Hi, On Thu, Jul 10, 2025 at 4:26 PM Miklos Vajna <vmik...@collabora.com> wrote: > * QA stats > + propose is to drop the "bibisected" keyword from Bugzilla + ESC > minutes (Mike) > + see > https://lists.freedesktop.org/archives/libreoffice/2025-July/093479.html > + Áron's proposal: > https://lists.freedesktop.org/archives/libreoffice/2025-July/093516.html "use > bibisected for the complete and partlyBibisected for the incomplete case" > + made sense in the past (Xisco) > + bibisect: not a specific commit, just a range > + bisect: specific commit > + "range" commits are quite rare by now > + same at the end usually: both bibisect and bisect for most bugs > + can look at the bibisect removal (Xisco) > + AI: do this (Xisco)
I see my suggestions mentioned, but don't see any comments related to the points raised there, in particular: - still many bugs exist that are only bibisected to a range, - right now 1070 open bugs have the bisected keyword vs. 176 that don't, but have the bibisected keyword, - these bugs could use further specialized triaging to identify the regressing commit, - the use of two keywords together (bibisected + bisected) can be avoided by having two distinct, more descriptive keywords. Could someone clarify if these have been discussed, and whether this option has been compared to just removing one of the keywords? Thanks, Aron
