[Libreoffice] [Libreoffice-qa] 3.5.0 QA ... from BHS 1 to BHS 2

[Cor Nouws]

Hi Pedro,

Thanks for replying. And as we wrote, I reply to the lists too.

Pedro Lino wrote (09-01-12 01:34)

>> So time to prepare bug hunt session two.
>>   http://wiki.documentfoundation.org/QA/Improving_QA-Release-3.5
>>   and
>>
>> http://wiki.documentfoundation.org/QA/Improving_QA-Release-3.5#2nd_Bug-hunting_session
>>
>> Or is there something that should make us change our mind for that?
>> As far as I've been able to notice, preparations for beta3 go fine, so
>> that's OK!
>
> I believe the second hunting session is only after the first RC is released?

Yes, that is the intention.

>> As may be read from the other thread (thread 'Evaluating first 3.5.0
>> Bug-Hunt Session' on the QA list) I'm far from sure about the impact on
>> level of issues, but looking at the overall activity/awareness, I think it's
>> worth.
>>
>> We could send a mail to the people that submitted their fist bug during the
>> session, asking their feedback for possible improvement. Would that be OK,
>> do you think?
>
> I think that sending an email might be a little overkill... Announcing
> at the main LO page should be enough.

Hmm, the mail is not intended to announce, but to ask feedback. (Of 
course one may think that if there is a strong feelng for doing so, 
people would have done. But anyway..)

> Having BHSs too often might discourage people. Unless TDF/LO can show
> that a substantial proportion of the bugs reported were fixed (or at
> least analyzed). Currently there are over 350 unsolved/unreviewed bugs
> for Windows only...

A good alternative might be that it's obvious that the sessions help to 
find bugs in time.
But I agree: many untouched (recent) bugs, does not look nice...
One of the effects of the sessions of course also is that (hopefully) 
some more people get aware of and involved in a bit regular work on QA.

> There is simply no time to triage/review/check bug reports if LO keeps
> pumping releases at this pace... The schedule is not realistic for the
> number of people working on QA. It might work fine for Ubuntu which
> has a lot of money and a lot of people...

Hmm should I agree or disagree...
I think it is correct that the enormous speed at which the developers 
keep working, is a problem for the people doing QA.
( Related is the great ignorance that I sometime notice, that it is 
important to prevent unnecessary and improper bugs, and that on the 
contrary, getting clarity as early as possible (before submitting smthng 
in BugZilla) helps. )

On the counter side, I think that because of the relatively small steps 
in the most releases, that alone does not put a high load on testing.
Thinking of structured testing however (via Litmus) I can imagine that 
for some it is not attractive to do that oh so often.

> Also, I think it's really BAD that the *public* wiki page talks about
> a reward for a Bug-hunting competition but there is/was no reward for
> the BHS "hero". How is it different?

I think I have not been secure enough here, which gave confusion and 
maybe wrong expectations and a half realised idea . Sorry for that. 
Anyway the bug hunt hero is mentioned at the blog: Gustavo Pacheco.
 
http://blog.documentfoundation.org/2011/12/31/thanks-for-all-that-supported-our-first-libreoffice-3-5-bug-hunting-session/
Should have announced that too.

Thanks,

  - Cor
  - http://nl.libreoffice.org