Reforming the free software message | Hacker News Reader

Reforming the free software message

drewdevault.com

54 pointsTomte35 comments

Many of the points in your original post follow from your first point (Reform the leadership, remove RMS)

So much of the FSF's myopic vision and ineffectiveness that you accurately point out stem directly from RMS playing a leading role in the FSF. Why does the FSF spend so much time browbeating people for saying "Linux" instead of "GNU/Linux"? That's entirely because 30 years ago RMS's ego wouldn't let him admit that something he had no part of was becoming popular so he had to make sure he got credit for it. Why does the FSF spend so much energy insisting on stuff like free BIOS? Because RMS thinks anything other than a 100% free ecosystem is unacceptable, so significant resources must be invested into endeavors like that. Enjoy your 12 year old ThinkPad with external Wifi dongle!

If you remove RMS and his cronies, theres a chance the organization could reform into something with reasonable goals. But until then it's just going to be a vehicle for his personal bugbears, which as time goes on become increasingly out of touch with reality.

> Why does the FSF spend so much time browbeating people for saying "Linux" instead of "GNU/Linux"?

It's even worse than that, because Stallman thinks it's fine to refer to it as just "GNU":

The main error is that Linux is not strictly speaking part of the GNU system—whose kernel is GNU Hurd. The version with Linux, we call “GNU/Linux.” It is OK to call it “GNU” when you want to be really short, but it is better to call it “GNU/Linux” so as to give Torvalds some credit.

What a generous person he is, wanting to give "some credit" for the small contributions this little upstart student from Finland did.

[1]: https://www.gnu.org/gnu/incorrect-quotation.html

Am I missing something, or are the only GNU parts of a modern distro just some barebones terminal utilties for copying files?

I'd think the kernel, desktop environment, and other components like wayland, systemd, pipewire, etc are the actual OS while gnu mv is just a user level program.

There's still some pretty critical components that are used by almost all Linux systems: GNU libc, gcc (+ associated stuff, like binutils, gdb, make, etc), coreutils (+ associated stuff, like bash, grep, tar, etc.), and GRUB.

Even systems like Alpine and OpenBSD use some of that.

And there's a bunch of stuff that's not "critical" but is fairly widely used: gettext, gawk, nano, gpg, parted, R, Emacs.

I agree with your broader point that GNU is "just" one component out of many and "it should be GNU/Linux/X11/Perl/..." is an old criticism of "GNU/Linux", but the GNU components are certainly more than "terminal utilities for copying files". Stallman gives GNU too much credit, but you're giving it (far) too little.

GNU does not seem to be acquiring new projects at a great rate though, and does seem to be losing influence. The goal of a functional "GNU operating system" is further away than ever because much of it is focused on creating Unix circa 1989, which is not very "functional" for most people's meaning of that term. One reason for this is that there's not much advantage to being a "GNU project" because it offers very little in material support and risks Stallman coming in and badgering you if you're doing "unethical" things such as merely recognizing the reality that non-free software exists.

(GNU did have a modern desktop environment in GNOME by the way, but Stallman spent 15 years alienating the GNOME developers until they got fed up with him)

You are missing something. Many command-line tools that are famously less bare-bones than on other OSes / POSIX, the compiler, gettext, and libc come to mind.

> That's entirely because 30 years ago RMS's ego wouldn't let him admit that something he had no part of was becoming popular

Just a reminder that Linus was inspired by the existence of Minix and by the GPL. So RMS had an important influence on Linux's success. "Influence" =/= "role", but is generally (and in this case) a "part".

None of this justifies the dogged pursuit of the "GNU/Linux" nonsense , but I do think this is an important clarification.

BTW apart from RMS himself I've never heard a single person refer to "GNU/Linux" except in a joking fashion, and I hang out with a bunch of people you'd think would, if anyone took it seriously.

I just meant that RMS was not a code contributor to Linux, and that his primary motivation for insisting on "GNU/Linux" was ego and jealousy.

Some people do call it GNU/Linux officially, such as the Debian Project, but casually yeah it's not really used so much because its weird and awkward to say. 99% of the world just calls it Linux and that's fine.

Maybe I'm out of the mainstream, but his dogged insistence on it is pathetic. It actually annoys me when I see "GNU/Linux" every time I log into Debian. It feels less like a tribute to the GNU project and more like giving in to his browbeating.

Honestly, if RMS were less of a jerk about it, I'd feel the other way: he and his organization did a lot, and it'd be great of Debian to credit them. But his juvenile bizarre obsession with it, at least for me, is a perfect example of the backfire effect.

I don't really have a horse in this race; the insistence on GNU/Linux seems silly to me and misunderstands branding pretty severely. But to RMS's larger point, a kernel is not an operating system and the Linux userland exists because of the GNU project. This is less true in CURRENT_YEAR when plenty of Linux offerings are entirely GNU-free, but for most of the history of the Linux project, virtually all distributions used GNU libc, gcc, etc, etc, etc.

Don't worry: I wasn't disagreeing with that part of your comment, or actually the comment in general.

RMS, the FSF and the GPL have had a phenomenal impact on software. But apart from RMS's own shortcomings, the result of that phenomenal impact is that his role is diminshed, to a de minimus level.

It's like Tim Berners-Lee: massive impact on the world but by now no real role. Nothing wrong with that (especially when it's such a nice person as TB-L).

> Enjoy your 12 year old ThinkPad with external Wifi dongle!

I'm curious why that Wifi dongle you casually refer to runs firmware that is MIT-licensed and available on Github.

According to Wikipedia, it didn't require any reverse-engineering. So what was the business use-case for FOSS firmware here? And in the category of Wifi chips, why hasn't there been another such use case for modern hardware?

