Hi Junio, On Wed, 14 Jul 2021, Junio C Hamano wrote: > Johannes Schindelin <Johannes.Schindelin@xxxxxx> writes: > > >> It wasn't quite obvious why we justify spending 370 minutes one more > >> time only to rerun 30-second job, though. > > > > True. > > > > And this is not a new problem. Every time anything happens in those > > `osx-gcc` or `osx-clang` jobs (e.g. that transient problem with the broken > > pipes in t5516 [*1*], that's a fun one), a full re-run is necessary, or > > else the commit and/or PR will remain marked as broken. > > > > In other words, while it is totally appropriate for me to explain this to > > you in this here thread because it came up as a tangent, it would be > > inappropriate to stick that explanation into this patch's commit message. > > We do not make a habit of adding tangents that came up during patch > > reviews into commit messages, and I do not intend to start such a habit > > here, either. > > I do not agree; a brief mention "even though piling more and more on > the primary workflow would make it even less convenient to re-run, > it is already so bad that another one would not hurt too much more" > would be a clue good enough to motivate others to do something about > it if they feel like it. By that rationale recent additions to `.github/workflows/`, which have a much much much much bigger impact on the runtime (such as the change that lets `linux-clang` run the test suite with SHA-1 and then again with SHA-256, almost doubling its runtime), should have added that apology to the commit message. But that did not happen. Besides, for all we know the problem might go away at any stage because pretty much all other main CI systems have a way to re-run only failed jobs. GitHub Actions will probably get it at some stage, too, I vaguely remember seeing it on some public roadmap somewhere. So I really do not appreciate this pushing for including an explanation in the commit message for something that is only relevant (if at all) in the context of an utter tangent (don't we have many of those over here!) during the review of the patch. It's just some curiosity that will eventually be an exclusively historical curiosity, of no interest except to a few CI nerds. The issue has everything to do with a currently missing feature in GitHub Actions, and it has nothing to do at all with what the patch is about, which is: to add a `sparse` job to our CI runs. Ciao, Dscho
