ESC meeting minutes: 2025-07-03

[Áron Budea]

Hi Mike, Ilmari,

To add to this, there are still about 170 bugs that have bibisected
keyword, but don't have bisected. They're likely mostly older bugs from
before the era of bibisect repos with individual commits, but there could
be newer bugs as well, as ranges where builds failed still exist
occasionally, or perhaps the steps can't be completed due to a crash.

The distinction between a bug being completely or incompletely bibisected
is big (ie. the failing commit is known vs. still not known), and I'd say
it's important to denote that, as the latter could use more specialized
triaging even if someone doesn't want to work on a fix.

However, I also agree that having to use two keywords at the same time
isn't ideal, and the current keywords can be confusing. My idea would be to
use bibisected for the complete and partlyBibisected for the incomplete
case (or other variants that make the result obvious).
And ESC reporting of the numbers isn't very helpful as it is now: one
should show the number of bibisected bugs that also include the keyword
bisected, and the other the ones that don't include the keyword bisected
(assuming the current combination of keywords).

Cheers,
Aron
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/libreoffice/attachments/20250710/611111a2/attachment.htm>