Due to previous misunderstandings, I will begin this article by stating its premises.
"It's my personal belief that premises do not change during an argument; an argument with different premises is actually a different argument."Not everyone abides by this, and many debates are dishonest or incomplete for it -- too often intentionally; people will tease you with the possibility of winning their agreement on something, but as long as they are still arguing, the premises have not changed.
Here are the premises of my article:
1. Political is more important than Apolitical.
2. Free Software is more important than Open Source.
3. Individuals have a right (even if it's foolish) to be apolitical.
4. Those individuals will likely be exploited anyway (for political reasons).
5. Copyleft is important -- as long as it works.
6. The ultimate goal of Free Software is for all software to be free.
7. The ultimate goal of Open Source is to co-opt Free Software.
The goal of this article is NOT to dismiss copyleft. My position on that continues to be that it is probably vital to (at an absolute Minimum) flagship software projects, such as the GNU Project and other large, prominent Free Software applications and suites.
"The funny thing about obfuscated source code is that it is tricky to define, but people "know it when they see it"."All software should be free -- and the specifics matter. The specifics are outlined in the Free Software Definition, summarised by the Four Freedoms stated in that definition, and I would add the de facto freedom written by Peter Boughton: "the freedom to NOT run the software" made possible by a reasonable amount of modularity.
This was a de facto freedom, because a reasonable level of modularity existed when the GNU Project was being created. Now that it too is under attack, I think we should consider this freedom more explicitly -- I place trying to tie together too many formerly-separate things (a la systemd) up there with obfuscated source code, which the Four Freedoms discourage or even disallow.
Notice I say "tie" together -- a true aggregate or library of modules that do not have gratuitous amounts of interdependency (Steve Litt) would not as likely be a problem. The latter is traditionally what a distro is -- an aggregate of reasonably modular components, from which you can reasonably easily remove or add what you want. That was a good thing, which Red Hat and IBM have tried to destroy. What do they do once they've destroyed it? Move on to the next thing.
"My argument is already that copyleft is better for freedom, and I do not intend to abandon that premise."The funny thing about obfuscated source code is that it is tricky to define, but people "know it when they see it". People write utilities (not always for nefarious purposes -- but for compression and reducing bandwidth costs or download time) specifically to "obfuscate code". This is not a problem, so long as unobfuscated source code is also available.
As to whether people have a right to create permissively licensed software, I think they absolutely do. The matter of permissive vs. copyleft is about the effectiveness of the movement and its goals. You (normally) have identical rights with permissively licensed software and copylefted software. The latter adds a way to protect those rights in future versions. But it is very important to know that permissively licensed software is still Free Software by definition -- it is Free Software with less protection.
My argument is already that copyleft is better for freedom, and I do not intend to abandon that premise. Outside of this argument, however, I still think it would be wise to have a taskforce (my concept of a Free Software taskforce is rarely if ever belonging to a single organisation or authority, but rather would ideally exist voluntarily along something akin to the THRIVE guidelines) in charge of determining whether copyleft remains an effective strategy. Copyleft is basically a security mechanism for software freedom, and in the security world practice matters just as much as theory.
"The difference between Free Software and Open Source on this matter, is that Free Software tends to push for copyleft even though it has made exceptions -- while Open Source tends to push for permissive licenses even though it has made exceptions."If someone manages to (for example), turn copyleft against itself not unlike copyleft turned copyright monopoly against itself -- even copyleft proponents would benefit from knowing when it helps and when it it doesn't. This flies in the face of faith-based activism, though the FSF says there are times (for very small programs for example) where copyleft's benefits may outweigh its costs -- and it is not worthwhile.
My premise remains that it is better (protection) overall, and I believe it is. That is not altered by examining the circumstances of when it is effective -- I still believe that will support my premise. Many safety devices (I am wary to mention a couple that come to mind) are better overall, but in certain circumstances they are actually safer left unused.
An ideal strategy becomes to use them most of the time, and when reasonable to avoid them in the exceptions where they are unhelpful or more dangerous. This is a scientific approach, that as its very cause applies only when it is likely to help.
