New statuses in Bugzilla?

classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

New statuses in Bugzilla?

Tony Mechelynck
Please comment on the following (Gerv, are you there?)

I think we need (when and if possible) two new statuses (or maybe keywords?)
in Bugzilla:

- 1) the report is too vague: NEEDINFO has been suggested. (cf. the thread
'New "INCOMPLETE" resolution in Bugzilla' in this same NG and a few others)

- 2) the bug has been sighted on and off but no way to produce it "every time
for every user" has been found: I suggest UNSYSTEMATIC. Note that the fact
that the bug appears and disappears with no apparent cause might be
"unobvious" at the time of reporting the bug.

Both of these are akin in that they point towards a true bug, but which the
devs aren't able to reproduce at will. The difference is that in case (1)
there might possibly be a way to reproduce the bug, if only the reporter were
more loquacious; in case (2) one or more people have seen the bug at least
twice, and the bug still sometimes appears uncalled-for, but (despite their
best efforts) no one has been able to make the bug appear at will: this is
what "Reproducible: Sometimes" is for, except this setting should be settable
at any time, not just at reporting time.


Best regards,
Tony.
--
The best way to make a fire with two sticks is to make sure one of them
is a match.
                -- Will Rogers
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Andrew Schultz-2
Tony Mechelynck wrote:
> - 2) the bug has been sighted on and off but no way to produce it "every
> time for every user" has been found: I suggest UNSYSTEMATIC. Note that
> the fact that the bug appears and disappears with no apparent cause
> might be "unobvious" at the time of reporting the bug.

I don't know what this gets us.  If a bug appears occasionally for a
developer, they have a chance of fixing it (although having consistent
steps to reproduce would certainly help).  Some bugs get fixed without
the developer seeing it at all (just code inspection).  I think if we
try to make everything too fine-grained, there will be 100 statuses and
all we can guarantee is that most bugs will have the wrong one.

If the developer thinks better steps to reproduce are needed, then the
developer can change the status to NEEDINFO.

--
Andrew Schultz
[hidden email]
http://www.sens.buffalo.edu/~ajs42/
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Mats.Palmgren (Bugzilla)
In reply to this post by Tony Mechelynck
Tony Mechelynck wrote:
> - 1) the report is too vague: NEEDINFO has been suggested.

Yes, we really need a NEEDINFO *status* (NOT a resolution!).
We have been needing it for the past 5 years!
https://bugzilla.mozilla.org/show_bug.cgi?id=136107


> - 2) the bug has been sighted on and off but no way to produce it "every
> time for every user" has been found: I suggest UNSYSTEMATIC.

The status field should indicate the status of the /bug report/.
"UNSYSTEMATIC" seems more like a characterization of the /bug/.
It can be added to the Whiteboard, or added as a keyword to allow
queries if necessary.

_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Tony Mechelynck
Mats Palmgren wrote:

> Tony Mechelynck wrote:
>> - 1) the report is too vague: NEEDINFO has been suggested.
>
> Yes, we really need a NEEDINFO *status* (NOT a resolution!).
> We have been needing it for the past 5 years!
> https://bugzilla.mozilla.org/show_bug.cgi?id=136107
>
>
>> - 2) the bug has been sighted on and off but no way to produce it
>> "every time for every user" has been found: I suggest UNSYSTEMATIC.
>
> The status field should indicate the status of the /bug report/.
> "UNSYSTEMATIC" seems more like a characterization of the /bug/.
> It can be added to the Whiteboard, or added as a keyword to allow
> queries if necessary.
>

IMHO we need something other than UNCONFIRMED to avoid triagers coming forever
back to an unreproducible bug (or a bug not reproducible at will by following
a well-defined procedure); and I thought NEW was supposed to mean the bug
_was_ reproducible, or at least had been seen by some "trustworthy" user
(trustworthy enough, at least, to have been granted CANCONFIRM privileges). An
unsystematic (or sporadic, or whatever) bug may have been sighted twice or
more by one or more "grass-roots" users with no privileges at all.

IIUC (and please correct me if I'm wrong), the list of keywords is a closed
one: only keywords listed at https://bugzilla.mozilla.org/describekeywords.cgi 
may be used -- by "closed" I mean: adding a new one is a policy decision;
keywords not in the list must not be used (even if Bugzilla will allow them).

Whiteboard mentions, OTOH, are "open": any reporter may write there anything
that (s)he deems useful; but apart from that, what's the difference between
whiteboard and keyword? Maybe the whiteboard is not query-able?


Best regards,
Tony.
--
hundred-and-one symptoms of being an internet addict:
182. You may not know what is happening in the world, but you know
      every bit of net-gossip there is.
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Andrew Schultz-2
Tony Mechelynck wrote:
> IMHO we need something other than UNCONFIRMED to avoid triagers coming
> forever back to an unreproducible bug (or a bug not reproducible at will

triagers /should/ be going back to a bug that's not very reproducible so
they can find a better way to reproduce it (if it's so unreproducible
that fixing it is hard).  Perhaps one person has prefs or a system
configuration that makes it more reproducible, or just has some insight
that helps.

If nobody looks at the bug, how can it ever be fixed?

> IIUC (and please correct me if I'm wrong), the list of keywords is a
> closed one: only keywords listed at
> https://bugzilla.mozilla.org/describekeywords.cgi may be used -- by
> "closed" I mean: adding a new one is a policy decision; keywords not in
> the list must not be used (even if Bugzilla will allow them).

