[Libreoffice-qa] minutes of ESC call ...

[Bjoern Michaelsen]

Hi,

On Tue, Dec 23, 2014 at 10:31:35PM +0100, Cor Nouws wrote:
> One thing that may be handled too: when confirming also check if it
> worked before (regression or not) and such.

True. In the end a fully triaged (aka a bug that a QA guys can do nothing more
about and that only can be moved forward by a developer) is either:

1/ a regression, which has been bibisected
2/ a non-regression, missing feature (enhancement)
3/ a non-regression, bug introduced with new feature from the start

We have proper tags to identify the first two of the above, but nothing to tag
the last group. If we _had_ a whiteboard status for number three too, we could
query for confirmed bugs that are:

- not a bibisected regression
- not an enhancement request
- not marked as a bug introduced with a new feature from the start

and have a list of confirmed bugs that QA can still help about. As such Id
suggest maybe a whiteboard status "newfeaturebug" for this. Opinions?

Best,

Bjoern

P.S.: Note there is some wiggle room in distingushing group 2/ and 3/. IMHO if
a developer of a feature says its group 2/ (aka that it was not in the planned
scope of the new feature) that should be accepted in general (and possibly also
documented that way).