The difference between Free Software and Open Source on this matter, is that Free Software tends to push for copyleft even though it has made exceptions -- while Open Source tends to push for permissive licenses even though it has made exceptions. The exceptions (from either side) are not the real problem; it is that Open Source tries to increase exceptions until permissive becomes the rule and Free Software is vulnerable.
Thanks to Techrights (and the Halloween documents) we have evidence dating back 20 years to show this was a deliberate strategy, and more recently we (that is, the public) have leaks that show examples of this strategy being used against Free Software from the past few years.
Addressing this constructively and assertively (but with emphasis as usual on actually being certain it is addressed) is part of what we must do to make software free.
"...Open Source continues to be taken over (corrupted) by sponsors."While some people will remain apolitical, we must continue to stand for freedom. If we do not, we know we will lose further ground. When nobody advocates, people lose. Even when we do advocate, we sometimes lose anyway.
And we know there is pressure from the higher (organisational and sponsor) levels of Open Source to encourage people not to advocate -- unless this is part of some wild, Robert-Anton-Wilsonesque plot to reverse psychology people into advocating more (it doesn't seem to be working out that way) the most likely reason for this is that Open Source continues to be taken over (corrupted) by sponsors.
I have also argued that this is happening to Free Software, and the leak mentioned a few paragraphs earlier is a prime example. GNU Maintainers using GitHub is another prime example.
"...there’s no way to avoid this danger. But that’s no reason to put our head inside Microsoft’s jaws.”
-- rms in a reference made years ago to Mono, though it could just as easily be be a warning to the new FSF president about his use of Microsoft GitHub.
"Addressing this constructively and assertively (but with emphasis as usual on actually being certain it is addressed) is another part of what we must do to make software free -- and it is very clear that the FSF and its new president (along with its new board member) will NOT do so."The control of users (Mr. Knauth, a very useful and informative way to read "control of users" is "lack of freedom", which your organisation claims to still fight and stand against for us) that GitHub helps Microsoft to gain is so worthwhile to them, that even before paying billions of dollars for it they spent years trying to control (bribe) the media around the future acquisition.
Addressing this constructively and assertively (but with emphasis as usual on actually being certain it is addressed) is another part of what we must do to make software free -- and it is very clear that the FSF and its new president (along with its new board member) will NOT do so.
Apart from watchdogging copyleft (as it is co-opted by SFC sponsors and the like, who will continue to try to turn it from a defensive feature of Free Software into a weapon against it) and trying to drag the movement out of the "Jaws of Microsoft", what else can be done?
"...I do not think we are far away from a situation where Free Software is so co-opted by Open Source that it barely exists."Again, the goal of Free Software is for all software to be free. I think that is a fine goal, but as a librarian at heart I know there will be historical value in preserving some non-free software (if only so it can be legally / cleanroom reverse-engineered or rewritten, as GNU rewrote UNIX) for use in emulators / museums and archives. This would need to be WITHOUT DRM, I might add, or with broken DRM that we would work to legally break for this purpose. That in turn may require fighting legal battles against laws like the DMCA or at a minimum, fighting for exceptions.
Though I still (as I have long advocated) propose we try to make free versions of more things whenever possible: distros, programs and projects. Hyperbola is working on a free version of OpenBSD (they are actually starting from another effort to do the very same thing) and I think this sort of project is exactly what we need.
But speaking broadly, I do not think we are far away from a situation where Free Software is so co-opted by Open Source that it barely exists. (We may already be in that situation, which is why I say the movement is dying).
I do not advocate becoming Open Source instead; Open Source is where I came from and what I put behind more than 10 years ago -- I now call Open Source "a corporate scam." It was a weapon used against Free Software. It continues to fool many people, as it fooled Perens himself.
At which point it became that weapon, I generally try to find the point at which Open Source became corrupted. I've never managed -- already one year into OSI, Bruce Perens had protested what OSI had turned into. The term "Open Source" in this context was coined in February of 1998, the year (the very month) that OSI was founded. The first Halloween document was from later that year. Perens resigned the following February, just 12 months after he helped found it.
