can appreciate the idea that software isn't free in the sense of community, fixes, labor, etc aren't instant and don't come out of thin air, but I reject this comparison too.

    > That's not the point nor the topic.

    I went through those point by point. Did I still miss the point or the topic?

    > I am not slighting BSDs, they're great OSes.

    Thank you for that! I've been using BSD since Ultrix and AIX 2. My feelings might be closer to "software always has a cost even when it's free" and BSD is just one possible navigation of that.

    > "let's make a new OS"

    redox-os.org/ looks cool and might have a lot of potential. I'm not an OS purist =)

    Seriously tho, I don't mean to make light (too much so at least) of the absolutely incredible amount of work that went in to Linux. Having your project get taken over then maligned...

    💬 1🔄 1⭐ 0
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 2:04 PM

    @librarysquirrel is terrible and that shouldn't get lost in this. There's a human factor. I'm sorry for everyone caught in the middle of that (including you, if that's the case). I think your post was probably in good faith and I understand this is an emotional topic too, so I don't want to argue it too much. People need to make their own decisions... users, developers, etc. It's fundamentally a rotten situation. Linux isn't usually my first pick for personal use, but I never wished for anything bad to happen to it or the community.

    💬 0🔄 0⭐ 0
  • Göktuğ Kayaalplibrarysquirrel@sunny.garden
    Oct 14, 2026, 2:18 PM

    @scrottie Yes, alas this is still besides the point. The point being, what if tomorrow NetBSD goes slop? We can't forever switch to the next best thing, or start from scratch, leaving the vast majority of people in the cold for whom that's not an option. That's what I am talking about, the public, and for their sake we should recognise software as social infrastructure, and don't let private individuals fuck it up with greed or stupidity, much like how we don't let tobacco companies advertise or how we wouldn't let Cadbury's put bleach in their wares and wouldn't say "well just buy Hershey's then" in response.

    You're telling me BSDs are workable. They are indeed, no objection. I quite enjoyed FreeBSD on my desktop (which was a shitty laptop) about a decade ago now. Emacs, vtwm, Ctrl+T, fun stuff. It's probably much better now.

    💬 0🔄 0⭐ 1
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 2:40 PM

    @librarysquirrel

    > We can't forever switch to the next best thing, or start from scratch, leaving the vast majority of people in the cold for whom that's not an option

    Ok, thank you for making your point explicit.

    Kinda yes and no... charge winds up happening eventually. Trying to stop it eventually winds up being impossible. Microsoft/Novell/SCO/the Business Software Association tried *hard* to stop Linux from happening, but it's good that Linux did happen. I used Novell as an example. It's incredibly good that Novell went away. Sometimes change is bad and is forced on people, leading to other change. Conflict between people want to extend 386BSD and the couple who owned 386BSD is part of ancient x86 BSD lore.

    Agreed that people should fight the Microsoft-AWS take-over of Linux and the enslopification of Linux.

    There's a lot in Linux that can't easily be moved... Steam/Proton/winehq/etc has some work on FreeBSD but getting that to 100% (which is already shy of 100% of Windows support on Linux) would be a big project... but... and I'm not just arguing for BSD for BSD's sake (way too cynical for that myself)... one way to resist the Microsoft-AWS takeover of Linux is to make it a less valuable target. When Oracle took over VirtualBox, that hurt a lot of people, but things like qemu and KVM softened the blow. Oracle wasn't able to complete destroy free virtual machine software. Likewise with the many things Google has bought and destroyed... having many alternatives prevents complete destruction of free alternatives.

    Not arguing for this and it doesn't make sense, but in an extreme case, if everyone did switch to some BSD overnight, Microsoft-AWS would lose any incentive in embrace-extending-extinguishing Linux.

    imo, fight like hell for Linux, but "be like water" too.

    Random story... I was for a while out of love for shared hosting and hate of silo'd "cloud" trying to keep a local copy of OpenSSI Linux patched with backported security fixes. OpenSSI is the "nonstop" tech that went way back in commercial Unix and went through HP before being open sourced... it lets you build a single system image cluster without NUMA...

    💬 0🔄 0⭐ 0
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 2:42 PM

    @librarysquirrel with load balancing, distributed filesystems, and a ton of awesome shit. I didn't want that to get lost. It's now unusably out of date and that project never gained traction (no one wanted to use my crazy hosting thing where the shared host had a large number of CPUs available). I wish that were mainline Linux. But at least there's NUMA. So I hate it when cool shit gets lost. An incredible amount of work went in to that back in the day.

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • scrottie (he/him/them)scrottie@bsd.network
    Oct 14, 2026, 2:52 PM

    @librarysquirrel Gah. There was one point where the NetBSD Project tried to hijack the NetBSD Foundation (or vice versa, I forget which is which), where essentially vendor stuff tried to take over the code and development, and that was a battle. Fighting private enclosure of the commons is a never ending battle and you're right that moving tech doesn't automatically solve that.

    💬 0🔄 1⭐ 0
  • anomaly 🌹anomaly@jorts.horse
    Oct 14, 2026, 1:34 PM

    @librarysquirrel it's because it's so easy to fork something initially - just copy it! But the difficulty comes the moment after, when you have to start maintaining it. Fixing security issues, supporting new technologies, updating libraries that are also constantly changing, governing your contributors, etc. Now all of a sudden it's a messy human organization.

    💬 0🔄 0⭐ 0