Yes.  But they're easier to add and would be more appropriate for
something like you're suggesting.  In fact, in situations like this now,
people might use the qawanted keyword.

> Whiteboard mentions, OTOH, are "open": any reporter may write there
> anything that (s)he deems useful; but apart from that, what's the
> difference between whiteboard and keyword? Maybe the whiteboard is not
> query-able?

openness is the only difference.  Keywords are more easily queryable and
are better defined.

--
Andrew Schultz
[hidden email]
http://www.sens.buffalo.edu/~ajs42/

_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Tony Mechelynck
Andrew Schultz wrote:

> Tony Mechelynck wrote:
>> IMHO we need something other than UNCONFIRMED to avoid triagers coming
>> forever back to an unreproducible bug (or a bug not reproducible at will
>
> triagers /should/ be going back to a bug that's not very reproducible so
> they can find a better way to reproduce it (if it's so unreproducible
> that fixing it is hard).  Perhaps one person has prefs or a system
> configuration that makes it more reproducible, or just has some insight
> that helps.
>
> If nobody looks at the bug, how can it ever be fixed?
>
>> IIUC (and please correct me if I'm wrong), the list of keywords is a
>> closed one: only keywords listed at
>> https://bugzilla.mozilla.org/describekeywords.cgi may be used -- by
>> "closed" I mean: adding a new one is a policy decision; keywords not
>> in the list must not be used (even if Bugzilla will allow them).
>
> Yes.  But they're easier to add and would be more appropriate for
> something like you're suggesting.  In fact, in situations like this now,
> people might use the qawanted keyword.
>
>> Whiteboard mentions, OTOH, are "open": any reporter may write there
>> anything that (s)he deems useful; but apart from that, what's the
>> difference between whiteboard and keyword? Maybe the whiteboard is not
>> query-able?
>
> openness is the only difference.  Keywords are more easily queryable and
> are better defined.
>

Thanks for info.


Best regards,
Tony.
--
Any sufficiently advanced technology is indistinguishable from a rigged
demo.
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

majken@gmail.com
In reply to this post by Tony Mechelynck
I agree that "unconfirmed" is the right place for things that aren't
easily reproducible.  IMO a new default state of UNTRIAGED is more
accurate for what you're trying to achieve.  NO idea how much work
that would be.

_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Gervase Markham
In reply to this post by Tony Mechelynck
Tony Mechelynck wrote:
> I think we need (when and if possible) two new statuses (or maybe
> keywords?) in Bugzilla:

It's unlikely that we'll get new statuses until Bugzilla supports
customised statuses natively. When it does, we can decide which ones we
want :-)

There already exist keywords which attempt to subdivide statuses - e.g.,
"clean-report". But I don't think they are widely used.

Gerv
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Tony Mechelynck
Gervase Markham wrote:

> Tony Mechelynck wrote:
>> I think we need (when and if possible) two new statuses (or maybe
>> keywords?) in Bugzilla:
>
> It's unlikely that we'll get new statuses until Bugzilla supports
> customised statuses natively. When it does, we can decide which ones we
> want :-)
>
> There already exist keywords which attempt to subdivide statuses - e.g.,
> "clean-report". But I don't think they are widely used.
>
> Gerv

Hm. How can we deal with the following problem:

- User U1 files a report.

- Triager T1 looks at that report, finds it wanting, and leaves a comment
asking for more info.

- U1 never replies.

The bug is still UNCONFIRMED, but as long as there is no additional comment,
any look at it by a triager (including T2 or T3) is a waste of time.


If we have many bugs like that (and I think we do, but I haven't checked),
we're wasting a lot of our triagers' time.


Now I don't do triaging; what I do (concerning Bugzilla) is filing bugs,
trying to get them into the proper category (I don't always succeed),
answering (if I can) questions for additional info, etc. So my chip in this
game is not about not wasting my time but about getting my bugs fixed, which
includes not wasting the _other_ guy's time.


Best regards,
Tony.
--
Bore, n.:
        A person who talks when you wish him to listen.
                -- Ambrose Bierce, "The Devil's Dictionary"
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Andrew Schultz-2
Tony Mechelynck wrote:
> The bug is still UNCONFIRMED, but as long as there is no additional
> comment, any look at it by a triager (including T2 or T3) is a waste of
> time.

Earlier in the discussion, people proposed a NEEDINFO for this purpose...

--
Andrew Schultz
[hidden email]
http://www.sens.buffalo.edu/~ajs42/
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning
Reply | Threaded
Open this post in threaded view
|

Re: New statuses in Bugzilla?

Tony Mechelynck
Andrew Schultz wrote:
> Tony Mechelynck wrote:
>> The bug is still UNCONFIRMED, but as long as there is no additional
>> comment, any look at it by a triager (including T2 or T3) is a waste
>> of time.
>
> Earlier in the discussion, people proposed a NEEDINFO for this purpose...
>

<quote src=Gerv>
> It's unlikely that we'll get new statuses until Bugzilla supports customised statuses natively. When it does, we can decide which ones we want  :-)
>
> There already exist keywords which attempt to subdivide statuses - e.g., "clean-report". But I don't think they are widely used.
</quote>



Best regards,
Tony.
--
Actors will happen even in the best-regulated families.
_______________________________________________
dev-planning mailing list
[hidden email]
https://lists.mozilla.org/listinfo/dev-planning