"Fix all tickets" experiment interim report

Regina Obe lr at pcorp.us
Sun Jun 21 22:09:01 PDT 2026


> >  - some improvements in tests (new oss-fuzz targets) uncovered more
> > issues to get fixed in existing codebase - thanks Even Rouault for
> > addressing some of these before.
> 
> I don't appreciate at all to have served as a guinea pig of your experiments,
> and I'm not sure I'll fix any other PostGIS issue anytime soon, which you
> probably don't care if there isn't any more remaining.
> You don't conduct such experiment that creates hundreds of pull requests over
> a couple of days, and merging tens of them on the course of a week end
> without letting chance to others to review them, without having informed
> others in the community at large before (not speaking about potential private
> chat between insiders that might have occurred).
> 
> As a remainder:
> https://postgis.net/development/rfcs/rfc01/#when-is-a-vote-required :
> "Adding substantial amounts of new code", "Anything that might be
> controversial"
> 
> --
> http://www.spatialys.com
> My software is free, but my time generally not.
> LLMs contribute to global warming and brain rot


Darafei,

I think we should not make any more changes to 3.7.0 until we've released except for any regressions that show up as a result of 3.7.0alpha1.
I'm hoping to ship out 3.7.0alpha1 this week, so the less churn we have the better.

Don't take that as a message that I don't appreciate the work.  What I see here so far is great and I do believe you spent a great deal of time checking correctness.
Given that you managed to not crash the garden tester.  People always crash the garden tester when they make a pile of changes.

It's just a little too fast for comfort and making several people very concerned at the ability to really confidently test all of this. 
I'd like to see if users run into any issues in 3.7.0alpha1 with what we've changed so far.

As part of an LLM policy, we should require an additional committer check the code before it gets committed unless if it is less than x lines of code (to be agreed on).

Thanks,
Regina



More information about the postgis-devel mailing list