I can think of at least a few counterexamples. The logic behind the hypothesis is decent and tells a plausible story, but I'd like to see a more robust analysis; has anyone done one since the article was first posted 10 years ago?
Noteable counterxamples (excluding e.g. Linux):
- Git (GPLv2)
- Blender (GPLv2 & 3, from the looks of it)
- Krita (GPLv3)
- MySQL (GPLv2) -- still seems very popular in 2025
- QGIS (GPLv2)
Don't forget Qt (though I think that's more a corporation wanting to create an incentive to make commercial users pay for a license)
Git - jj[1] uses gitoxide for its git interop/backend (git being the backend of jj you're expected to use), both Apache/MIT; it functioned so cleanly I assumed it was invoking git directly, but apparently it's all custom!
For new things I'd guess PostgreSQL vs MySQL is probably on the same order as llvm vs gcc; probably with PostgreSQL being better off than llvm even.
[1]: https://github.com/jj-vcs/jj/ (started by a Googler as a hobby project, now their full-time job at Google, so kinda fits into the "big orgs will replace things with permissive-licensed versions" narrative)
I always use git as a binary and don’t have any interest in forking it.
Most of the new databases of last 15 years have some kind of restrictive license that aims to stop AWS from selling a managed version of it that also limits what I can do it. So they are dead to me. Postgres is growing not just because it is a good database which is taking ideas from document databases, but because you can build on top of Postgres and feel confident that you can commercialize or open source your work and people will be able to use it.
Notably, libgit2[1] is GPLv2 with a full linking exception, while e.g. go-git[2] is Apache2.
Blender used to be proprietary,the community had to raise money to buy the source code to make it free software. What a crazy success story for opensource.
> Notable counterxamples (excluding e.g. Linux)
These are mostly executables/end products, aren't they?
GPL allows protecting the business logic by discouraging hostile forks: after all, forks of such projects can't relicense GPL files w/o permission, and upstream can just copy the changes back w/o having to ask.
On the other hand, permissive license favor dissemination. Would {fmt} (libfmt) be as ubiquitous as it is, if it were GPL instead of MIT?