[throwaway50203]

> far more abstract principle of the home PC user being able to recompile their own software stack from source.

I contributed hundreds of hours of work into libraries. They are used by the same companies that sell me locked-down devices that I cannot modify or even inspect. They even invade the my privacy and I cannot disable that.

At least GPL tries to mitigate this problem. MIT turns volunteering into unpaid labor for some billionaires.

Calling it "far more abstract principle" is quite absurd.

These companies try hard to push the narrative that MIT is "practical" and GPL is "political" - guess why.

[pmoriarty]

I release all my code under the AGPL.

However, I'm under no illusion that it'll prevent my unpaid work from being used by billionaires.

Setting aside those people/companies that are willing to violate the AGPL, even the ones that respect it are free to use it as long as they redistribute the source to modifications they make.

Anyone (including billionaires) could do that, and not every business model relies on keeping their source code secret.

[throwaway50203]

> However, I'm under no illusion that it'll prevent my unpaid work from being used by billionaires.

There's plenty of evidence that AGPL is reasonably effective at that. SaaS providers are clearly afraid of it.

> not every business model relies on keeping their source code secret.

I was never talking about secrecy and business model.

It's about allowing end users (including myself) to have some reasonable chances to inspect / update / reflash their devices.

[est31]

Note that the definition of "use" is important here. Many MIT advocates regard shipping parts of MIT licensed code inside a proprietary product as "using" it. And yes, in some ways they are right, but it's definitely not the only meaning of the term.

The GPL has a different usage definition: run and execute. E.g. think of an image manipulation program under the GPL. If you design a game, you run the program to create assets, but none of the created assets contain a trace of the original program any more. The game you ship doesn't either. The GPL doesn't affect you. However, if you tried copying one of the algorithms implemented in the program into your own proprietary competitor, you infringe on the copyright now.

[geofft]

Are you happier with locked-down devices that you cannot modify or even inspect that invade your privacy as long as they don't use your code?

Anyway, I certainly think the MIT license is political - I also believe in the political position of free education for all people, even if they're billionaires (or children of billionaires), and in the political position that I should be taxed to provide these services for people richer than me. I donate to charities that help people who do not repay me. And so forth.

If some billionaires want to profit off of unpaid labor, don't think you can stop it by using the GPL, which relies on the legal system and copyright laws that made them billionaires in the first place. If you sue them for infringing the GPL, you might have a great strategy to win their game, but you're still playing their game in their courts. There's a saying from a totally different context that applies here too, I think: the master's tools will never dismantle the master's house. We've had the GPL for decades now and the result is there's a lot of billionaires whose fortunes were made via GPL'd software like Linux.

If you want to get rid of billionaires, or of freeloading billionaires, or of locked-down devices, or of invasions of privacy, take action against that thing. Do not fall into the trap of "Something must be done, and the GPL is something."

[carapace]

This. I try to remind folks that the whole GPL thing got started when RMS wanted to fix his printer and Xerox said "no" despite having (IIRC) used some of his work to write the printer software he wanted to edit.

[qsort]

> MIT turns volunteering into unpaid labor for some billionaires.

Um... no? The idea behind MIT/ISC and the like is to let other people use your software, which they aren't automatically allowed to do because copyright is broken by default.

There's a lot of software that I would like to use in my day job that I can't touch because it's GPL. This is by design.

> These companies try hard to push the narrative that MIT is "practical" and GPL is "political" - guess why.

I have no sympathy for FAANG, but that happens to be correct. GPL is explicitly political in its intent.

[throwaway50203]

> There's a lot of software that I would like to use in my day job that I can't touch because it's GPL. This is by design.

Good: the license is working as intended.

> I have no sympathy for FAANG, but that happens to be correct. GPL is explicitly political in its intent.

Don't be cheeky. The sentence was [ push the narrative that MIT is "practical" and GPL is "political" ]

The MIT is equally political - just siding with the freeloaders.

[dnautics]

For some people the fundamental problem with the GPL is that it is entirely dependent on the political structure that they believe to be flawed. If you are for copyright abolishment, using the GPL is cynical at best, hypocritical at worst.

[throwaway50203]

> GPL is cynical at best, hypocritical at worst

It's not different from soldiers taking weapons from the enemy they are fighting. A pretty common tactic.

Fighting fire with fire. Where's the hypocrisy?

[dnautics]

Sure, but in war there is at least an exit plan (a way for eventually not having to kill anymore), even if it's shitty and ill-planned. If you are in software freedom for the purposes of abolition and you pick the GPL, you are allying yourself with people who NEVER want to give up copyright, because it lets them "exert control" over the end user, e.g. corporations, so your ally has a very strong real vested interest against what you take as a desired end goal.

What does the endgame look like for an abolitionist?

[2Gkashmiri]

One quick question. GPL forces downstream users to disclose their source code and changes. Does it also force them to push those changes or improvements upstream or not ? As in, I make an improvement to a software and I am selling and also giving source code but am I supposed to push my changes back up or not?

[codeguro2]

The restrictions of the GPL only kick in when you distribute your binaries. You have the freedom to run and modify the source as you wish. The only rule is, "The source code travels with the binaries" - meaning /if/ you choose to distribute it, and you may choose who you distribute it to, that those who receive the binaries must also receive the source code and can take full advantage of the GPL just as you have.

[a1369209993]

It specifically doesn't require that for logistical reasons. See, eg, the "desert island test" from DFSG. If you aren't able to contact upstream to send them your changes, such a requirement would effective forbid you from making changes to the software.

[amw]

MIT ends up just as political, it just does so by assuming society's default politics. It "just so happens" those default politics are great for billionaires.

[crankyoldcrank]

Ugh, they're also great for anyone that wants to feed themselves with their software, whether self employed or not.

God forbid people use a license that makes it easier to turn a profit so they don't have to beg to stay afloat.

[wizzwizz4]

MIT actually makes it harder for people to profit from their software than GPL, because the developer of GPL-licensed code has the exclusive right (by law, as author) to produce proprietary forks of that code; if the code's MIT'd, somebody else could fork, clone the proprietary features and add a few more (preferably going against the original program's general philosophy in doing so, to reduce the chances of those features being introduced into the original), aggressively market, and basically steal the whole project and its user base… then the developer's unemployed, the whole thing's proprietary and some unethical organisation is profiting off the work of another.

The “rolling release” model, where the latest version's proprietary but six months ago's version is libre, simply cannot be done with something like MIT unless you're willing to risk somebody else taking it all for themselves.

[dnautics]

> MIT actually makes it harder for people to profit from their software than GPL, because the developer of GPL-licensed code has the exclusive right

Can you name an IC that profits from their software with the GPL?

By contrast I know of a very well-known IC that profits (handsomely) from their software with an even looser license than the MIT - SQLite.

[cercatrova]

MongoDB, Sentry, Plausible Analytics et al. use GPL, AGPL or even stronger proprietary versions such as SSPL to prevent cloud vendors from using their products. Now imagine they used MIT instead, cloud vendors could rip them off without any contributions back.

This is exactly what happened even when they used AGPL, something that was thought to scare cloud vendors, but instead it wasn't scary enough, so these companies had to invent new proprietary licenses.

[wizzwizz4]

> There's a lot of software that I would like to use in my day job that I can't touch because it's GPL.

But my day job's fine with me using that software. Why? It's not using the law to restrict people's ability to use their computers when that software is installed.