Conversation

the simplicity of the ISC and MIT license is why i've largely switched over to it for anything that doesn't make sense as AGPL. copyleft doesn't matter because the community will keep the free releases alive anyway.

Quote Tweet

Matthew Garrett

@mjg59

Jun 10, 2021

It really shouldn't need pointing out that the way to win hearts and minds isn't to fork code under a permissive license just so you can add copyleft material to it

Show this thread

Ariadne Conill

@ariadneconill

Jun 10, 2021

the reality is that 95% of FOSS releases are done because somebody needs a thing, and so they make the thing, and they don't really want to think about the consequences of making the thing. GPLv2 worked because it was the standard, GPLv3 fragmentation broke that.

Ariadne Conill

@ariadneconill

Jun 10, 2021

and to be clear, i absolutely adore and agree with the goals of GPLv3, but at the same time, i want a hassle-free life, and so unfortunately, the best GPL to choose is none.

Replying to

The goals don't match the reality of an incredibly complex and restrictive license which restricts usage including preventing projects mixing GPLv2-only and GPLv3 code and other useful things. The ends don't justify the means and it's not clear what it actually accomplishes.

Replying to

and

It's pretty bad having an OS where you can't even share code/knowledge properly between the kernel and libc due to restrictive licensing. Who wants to deal with that? If someone worked on both Linux and glibc they would end up actively infringing on kernel licensing in practice.

7:35 AM · Jun 10, 2021Twitter Web App

Replying to

and

Causes a whole lot of complexity and annoyances for open source developers while seemingly not accomplishing anything for end users. How many products really chose to allow the user to replace the OS only because they wanted to use GPLv3 code and chose that over avoiding it?

Replying to

and

Meanwhile, users don't have a mainline Linux kernel ZFS due to GPL license restrictions. It's also a major part of why there's so much fragmentation between TLS libraries in a way that it's not feasible to use a single one like Android (BoringSSL) on a traditional distribution.