https://bugzilla.gnome.org/show_bug.cgi?id=696182 sysadmin | Git | unspecified
--- Comment #5 from Colin Walters <[email protected]> 2013-03-20 17:14:43 UTC --- (In reply to comment #4) > (In reply to comment #3) > > The jhbuild build bots would immediately fail though, and we'd hopefully > > notice > > and fix it quickly. > > True, but this means lots of people get notified (e.g. through Jenkins) and > module health stats get skewed unnecessarily. I think it’s better to get the > original committer to fix their mistake before pushing, then nobody has to > waste time context switching to “what’s gone wrong with Jenkins” mode. Right; but as a general rule I think one has to carefully constrain what code gets run as part of a pre-commit. Checking for trailing whitespace and that sort of thing is pretty easy, runs in a predictable amount of time, etc. But growing that closer to "make check" type stuff is another world. There's the simple fact that we're running arbitrary code from the repositories on the git server itself (does this code have write access to the repo?). The other way to ensure that bad code doesn't make it into the repos is to automatically check incoming patches before they land. -- Configure bugmail: https://bugzilla.gnome.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are watching the QA contact of the bug. You are watching the assignee of the bug. _______________________________________________ gnome-infrastructure mailing list [email protected] https://mail.gnome.org/mailman/listinfo/gnome-infrastructure
