[Libreoffice-qa] Reminder: QA Meeting on Wednesday

[Joel Madero]

Hi Pedro,


> It is impossible for me to join at the 13:00 but I would like to leave two
> questions:

That's unfortunate, we should talk about moving the time and/or finding
some other way to get dedicated people like yourself on live chat during
meeting.
>
> Isn't 16% of Regressions something that TDF should be worried about? 
I'm not sure what "worried about" really means. If a major issue was
resolved and the side effect is a minor regression, not so much. If
instead there are a ton of minor fixes that have created crashers, more
of an issue. I'd need a lot more information to adequately judge the
problem.
>
> How can QA (Quality Assurance) and TDF motivate the Devs to worry about the
> consistency of the program?

From QA perspective it's not our job to judge or dictate how developers
work. They are volunteers, and they should be respected like volunteers.
What we can do is the same that we could always do:
1) Narrow down the version where the problem came to be;
2) Bibisect/bisect -- CRUCIAL
3) Appropriately tag whiteboard
4) Appropriately deal with severity/priority.

Beyond that, it's out of our control and stressing about how developer
volunteers spend their time is not in anyone's best interest....in
particular if we're falling short on our side (specifically bibisecting).

Finally, the only other advise I would give is raising the worst (and
*fully* triaged) regressions up to Robinson so that he can bring them up
during ESC. This is particularly useful when we can identify the
specific person who accidentally caused the regression and try to see if
they can fix it quickly.

Before demanding a lot of other's time, I personally would prefer we do
our job as good as possible.


Best,
Joel