ML @ml@qoto.org

@wistahe I don't know if it's unusual enough, but I thought this page talking about the pre-web internet was interesting:

gemini://gemlog.blue/users/Sardonyx/1598548115.gmi

I did need to eat something, but talking with a friend also helped :)

I need to continue preparing a presentation but I'm too stressed to focus. I think I also need to eat something

Wow, Spectre seems like a really cool clojure library:

http://nathanmarz.com/blog/clojures-missing-piece.html

I really want to check it out!

gemlog.blue is a convenient site for hosting a gemlog (gemini blog), for fake nerds like me who won't get around to hosting one themself for a while

https://gemlog.blue

@louiscouture I haven't been keeping up with it very much since version 8, but Java is at version 14 now, and will be at version 15 in two weeks. (Java switched to a fast release schedule a few years ago). Java is mostly backwards-compatible, so you're only likely to have to deal with new features, like the local-variable type inference added in Java 10.

I'd say it would be fine to just learn Java 8 and then find resources explaining what's been added in all the new versions, that's how most people are experiencing this anyway.

The view of the city on the way home from my parents' house is gorgeous. Unfortunately I can't enjoy it properly; gotta keep my eyes on the road!

@freemo I don't understand this, but it looks very cool! Makes me want to add electrical engineering to my already huge pile of things to learn lol

@mathlover O.O another ML! I do love math as well, though that's not what ML stands for in my case (it's just my initials)

I've been feeling antsy all day, but now I took out a piece of paper and started doodling and it's helping

I don't think the GPL is overreaching, and I don't think it's meant to trap ignorant contributors. But it's an unavoidable side-effect of the way copyleft is implemented that the restrictions are stronger than "just don't abuse people": incompatibilities between different copyleft licenses are a prime example of that. Even though both are free software licenses, you can't link EPL and GPL code together.

The GPL is not compatible with Parity. If it turned out that Parity was more effective than the GPL at ensuring software freedom, and a large body of software was created under the Parity license, archivists and hobbyists could not use that software to "develop, operate, or analyze" old GPL'd software. Not out of any malicious intent from either the Parity or GPL users, but because of the way the copyright works. This is why I think the semi-permanent nature of the copyright can be a problem even in copyleft: it makes it harder to move forward to potentially better licenses.

Parity is not about tricking people either. Like the GPL, it's largely about not being absorbed by proprietary software. But I do object to your statement that "People shouldn't (and wouldn't) choose free software because they have no other choice, they should choose it because it is objectively superior." The FSF has used this tactic before, see CLISP:

https://en.wikipedia.org/wiki/CLISP#History

I agree that we shouldn't trick people, but I don't think it's wrong to require people who use your code to also make their code free.

The point of the GPL is "you modify our software, you have to contribute it back". The AGPL goes farther and says "you use our software as a web service, you have to contribute it back". Parity goes even farther and says "you use our software to develop, operate, or analyze software, you have to contribute it back". These kinds of stronger clauses were not as necessary in the past, but with the advent of web services, they became more necessary: the GPL does nothing to protect your code from being absorbed into proprietary web apps, and even the AGPL has weaknesses in that regard.

For the record, I often use free software even when its functionality is worse than proprietary alternatives.

Regarding competitiveness, I think a difference in our philosophy is that I see free software licenses as being about freedom first, competitiveness second. We can consider a project in two fictional scenarios. In one scenario, the project is released under the GPL and is frequently used in proprietary web apps. In the other, the project is released under the AGPL and is not used as much over all, but is used in more free software web apps. In the former scenario, the software might be more "competitive" since it was used more, but the latter one is the one with more software freedom. So I think even if the Parity license is not more competitive than the GPL when we include proprietary software, it might still lead to more software freedom overall, and so be a good thing.

Lots of things left to do, but for now my desk is clean.

@namark @fikran

Making it easier to re-license (but still requiring up to a decade, so not *too* easy) is still an advantage of this system. If a flaw was found in the GPL, or if a flaw was created by some court ruling, there are many projects which would be completely stuck. Projects that have the "or any later version" clause would be in a better position because the FSF could release a GPLv4, though that still requires trusting the FSF to do this. I do trust them for now, but we don't know who will be in charge of them in a few decades.

Sometimes software becomes out of date a decade later but is still played with by hobbyists. So it's not just a binary of relevant / irrelevant. Once it's a decade out of date, I think the whole free vs proprietary fight becomes less relevant, so I think it makes sense to loosen restrictions then. It makes things easier for hobbyists, archivists, and so on. Though I'll admit that my argument is at its strongest when we consider a license like Parity, since its much stronger requirements actually could make things significantly harder for hobbyists and archivists. Since you reject the Parity license, it's easier to reject the above reasoning, but maybe we can agree that conditional on licenses like Parity being widely used, it would be a net good to have free software enter public domain before long?

The Parity License isn't for every project, much like the FSF doesn't recommend the GPL for all projects (sometimes they recommend the LGPL, and in some cases they even recommend permissive licenses). Strategically, Parity is most useful when free software already has a lot of power: if the only tool that does the job is Parity-licensed, then people will have to choose between either not doing the job, or releasing their code under a free license. So for things like C compilers and text editors, which are very competitive, Parity is a bad choice. For projects that do something novel, though, I think it could be a good choice sometimes, as it would make the novel features be accessible only to people developing free software.

So, I think Parity has its place in the grand Free Software Takeover project.

@tim Yes, that's right! And Gemini does have at least two search engines right now