everything wrong with free software
"obedience breeds foolishness"
*originally posted:* aug 2022
i have left the gnew project, and there doesnt seem to be very much left of it. efforts were made to salvage it, and i wont call this a gnew "post-mortem" as i couldnt really tell you if its dead.
right now, it certainly doesnt look good for gnews future.
but i dont know if it will be picked back up-- my leaving gives them a chance to do whatever they will without me. if gnew managed to flourish, thats great-- free software could really use a success story these days.
ewwfs is largely about whats wrong with free software, as free software is extremely important but so is addressing major problems. failing to address problems has gutted the free software foundation, allowed open source to take over and hurt countless software projects as well. while gnew may have grown to tackle problems such as these, i considered it to also be about how things COULD improve-- not just how they havent.
for now, however, things have not improved very much.
speaking personally, gnew would have been a great place to put all kinds of ideas, from teaching, promoting and simply using homemade software-- rather than the new, very open source ideal where all free software becomes controlled by for-profit companies. you can never say this without pointing out that free software itself was never technically against commercial exploitation-- or rather it came to be this thing where commercial exploitation was ok, as long as the tools used in exploiting users were under a free license. whether you stand against the sort of exploitation that has happened or not, i consider (to address a peripheral issue) non-commercial licenses to be worthless, and they do more to limit the public good than they do to limit companies.
its very true that without the free license, software is non-free and thus a way of exploiting users. this assumes (as i do) that software being free (and you having control of your own computer) is a reasonable expectation-- that people who push non-free software on you are actually letting you down.
but in practice, a free license is not enough by itself. as the fsf focuses increasingly on licenses and marketing to the exclusion of all else, it transforms its own cause into something superficial and extremely lacking-- not in quality, but in its own core goals.
one of the greatest tragedies of free software is that the movement was overshadowed by the organisation. this was a strategic mistake; a movement can last for generations, but organisations come and go. you might be tempted to think of a model where orgs outlast founders and movements outlast orgs, but its not as simple as that. if stallman were not essentially a hostage of the new fsf, he might manage to have a voice and convince people to solve problems the fsf simply cannot.
instead, he is more like john lennon around the time he left the beatles-- lennon was stifled by being a member of the band, and it wasnt so much that he wanted to do everything solo; musicians collaborate all the time, and the beatles separately went on to work with wings, the plastic ono band and various rock superstars that ringo starr enjoyed collaborating with. im personally very fond of what george harrison and lennon did without the beatles, even though i remain a very big fan of the beatles to this day. the brand was nonetheless forcing lennon to be "a beatle", when he wanted to write and play music that wasnt beatles music.
if you join the fsf today, you wont be helping free software nearly as much as youll be a hostage of the fsf and its bullshit. its not the fsf it was, nor is it (nor will it become) the fsf it ought to be. one of the reasons organisations come and go, is they fail. an organisation can potentially outlive its founder, but it can just as easily become the opposite of what its founder intended. regarding the fsf, i think it has clearly become about open source-- what is open source? the cynical, misguided astroturf "movement" to package up free software and transfer it wholesale to the very corporations who wish to control users.
the fsf cant do that AND still be the fsf. but control of the brand means they can call themselves the fsf anyway, even long after they put their own values behind them. i think todays fsf is nearly as cynical as open source, even while they say the words "free software"-- and i dont think its been honest or helpful to users who want to be more free in most of a decade (possibly more). i continue to acknowledge, and once enjoyed, work that came mostly from stallman and the fsf and their efforts.
today, gnu/linux (gcc too) is a corporate juggernaut-- it is being manipulated, twisted, repackaged and rebranded. the "foundations" that get put in charge of many key projects are in turn infiltrated by people from the companies who wish to control (thus mitigate or even negate) free software. any free software organisation that cannot see and admit this is useless to me. gnew, to its credit, was different than this.
i was not the true founder of gnew, nor did i lead it. my role was advisory, and i was proud of my role. it was nice to be part of something that wasnt just me, advocating for real freedom (like we used to) while everyone sells out to github and google. its not that im against developers making money-- if you work for a company, you deserve payment for your labour. if you volunteer, it should be for the benefit of you (if nothing else, you should be able to feel good about it) and hopefully the world. the tradeoff between users who want control of their computing, and companies that want control of users, is something open source doesnt give a shit about at all.
rather, the state of free software today is where volunteers are exploited in a for-profit way, essentially another example of the poor and "unimportant" subsidising the rich and well-connected. basically, everyone is a fucking intern working for a company, rather than a volunteer working for everyones freedom or an employee getting paid to do work. this did not "just happen", rather it is the result of a series of mistakes on the part of the fsf and others.
and this is absolutely the legacy of open source. eric raymond is a fucking idiot who wont, and probably cant tell he sold out stallman and everyone else, and while he thought he was craig mundie and microsofts "worst nightmare", its highly unlikely that either ever worried late at night about the prospect of raymond building the perfect weapon for co-opting free software and then letting it fall into enemy hands. if youre letting this guy plan some revolution, consider that for two decades he has been the greatest (most likely unwitting) ally of his own worst enemy. he is a slave to his ego, and his ego is a fool.
not to put too fine a point on it, but while he claims stallman is a "friend", the very takedown that took down torvalds, tso and stallman was a plan that had been sitting in a desk (or the equivalent thereof) at osi for years. or so ive been told at least, via someone who claims perens said so. while i dont think very much of perens, his open letter on the debian mailing list following his original resignation from osi in 1999 is a must-read. it explains whats been wrong with osi since its first year, and it never got any better since.
regarding gnew, i still think it had real potential. ideally there would have been software development, as that played a very significant role in the gnu project, but the chief gnewsance had the idea of making it "about education".
