Closed Freed-Wu closed 9 months ago
Look at the past merges. You have to change more than repos.txt
Update build.py
Also the readme and the tests.
Updated
Related discussion: https://github.com/nvim-treesitter/nvim-treesitter/discussions/5588
@grantjenks
You can’t just delete all the licenses from the readme. Distributing in binary form doesn’t necessarily remove all license restrictions.
Can we have some method to update license automatically like https://github.com/nvim-treesitter/nvim-treesitter/discussions/5588#discussioncomment-7415038?
You can’t just delete all the licenses from the readme.
I have added the licenses.
Distributing in binary form doesn’t necessarily remove all license restrictions.
nvim-treesitter doesn't add any license in README. Is it necessary?
@grantjenks
And, I advise to add a scheduled CI/CD to update the version.
@grantjenks Any other problem? If you are busy to review this PR, can you add me to collaborators?
Looking at the project licenses, I don't think you can add GPL'd code and maintain this project as Apache 2.
ChatGPT thinks so too: https://chat.openai.com/share/fe181dca-6ebc-4965-890e-d8ef20cfcb6b
IANAL
add GPL'd code and maintain this project as Apache 2.
Where is GPL's code? This project's license shouldn't be changed.
Could you move the hashes of the subrepo to stable commit (or the commit with latest release attached if possible)? I don't think all of them are stable.
Also a lot of them are missing src/parser.c
. (Written in JS)
Those will fail to build.
a lot of them are missing src/parser.c
Not sure why missing. Let us delete them before we know why.
There will need to be an ALLOWLIST of Apache2 compatible software licenses.
How to get the allowlist?
There are many languages information is stored in https://github.com/nvim-treesitter/nvim-treesitter/blob/master/lua/nvim-treesitter/parsers.lua#L79-L2269 and updated very frequently. We should have a method to extract them to get the latest parsers.
Is it really a better approach though?
I find locking down to commit id's to be a much more sensible option.
Fix #32