"Monopoly isn't freedom, and the trajectory outlined in the Halloween documents (as a strategy at Microsoft for fighting the "threat" of Free Software) including the false "Loves Linux" campaign, the takeover of GitHub, and the attacks on rms."Since then OSI has been further and further co-opted by monpolistic companies. Monopoly isn't freedom, and the trajectory outlined in the Halloween documents (as a strategy at Microsoft for fighting the "threat" of Free Software) including the false "Loves Linux" campaign, the takeover of GitHub, and the attacks on rms.
As with Debian, the goals of the thing rms created (OSI defines "Open Source" via the OSD, which is based on the Debian Free Software Guidelines and written by the same author) were betrayed by people close to the founder. What happened at OSI? If I knew, I would tell you.
What I do believe is that for most of its existence at least, OSI has been used against our freedom, not for it. It has replaced activism with shallow promotion, and this pattern now affects FSFE and the FSF as well. It took longer to break Free Software than Open Source, if the latter was ever not corrupt.
I actually want to believe there was a point in history (however brief) when Open Source was actually a separate thing from monopolies like Microsoft. The commentary on the Halloween documents, written by the other OSI co-founder Eric S. Raymond, do not seem like a plan for ESR to betray us, even if he did later on. (OSI plotted long ago to attack rms, not entirely unlike the way he was finally ousted nearly a year ago. This plot was acknowledged by Perens).
I doubt we will ever prove what really happened with OSI, though the leaks coming out lately make me think it's too early to say that's impossible. Meanwhile, a few projects that called themselves "Open" and even used the phrases "open source code" or "open" referring specifically to the availability of the source code, do predate OSI and the coinage of the term by Christine Peterson in 1998. These include Caldera (yes, the infamous SCO -- though McBride was not CEO yet) OpenDOS (from 1996?) and OpenBSD from 1995.
"I doubt we will ever prove what really happened with OSI, though the leaks coming out lately make me think it's too early to say that's impossible."Whether there was ever a brief (and historically fascinating, and possibly increasingly relevant) stage where Open Source was not something corrupt and monopolistic, I still believe it "turned into" that fairly quickly. Since I never manage to convince myself that it was planned that way all along by ESR (let alone Perens) I can only conclude that ESR was compromised by his ego and led by the nose. He does not fight what happened, though in many ways neither does Linus (who took part in the scam and then had the same tactics turned around on him when he was no longer useful) or even rms. This whole movement was started by rms, and I do not think he abandoned it. I simply think he was (we were) beaten.
Organisationally I continue to have the most hope for Techrights, FACIL and anything (truly and meaningfully) like them -- which at the moment is not a long list. Project-wise I have the most hope for Hyperbola, and I am now running and getting more familiar with the operating system that Hyperbola is based on (OpenBSD). I still run Tiny Core GNU/Linux as well.
What I think we should do with both of these (and not only these) is work to remove non-free software and ALSO work to remove software that is GitHub-encumbered. For FreeBSD, their package management and several other organisational functions now depend on GitHub. (So, very unfortunately, does Daniel Pocock for his website; at least Techrights is republishing some of his articles so that we don't need to opt-in to Microsoft's surveillance on those pages -- of not only users, but of how many visits each page gets, of Daniel as an author, etc.)
Of course no system is GitHub-free now, I've looked into the possibility for months and months; OpenBSD does use "less" for example, which is written by Mark Nudelman and he still maintains it on GitHub.
"Of course no system is GitHub-free now, I've looked into the possibility for months and months; OpenBSD does use "less" for example, which is written by Mark Nudelman and he still maintains it on GitHub."I think the most "recent" version of less that OpenBSD includes in base is from 4 years ago, before the Microsoft acquisition, but if they ever update (maybe they will just do their own work on it, they already respond locally to disclosed bugs and vulnerabilities) but OpenBSD could be one of the least GitHub-encumbered operating systems left. They have (unlike other BSDs) a strict policy against GitHub pull requests -- and unlike some projects, they seem to really mean it.
Self-hosting is a security benefit, particularly for large projects (if you don't know what you're doing, self-hosting is a security problem -- though giant networks-effects-exploiting repos like GitHub become a major target and invitation). As noted recently, Microsoft tends to abandon users and maintenance once they have successfully moved into a territory, just as IBM is doing right now with Red Hat and especially Fedora; using GitHub for critical project infrastructure is a huge mistake. But at least OpenBSD (less FreeBSD, less NomadBSD, less even HardenedBSD) gets this.