on two different levels, i liked this idea. on the one hand, making it about education focused on something the fsf always lacked-- the fsf really did not excel at this, instead it hoped to convince schools (taking donations and advice from the b-and-m foundation) to "support free software". thats just not enough to sustain the gnu project. the fsf could have done a lot more in terms of education, but it wanted to delegate that to parties that were barely sympathetic to their cause.
but software development was certainly not forbidden-- there was no mandate to avoid "doing software" as part of gnew. so as a way to hold over until gnew had sufficient forces to become more like a gnu project, this was a reasonable compromise; especially if it was temporary.
still, gnew got several things right that the fsf no longer did. and for that, it had potential.
what went wrong is the chief gnewsance withdrew, when there was not enough of gnew without him to carry things forward. he made some appreciated efforts, but this was simply not enough. trying to repair and salvage gnew without the thing that held it together was an exercise in futility, and if i wanted to promote a movement that has no legs and no one at the helm, i could just spin a sign around for the fsf next to the cash-4-gold guy.
im being frank about this because its the truth of the matter, and if the goal here was to trash talk gnew i could go much further. thats not what i want-- if gnew manages to reboot itself thats great, but even if it doesnt, ill be doing what i can with the ideas i learned while being peripheral to its cause.
everyone involved with gnew in the early days should have given at least 1%-- and to be certain, they did at first. thats not a typo, its a warning to anyone who wants a viable free software (or free computing) organisation. you can do a lot with 1%, but you cant do anything with 0% or something too close to it.
my own contributions were certainly not enough to carry gnew, and i wasnt going to stand in an empty-looking room and call it a meeting. one person is not a movement, even if that one person is doing some of the same things a movement does. but gnew wasnt just one person, rather it became more like that as people withdrew.
it should be noted, with the rest of what im saying here, that people who devote themselves to the launch of something are probably going to put in less effort later, than they did to get things moving in the first place. this is not a sin. when some people withdraw, when others pick up where they left off, thats natural and unavoidable. when people remain involved on paper, but they withdraw part-way, thats unavoidable too. expecting levels of commitment to be total or even constant is not just unrealistic, its unfair.
but withdrawing from a project can lead to procrastination, excuses and even an unrealistic perspective about future involvement. my intervention may have killed the whole thing; but i think its very safe to say, it was going nowhere. do not underestimate the benefits of a good leader-- theres a reason that people trying to destroy a movement target the head of the organisation. for all this pseudo-left-wing bullshit about leaderless this and non-hierarchy that, every major leftist success story had someone at the helm. the difference is that collectively, the workers / volunteers mattered more-- not that they were leaderless.
the leader of an organisation is like the ruby in a laser; if you take the ruby out (yes, there are other designs that use something else-- for this metaphor theyre all rubies though) its not a laser anymore, its just a fancy expensive light tube. when the chief gnewsance was active, there was a relative degree of coherence.
part of this is because he had a lot of good ideas. but the other thing was, we could appeal to him for direction. its difficult, if not next to impossible, to keep an organisation together without that. even if youre full-on marxist about this, you can say "all power to the workers", but you would never transition to a pure and stateless democracy without first transforming the state itself. even in its most modest form, leadership is not optional but a necessary component. perhaps someone can show me a working example to the contrary, but for now the withdrawal of the chief gnewsance was no exception to this rule.
yes, efforts were made to replace him-- with his blessing, i might add.
thats all i have to say in retrospect, other than to talk about what could have been. the ideas that went into gnew are important, and i dont have any plans to give up on those ideas. in that sense at least, i will continue to promote gnew as i continue to promote free software. but whether gnew exists or not, whether the fsf is really the fsf or not, its the things they stood for that im going to focus on instead.
gnew was not the only project i was involved with, and its while i tend to other things that the ideas of gnew and free software are most likely to be useful. i still consider this a setback.
i still write code, and i still talk about free software. i still hope for the future of the movement, though what happened with gnew is an important loss. i personally considered it part of this federation people talk about sometimes, and hopefully that federation will pick up at some point. its not centralised, so gnew could never have been at its centre. but gnew had its own (extremely lightweight and modest) central hub, and without that it would never have gone as far as it did. the position of chief gnewsance was based entirely on merit and potential.
even if you are a node in a (wisely) redundant network of networks, i still recommend you have a leader of SOME kind. i DONT think everything should depend on them, though theyre still useful for the things that depend on them whether anyone (including said leader) wants them to or not. without stallmans leadership for example, the fsf is dying. transfer would have been ideal. but it only counts if you can find someone-- not as good as stallman, and probably not better, either-- but it has to be someone good enough for the role.
for the fsf, thats a tall order. for gnew, that was a tall order as well. by all means, the fsf should have put in some redundancy so that if stallman failed in some way, they could help make up for that. ive been saying that for years; this isnt a reversal on that stance.
but that would have only helped the fsf continue to shine (rather than burn out) while a new ruby was located to turn it into a laser again.
other possible metaphors include the lens of a camera (to focus the light into an actual picture) or the kernel of an os. gnu to be certain, was a more important project than linux. and letting linux take too much power over gnu was a strategic failure, which gnu still suffers from today. (choosing hurd for a design was also a strategic failure, even if it was a cool idea. it would have made a great second kernel, but it was an awful choice for a first one-- based on the false assumption that gnu would not be a race against anything else, solely because it wasnt initially).
even though a nearly impossible kernel like hurd, or a somewhat traitorous kernel like linux can be enough of a problem to cost the project its success, no one argues with the idea that gnu required a kernel. i mean, you can just run one program on the computer at a time-- or you can have each program somehow do the job of the kernel. this is technically possible, whether it would help or be something anyone actually wanted to use for everyday work.