[Libreoffice-qa] retest ping to UNCONFIRMED bugs older than 6 months

[Tommy]

On Wed, 05 Nov 2014 22:46:03 +0100, Robinson Tryon  
<bishop.robinson at gmail.com> wrote:

> ....
> If one of you'd like to bring a proposal to the QA Meeting, that
> sounds good to me. Alternatively, I think it's fine to hammer
> something out on the QA list.

I have troubles joining the QA calls so I'll keep discussing this here in  
the mailing list.


> Beluga has some interest in some of the older groups of UNCONFIRMED
> bugs, so perhaps Tommy and Beluga could chat about some ideas here,
> then come back to the list or to the next QA Meeting w/a proposal.

here's a mail conversation between me and beluga on this topic.
I think it's better to share it in public

Il 2014-11-07 08:59 Beluga ha scritto:
> Hi,
>I read your discussion on the QA list.

I think it's better to keep the discussion open in the QA list so other  
may partecipate so I'm copying and pasting this email conversation over  
there as well.

> I agree that we can ping them, but I think we should do some tasks  
> before that.
> As you noticed, I've been going through the old ones. I'm currently at  
> July
> 2014, so not very old anymore :)
> The concentration of hard-to-repro bugs is high in the older reports.  
> Some are
> practically impossible to repro by a regular QA team member.

the rationale of pinging old bugs (which most of the time are unconfirmed  
just because they are hard to reproduce) is to ask the reporter to retest  
himself with a new release and see if the bug is gone or not.

nobody better than the original reported can reproduce a bug.

so the gain is that if the reported himself tell the bug is gone we will  
not spend anymore time trying to reproduce an hard to reproduce the bug.

so with a minimal effort (automated ping message) we can get rid of bugs  
 from the list and the reported will do the job... the same thing can apply  
to old NEW bugs as well