opensource

There’s a lot in here to agree with. Specifically, the idea that copyleft can create a sort of mandated flywheel of open-source acceleration; my words not his. Definitely worth a read, as always.

There’s a lot in here to agree with. Specifically, the idea that copyleft can create a sort of mandated flywheel of open-source acceleration; my words not his. Definitely worth a read, as always.  
https://vitalik.eth.limo/general/2025/07/07/copyleft.html

You're right about the reasons for wanting copyleft (nudge towards openness), but you're wrong that copyleft archives them.

1. Copyleft creates rewrites, not collaboration. There are tons of examples of people rewriting copyleft encumbered code, it's cheap and fast to rewrite just the part you need. I've even done it. There are ~no examples of an existing codebases being relicensed just to adopt a copyleft dependency. If we're lucky, the rewrite is OSS, but that's not the default! I'd go as far as to claim copyleft creates more duplicated proprietary code.

No one wants to maintain a fork if they don't have to, I'd rather upstream a fix and have it be maintained for free. Permissive licenses create more collaboration this way by default. Permissive creates less duplicated code and more OSS.

2. Strong copyleft is asymmetric and exclusionary. Copyright holders have special rights to relicense under different terms that the user/consumer does not have. Many companies ban touching strong copyleft code, and most people work for companies, so you're naturally losing a huge subset of possibly collaborators.

Permissive licenses are symmetric, everyone has the same rights including the author and users.

(Can expand more if you'd like, working on a essay myself.)

Too bad it's wrong 🙃
https://farcaster.xyz/shazow.eth/0x1f75abc5