No, AI is not causing a PR or commit author explosion in #curl (so far)
@bagder Did it cause a rise in submitted (but not necessarily merged) PRs, though?
While commit counts are nice to know for the stability of the project, submitted PRs (no matter if merged, closed or kept open) inform us on the workload of the people behind cURL.
@thibault I think there might've been a small increase, but not to a degree that very noticeable in graphs. I suspect there might be coming more in a future.
@bagder I don’t think AI somehow can contribute to any spikes in the number of commits in any open-source projects. It’s just a tool, like the PC was at some point in time in the past. Seasoned engineers will do things better using AI tools as advisers. AI tools without experience on the ground are worth nothing. The final decision will be on engineers who have the experience to justify the proposal.
@SerhiyMakarenko sure it can. It allows many more authors to contribute code. People that previously could not, now can because AI helps them do it.
@bagder Then why does AI allow the statistics to show it actually does not? I believe there is some misconception between understanding what AI is and what people can do with it. In my opinion, AI can boost someone who has experience and knowledge to do something, but can’t do the same for the person who knows nothing about software development, networking, or curl specifically.
P.S. I’m not advocating AI, I’m advocating you as an engineer who created curl and is keeping it up no matter what.
@SerhiyMakarenko I'm using the stats to show what is actually going on. Not what *could* happen.
@bagder @SerhiyMakarenko
It also lowers the barrier for experienced developers to get familiar with a large codebase, hence making it easier to get involved.
@bagder I hope I will continue to not explode.🤭
@bagder Will be interesting to see how 2027 will turn out. If the new tools will continue to find a high number of problems in the code or if this was it and things will return to the old commit levels again.