Has anyone used this project for a large project code base? I know there is 50 uses, but I have used all on previous project. I have also tried cody, same gpt4 model, but for some reason I felt cursor.sh is better, but I may be biased. Thoughts?
Define large? I think mine’s somewhere in the middle. Hundreds of files written over 25 years though
also curious about what you mean by “large”, but i will say that in my opinion, cursor is 100x cody in terms of its ability to rapidly “learn” your codebase (and relevant docs). i switched everything from cody to cursor after 5 minutes.
I’m still figuring stuff out but slowly learning how to make it work across a large number of files. So far so good. Could be better, but also is better than anything else I’ve used
editing multiple files simultaneously is not a strong point atm.
aider is better at this. (but way worse at understanding multiple files).
i am trying to form a voltron: running aider via rift inside of cursor…but currently it is more of a frankenstein
I use it for two big projects with some huge enterprise clients
Probably >100k LOC, and >50 devs
we use it for our pretttyyy big codebase and it works pretty good (probably even a bit overoptimized for big codebases over small ones ;( I would like cursor to be unbeatable at small codebases.)