IIRC one of the Linux kernel developers worked for Atheros at the time and was able to convince them internally to release the source code for at least two different WiFi devices of theirs. Atheros were acquired by Qualcomm and that policy is now long gone.

Do you happen to know who the developer was? I'd love to get more info about how that happened.

I believe it was Luis R. Rodriguez, one or more of these emails might work: mcgrof@kernel.org mcgrof@gmail.com mcgrof@do-not-panic.com mcgrof@winlab.rutgers.edu

https://www.do-not-panic.com/

Most of the time, the great visionaries don't make effective leaders. There are exceptions (Jobs, famously---not that he was perfect, but he was capable in that role, charismatic, and visionary).

We all owe RMS a lot. He had a principled vision, stuck to it, and did an enormous amount of work to advance it. But it's pretty obvious that there are two big problems: 1) he's just not a charismatic leader in any way; and 2) like many people, he is not capable of growing/changing his mind.

The FOSS world is different now. We have different challenges and operate in a different environment. That environment is also far more complex, and complex environments are less amenable to the kinds of no-compromise approach that can work in smaller, nascent communities. To Stevie's point, it's hard to take a guy seriously who is using decade-old hardware without WiFi because of "principles." It's a bit like refusing to wear pants because purchasing clothes cuts against your anticapitalist ethos. To some extent, you need to work within the system to reform it.

IMO, FSF needs to set _an example_. We are not going to get people excited about FOSS with nitpicky Usenet-style debates about verbiage and ancient hardware and excommunicating people who dare use an iPhone.

Also, their website desperately needs a redesign. It looks like it's from 1998.

IANAL, etc.

The Copyleft (both the short blurb and the longer detailed) explanation needs to make more explicit about the "distribution" aspect. GPL-v2/v3's copyleft requirement in particular only apply if you distribute the software. If you make changes/incorporate GPL-v2/v3 code but then never distribute that software, you are under no obligation to make those changes or the total incorporated project available. This is exactly why the Affero GPL came about, to close this "loophole", especially for network services (famous case is Amazon building a managed elasticsearch service).

There is also a reformed project for a subset of the GNU project: https://gnu.tools/ A good match, perhaps.

This is a valuable project, executed well.

A suggested edit: the term "copyleft" at https://writefreesoftware.org/learn/licenses is confusing and is falling into disuse. "Reciprocal licensing" is easier to understand and is growing.

Does Drew Devault have a HN handle we can @?

Not sure that "copyleft" is confusing or falling into disuse, I hear people talking about it without confusion pretty often. First time I've heard "reciprocal licensing".

I think "copyleft" is ubiquitous within the tech crowd but the first time I heard the term I was confused. Might be worth asking a few people you know in the real world who are not tech people to see if they knew what either means.

The only people who know the term "copyleft" are already familiar with GNU, the FSF, and the free software movement. That is a tiny pool.

“Reciprocal licensing” is not precise. I think ShareAlike is more descriptive (and consistent with CC licenses).

ShareAlike is also effective.

What we know for sure is that Copyleft is never going to work.

Never heard "ShareAlike" either, and it sounds like a proprietary file sharing feature for phones or something. I just say open source and everyone has always known what I meant.

It's a term from the Creative Commons family of licenses.

Does @ on HN actually work though? I didn't think that was a thing here.

No.

Great example of do not complain, but just do it better. Really like this! Of course made possible with Foss and cc-by-sa work. Quote heard from stalman ‘we should have called it freedom software instead of open source software’

fully agree there is work to do -- I believe that the word "FREE" is problematic in English, and is destructive to the ends espoused thereby.. There is no perfect fit, but "Libre" causes the English-reader to pause, while removing very loaded language available to critics.

Secondly, there is no secret that some cloud-environments are teaching paying customers that "GPL does not apply here"

This is another huge problem with RMS: the nitpicky obsession with language. Although, in fairness, this seems to be an issue with a lot of technically-minded people. Words are not like code; a word can be imperfect but work well enough in context. Code isn't like this, of course. Even if it's obvious from context that you're writing a conditional, using "=" instead of "==" won't work. But language isn't like that; "close enough" can be just fine, and over time, meanings can evolve.

"Open-source" is a term that a lot of people know and use. It's fine. It's not perfect, and its meaning does not perfectly encompass everything that FSF stands for. Who cares. It's good enough. These arcane, pedantic debates about "libre" (not an English word) versus "FOSS" (an acronym that must be learned) versus "free" (a word with a confusing valence of connotative meanings) are off-putting and distracting.

A lot of activists get distracted by this kind of nerd-bait. Somehow, smart people get convinced that if only we can get people using the right words "I call them used not users" then we'll win the argument via some linguistic cheat code or something. It just doesn't work like that. Here's a good discussion of how climate activists can't get out of their own way on natural gas terminology and some evidence for why the entire debate is pointless: https://strongmessagehere.substack.com/p/words-arent-magic

I get what you are saying, but .. this is not "nerd bait" I really mean it.. what is your suggestion then?

Oh, my suggestion is simple: use words people understand and stop arguing over terminology. No goal is achieved browbeating people over terminology. Oh, and whatever you do, don't _make up words_! Talk about inaccessibility.

This community needs to be more open and more welcoming, not more esoteric and closed-off.

yeah - I have been "open and welcoming" for decades, I am not at all sure that is what is missing. I don't like the word FREE in the USA, I think it is doing actual damage.

Let's avoid "browbeating" in any case, agree.

A centrist corporate-friendly reform that no one asked for to one of the last principled institutions in an industry incapable of reflecting on how much it has become the most hideous monster in the world.

I think it's worth actually looking at the link. The four freedoms and copyleft are explicitly in there and the copyleft licenses are at the top.

He's arguing that the messaging is messy, not the message.

Reply on news.ycombinator.com