If you aren't going to (or don't have the skills to) help Hyperbola directly, but still want to work on liberating something (and you're tired of chasing after GNU/Linux while the latter especially marches headlong towards Microsoft), then here are some other things you could do:
1. Hard fork GNU less, and maintain a version that is not on GitHub. You might need to rename it, but not necessarily. If it is reasonably compatible, you could offer it as a downloadable replacement. Otherwise you could offer it as an alternative.
2. Find other people to work on Hyperbola, or help them in some other way.
"If the goal of all software is to be free, and the effect of making copylefted software permissive is to be less free, and we are too considerate to do the opposite, then Free Software itself erodes in a single direction; Open Source co-opts it and we do nothing to get it back."3. Fork a permissively-licensed program and maintain it as something under GPL 3 or later. The original will still be permissive, your changes will be GPL 3. If someone with sufficient experience and expertise in the subject can refute this, by all means do so. But it is essentially what rms did to certain projects when he created the GNU Project.
The latter strategy (despite contributing to the creation of the GNU Project itself) was once discouraged by rms (presumably as the GNU Project already exists, so now we don't need to do more of it) but Open Source continues to adopt or rework copylefted software into permissively-licensed software; such as the GitHub-based project to redo binutils in Rust -- permissively licensed and dependent on a language which itself depends on GitHub.
If the goal of all software is to be free, and the effect of making copylefted software permissive is to be less free, and we are too considerate to do the opposite, then Free Software itself erodes in a single direction; Open Source co-opts it and we do nothing to get it back.
I'm not of course, advocating that we stoop to their level. Open Source is backed by bribery, a corrupt press, and a mountain of lies and doublespeak. When they say that "Open Source and Free Software are basically the same thing" because the software licenses overlap, what they're doing is what Open Source always does: bait-and-switch, conflate and co-opt; focus on the software, NOT on freedom. Use freedom only as lip service, all while shoving everything further into the gaping maw of monopolies (...Mr. Knauth).
"Open Source is backed by bribery, a corrupt press, and a mountain of lies and doublespeak."The GNU Project in particular, is adopting and applying the same strategy to oust rms (but along with him, everyone that truly cares about your freedom) that LibrePlanet used in 2018, that GNOME and MIT used in 2019, and really that OSI planned many years ago. Everything "Free" is being corrupted using the same tactics that turned OSI into basically a propaganda wing of Microsoft and IBM.
If that's what you call "Free Software", it really IS just Open Source -- because its politics ignore freedom, and only mention it in lip service.
Open Source is essentially the erosion of Free Software. If you care about the latter, you will (unlike the new FSF) fight this erosion. The three biggest weapons of Open Source and its bait-and-switch scam are (in order of historical relevance) are Lies, Bribery and GitHub.
The best weapons you have against it are History, Autonomy and (frankly) a stoic, truer love for the four (or five) freedoms. Modularity is an example of autonomy in action, or as a virtue.
"Open Source is essentially the erosion of Free Software. If you care about the latter, you will (unlike the new FSF) fight this erosion. The three biggest weapons of Open Source and its bait-and-switch scam are (in order of historical relevance) are Lies, Bribery and GitHub."As to OpenBSD, they were Open before they were corporate -- they were "open" before OSI existed. The software (I do not speak only of the software, nor do I advocate you do so) platform probably gives you as much autonomy as any at this point -- all it needs is the advocacy for freedom to go with it. If there is firmware that does not respect your freedom in OpenBSD, that's a perfect example of what we need Hyperbola for -- to remove the non-free parts.
By comparison, FreeBSD, apart from making use of GitHub (increasingly it would seem) also accepts code written under NDAs -- the very non-free plague that trapped code rms wanted to use, even before copyright legally applied to software (not before 1980 in the United States). I am not against someone making a fully-free version of FreeBSD, but it would be far less practical to try (you'll need an army) compared to OpenBSD. The Hyperbola team (their track record is impressive) made the right choice there.
But isn't it permissive? Yes, all BSD kernels that I know of (certainly the mainstream ones) are permissively licensed. The kernel that would become the basis of Hurd was also.
