[Libreoffice-qa] Stats I plotted + Getting rid of bugs

[Bjoern Michaelsen]

Hi,

On Sat, Nov 03, 2012 at 02:16:55PM +0100, Florian Reisinger wrote:
> I want to make life easier, but maybe it is the wrong approach, but
> IMHO bugs reported against old beta (and RC) versions should get
> checked, because (the possibility is higher in beta stadium) this bug
> has been already fixed / is a duplicate of another bug (BTW: Does
> anyone know a way to especially search for possible duplicates..?))
> AND what is most important the 3.3 and 3.4 won't be developed further
> (soon the same with 3.6).

But when there is no way to check if people checked with a newer version,
nagging them to do something which the possibly already did will cause
irritation and confusion.

> I excluded the enhancement request because of one reason: I don't
> understand why we allow people to file an enhancement request, because
> a) The developer codes features he likes / would like to have b) We
> create deceitful hopes, because the submitter thinks we will implement
> every feature request (This paragraph = My opinion)

The feature request fantasies need an outlet and as long as we can separate
them in queries, its fine to have them on bugzilla. Also it is a good way to
bring together people with a common interest (a feature), resulting in product
improvement.

> Why isn't it trivial to set the version to the one used at the moment
> of bug submission

Because there are a lot of bugs checked with multiple versions (but not all) by
multiple reporters. In a ideal world, "version" wouldnt be a field, but a list
of all the versions and then a BUGREPORTED/CANTREPRODUCE/UNTESTED tristate. It
would star with all in UNTESTED, with the expception of the reporters version
-- which is BUGREPORTED. Others could then add too that.

Best,

Bjoern