QA Meeting Minutes - 2014-04-21

Hi all,

Minutes from our QA Meeting are here:
https://wiki.documentfoundation.org/QA/Meetings/2014/April_21

Our next meeting will be on May 5th. Agenda is available here:
https://wiki.documentfoundation.org/QA/Meetings/2014/May_05

Cheers,
--R

Hmm, Im uncertain as to what the goal is here and if its reachable.
- Reporters (end users) will most likely ignore an attempt at using a
  standardized vocabulary
- Triagers will most likely have a good grip on all the words used for one
  topic and might even derive additional information from the nuances (as
  Terrence shows here :wink: )
- Ultimately: Who should be the consumer of these queries/standardisation? How
  does it help devs to address the impportant and urgent bugs quicker?

Best,

Bjoern

From the QA meeting minutes:

(*) SUGGESTION: Standardization of our summary field for Bugzilla
    (*) Or: When searching for one phrase, display results from a

< similar one (e.g. "image" -> "picture" or "graphic")

I have been thinking about the possibility of canned pieces of queries
to ease searching for a particular kind of crash. I mean ...
(*) A segmentaion fault is a segfault is a SIGSEGV is a signal 11.
(*) A crash may be a segmentation fault, except when it is something
    else. Something else could be an assertion, SIGABRT, or signal 6,
    It could even be a hang. And lots of reports say nothing more
    specific than "crash".
and so forth.

I have not said anything here before because I hoped to make a more
concrete suggestion. But if we assume standardized words in the
summary or whiteboard or keyword (without thinking about how that
standardization happens), things become very simple.

HTH,
Terry.

> From the QA meeting minutes:
>
> > (*) SUGGESTION: Standardization of our summary field for Bugzilla
> > (*) Or: When searching for one phrase, display results from a
> < similar one (e.g. "image" -> "picture" or "graphic")

Hmm, Im uncertain as to what the goal is here and if its reachable.
- Reporters (end users) will most likely ignore an attempt at using a
  standardized vocabulary

Absolutely right. In fact, we should not even mention it to a new
reporter until the need arises.

- Triagers will most likely have a good grip on all the words used for one
  topic and might even derive additional information from the nuances (as
  Terrence shows here :wink: )

Triagers would be the first to apply standardized vocabulary.
Subject, of course, to us wanting not to discourage people who a just
not interested in vocabulary. The goal is to be able to ignore
uninteresting bugs more efficiently.

- Ultimately: Who should be the consumer of these queries/standardisation? How
  does it help devs to address the impportant and urgent bugs quicker?

I see the primary user being a triager (or maybe a reporter) looking
for duplicates.

<sky color"blue">
  Developers do not matter, because we (QA) are going to do such a
  good job that no developer ever has to look at a bug she is not
  interested in and which is ready for her attention.
</sky>

Hey, I admit that that is "blue sky". How far can we expect to
advance toward that? How much value is there in the
yet-to-be-demonstrated incomplete result? I do not know, but the
sheer number of words I am writing suggests the improbability of a
good result. Sigh.

Perhaps the detail page for a bug should have a button meaning "I am a
developer, and I have looked at this bug, and I am sorry that I spent
my time this way." Well, a buttopn plus the opportunity to say what
enticed him to go there is the first place.

HTH,
Terry.

Bjoern Michaelsen schreef op 21/04/14 23:57:

From the QA meeting minutes:

(*) SUGGESTION: Standardization of our summary field for Bugzilla
    (*) Or: When searching for one phrase, display results from a

< similar one (e.g. "image" -> "picture" or "graphic")

Hmm, Im uncertain as to what the goal is here and if its reachable.

I did this suggestion due the fact a couple of days ago:
I found a new bug report I knew it already was reported earlier (dupe).
while searching for a bugs with summary 'color', I didn't found the
dupe. After 5 minutes I thought about searching for 'coloUr' which
showed the original bug ...
So my goal was to ease a bit our search for dupes.

Kind regards,
Joren