Thanks for sharing your opinion on GSoC. It is good to see. If we have at least 2 core devs that want to support GSoC, I think I'm of the opinion we should let them do so and they can mentor at least 1 student. If no one wants to champion it, then, yes, it should be abandoned.
Note that GSoC may also be the sole reason we have any money owned by the project. I have always been proud of SymPy's welcoming of new contributors and use of GSoC in that process. Long back we used to show a graph showing the number of contributors that have made varying counts of git commits and compared it to other projects like NumPy, SciPy, etc. It seemed we were less susceptible to the bus factor and that our approach with new contributors did pay off. I still believe this is true and it would be a shame if we lost this aspect of the SymPy community. I always admired Ondrej for his approach and instilling this behavior in the community, which has been linked with our GSoC participation. Jason moorepants.info +01 530-601-9791 On Sun, Feb 1, 2026 at 2:23 PM Oscar Benjamin <[email protected]> wrote: > On Sun, 1 Feb 2026 at 07:46, Jason Moore <[email protected]> wrote: > > > > GSoC has played a very significant role in the history of SymPy. I don't > think that can be understated. We have always gotten lots of poor quality > contributions and interactions during the annual application phase. AI slop > causes a rise in this. > > > > We should have some careful deliberation before deciding to not > participate in GSoC in the future. My opinion is that GSoC results reflect > proportionally the time/quality of mentorship in most cases. > > I agree but I think that reviewing all these PRs right now is not > possible. I have now decided to change my own approach to doing this > in such a way that almost all of the PRs opened by new contributors > hoping to do GSOC will be ignored by me. The new basic rule is that if > the PR is not perfect or requires any effort to review then I will > ignore it and I will not leave any feedback about what is wrong with > the PR. > > I don't think that we can reasonably select from GSOC candidates > without these PRs but I also don't think it is manageable to actually > review these PRs in a way that is supportive of novices. There are > just too many people who would require too much coaching. > > I think it would be better if sympy gave up on GSOC and focussed on > recruiting higher quality contributors rather than training an army of > novices almost all of whom will leave as soon as the GSOC selections > are announced. Even those who do GSOC rarely contribute again in > future. I know that GSOC has been different in the past but I think > that things have changed and it isn't worth it any more. > > The time spent with all these PRs would be better spent just doing > what the GSOC project was suggested to do. New contributors would > still come but they would be people with a broader range of > experience. They would be more likely to hang around long term making > it more worthwhile to coach them and their good PRs would not be > ignored because of being buried in a sea of bad PRs. > > -- > Oscar > > -- > You received this message because you are subscribed to the Google Groups > "sympy" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion visit > https://groups.google.com/d/msgid/sympy/CAHVvXxSoBT4%3DKS3AuGMpcx%2BLtVBMNTGZqR8%2BLvJDAYaNNJ85QQ%40mail.gmail.com > . > -- You received this message because you are subscribed to the Google Groups "sympy" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion visit https://groups.google.com/d/msgid/sympy/CAP7f1AgLM2fkRMXdks0uHncTgPeVmCPPvHGPBzypzabd6_FX%2BA%40mail.gmail.com.