Linux had an advantage of being GPL licensed. This suffered two setbacks, neither of which are now resolved:
1. Linux (the kernel) is GPL 2 only, and the GNU Project favours GPL 3 "or later" because they want to be able to fight Tivoisation
2. Linux (the Foundation) is essentially run by Microsoft, as is Torvalds himself.
Through lobbyists, Microsoft fought GPL3 and in particular, worked to keep Linux (possibly also Busybox) from adopting a version of the GPL that would stop abuse by DRM.
Even if Linux were GPL3 or later, (which might help to make Microsoft and certainly Apple less interested in it -- Apple is not even interested in GPL 2) Microsoft would still effectively run the Foundation (unless they lost interest? But they've spent nearly a quarter of a century so far, plus billions and billions of dollars without giving up on co-opting it) and there is nobody to fork Linux anyway--
"Autonomy is key to keeping Free Software alive -- kernels we can actually fork, components we can actually remove (in practice that is, not just because Mr. Poettering says so) and source code we can actually change -- because that's one of the four freedoms."If there were, we could try to do the same: make a free version of Linux. No, linux-libre doesn't count. I most certainly asked, several times! If Oliva says Linux can be forked, great. That's not what he said to me about it. Would that he were mistaken, or that I misunderstood. That would be wonderful news.
BSD has been forked several times -- often by small teams. Autonomy is key to keeping Free Software alive -- kernels we can actually fork, components we can actually remove (in practice that is, not just because Mr. Poettering says so) and source code we can actually change -- because that's one of the four freedoms. While systemd isn't obfuscated source, it is the next worst thing (by design, and even by gloating).
And make no mistake -- OpenBSD doesn't care about our politics. Not at all, they don't even want our GPL code -- but they come right out and say what they stand for and what they care about, and what they want. There is no bait-and-switch, just you get what it says on the tin. What you do with that is up to you. What Hyperbola does with it is probably the best we possibly can do, going forward.
Linus never cared about our politics either, and that didn't stop us from spending 20 years saying "GNU/Linux". If you want a copylefted operating system that is LESS co-opted, that we actually have ANY say over the development of, I believe it will be GNU/BSD (Or HyperbolaBSD, though I really hope Hyperbola won't be the only example). You can start by compiling GNU programs for OpenBSD (I believe they have already compiled GCC for you, even if they lean towards Clang).
"Don't be seduced by monopolies."Given what OpenBSD contributes, and what Hyperbola stands for, I am willing to stand for each. Note that Hyperbola has done more than any other project to convince me to try BSD in the first place -- I think right now there is more shilling and corporate treachery going on at the FSF than the OpenBSD Foundation, but 25 years from now who knows? I used to support Debian, and I wish I'd known how corrupt they would turn out to be. This is the sort of problem you get with a single flagship, don't you know? Don't be seduced by monopolies.
Meanwhile, gaslighting is a staple of what we now call Open Source, and it even gaslit its own founders, Perens and ultimately ESR. The FSF's actions against rms weren't better -- they did to rms and to use what Open Source did to everyone else (including Torvalds). This is a two-decade-long anti-pattern, and it's nothing to with freedom.
We already KNOW what freedom is. We have lots of things we could be doing, to make more freedom happen while other corrupt entities fight us.
We don't need a crusade -- We need a Free Software movement.
But it will be up to you, not them. They will not (except a few, like Hyperbola) help you take permissively-licensed software and add copylefted extensions and updates (which you ARE allowed to do. If you can take permissive software and turn it into non-free software, you can also take it and turn it into software with all additions under copyleft).
They will not (particularly the FSF and its new president) fight to keep you from the the Jaws of Microsoft. On the contrary, they will continue to shove the GNU head into it. I remain in favour of a rescue.
"If OSI is one example of an organisation turning against its own mission in just a years' time, it's definitely not the only one."But the FSF is not a movement, it is a 501(c)3 non-profit corporation which is doing exactly what most 501(c)3 non-profit corporations tend to eventually do -- sell out, chase sponsors, abandon values and increase "messaging" several-fold (while doing the exact opposite of what they used to advocate -- or simply do not much of substance at all). They might throw a little party now and then for peon-level sponsors, but they don't care about you. (Even Microsoft threw a little party for protesters who Marched onto the Microsoft campus to demand OEM install refunds. Why not? They weren't going to give them a refund).
