Want to hear from those who have actually used chatGpt/Gemini/copilot etc on open source repo.s to make changes to a complicated code base (or a repo where you had no clue about). Which repo and which model did you use and please provide examples on how did AI models help you add features or contribute to the open source?
World Conflicts
Yesterday
1068
I am Columbia alumni
Tech Industry
11h
2587
Asians - what are your thoughts on asian female white male ?
India
3h
171
Godi Media: Very little coverage of Prajwal case.
World Conflicts
Yesterday
674
American police seem to work only when Israel is challenged
Tech Industry
12h
340
What notice do you give when leaving job?
I use copilot daily. Mostly it is good if you can prompt it get it to tab complete what you want. This can be anything from a comment that I'm too lazy to form a full English sentence for to a function that's mostly the same but eg, operates on a different field of the struct.
Thanks. I tried it just now. It does not seem to read the repo. I'm helping someone with an open source codebase and we both don't understand the intricacies unless we put in significant time ourselves understanding the whole code base. I was hoping that the copilot would read the repo and tell us what changes we need to make in order to support something new.
It only uses the code around the area you're editing as context. It won't read the whole repo and help you make a big refactor or something