This is a common pattern for such organisations. They have a usefulness, sometimes lasting many years, then over time they typically become a giant turd. With loads of help from corporate manipulators, OSI did that A LOT faster. If OSI is one example of an organisation turning against its own mission in just a years' time, it's definitely not the only one. But in fairness, OSI's mission really sucked: "Sell Free Software to Corporations". "What, you mean exclusively?" "Eh, you worry about the details!"
It is a testament to rms as a person that the FSF did not do this a lot sooner.
You cannot follow any leader who (now) does nothing; you can be that person or just keep waiting, while the cause falls apart. You can also support an organisation that is All Talk, but which doesn't even speak to (let alone do anything to stand against) serious threats. Put another way -- if Geoffrey Knauth said as rms did, that using GitHub is a TERRIBLE IDEA, even if while still continuing to use it himself, even that would be a step Forward from where the FSF is today.
"RMS is probably the least corrupt individual in the history of Free Software (of course he founded it, but not all founders care half as much as he does) but he still isn't going to save it at this point."Either way, it's up to you. RMS is probably the least corrupt individual in the history of Free Software (of course he founded it, but not all founders care half as much as he does) but he still isn't going to save it at this point. The FSF is just too compromised. But as always I welcome them to prove me wrong. Throw out the backstabbers and the people who defrauded you as well as lied to us (you won't) throw out Microsoft GitHub (you won't) and actually fight for what you say you do.
That isn't how this works, though. Corrupted organisations don't come back, they just keep chasing money and (at best) talking the talk, until nobody cares about them anymore. Whether that's the future of Free Software or just Open Source and the FSF -- that's entirely up to (as the Gates-Foundation-funded "public" radio puts it) "listeners like you." If you don't have at least 100 grand to throw around, you can be fairly certain you haven't paid the FSF enough for them to listen.
"The Office" has so many parallels now, from the time that they had two regional managers for the same region, to the time they brought in someone from the steel company to sell paper, to the time that the State Senator dragging a Latino employee up on stage to show how diverse he is -- okay, that's more like IBM, but what we are getting at is tokenism here -- there's nothing honest about tokenism.
"If you think Torvalds wasn't cheap, divide the salary of people paid to develop the kernel over the number of unpaid people who made it a reality. "If you aren't a developer, you are still a user -- and Free Software should be fighting for your rights, but they don't. Real freedom is for everybody, and if you join the FSF, you will be a token user: someone to be paraded on stage so they can do fundraising -- not someone whose well-being matters to them. And you can rest assured that Open Source doesn't care about you either.
If you develop, they'll use you a bit differently -- as a corporate mouthpiece and cheap labour. If you think Torvalds wasn't cheap, divide the salary of people paid to develop the kernel over the number of unpaid people who made it a reality.
Anybody exploiting the kernel now got it basically at zero cost compared to a traditional project, and the cost of Microsoft's "acquisition" of the Linux Foundation was also dirt-cheap, compared to what such things usually cost. Although Microsoft will occasionally poach developers, on a 20-year scale it's much cheaper to just bribe them occasionally.
Rather than believing (as ESR did) that you can change that from the inside, do as Mako Hill suggested at LibrePlanet -- and change it a bit further away from the corruption, not by climbing into its maw. What do you hope to achieve by getting eaten -- causing some indigestion?
"Rather than believing (as ESR did) that you can change that from the inside, do as Mako Hill suggested at LibrePlanet -- and change it a bit further away from the corruption, not by climbing into its maw."This scam will continue to do to us what it did to rms, Torvalds (ehhh) and everybody else who ever got paid to shill or code for Microsoft. (People still developing on GitHub are effectively doing both). And it will continue to turn Free Software into a glitzy corporation-worshipping shell with no substance, which no longer cares about its promises. It just sells people out.
There are still a few people (but sadly very few groups) doing this for the right reasons. THOSE REASONS, like User Freedom (not subsidised labour for corporations) ought to be our priorities.
But anyway, Long live rms, long live Free Software (even if it dies for a while and is eventually dragged back into reality and existence) and Happy Hacking. ⬆