Re: Triaging: Close as needsinfo

[Aymeric Augustin]

Hi Shai,

On 11 mai 2013, at 22:58, Shai Berger  wrote:

> On one hand -- should I have closed it myself, and saved a core-dev the 
> trouble? The triaging guide[1] seems very inconclusive about this, advising 
> on 
> other reasons to close tickets but not on needsinfo (except "tickets which 
> are 
> really feature requests").

Yes, this is the best course of action. I'll clarify the contributing guide.

There are several reasons for this:
- it increases the signal-noise ratio in open tickets,
- often the reporter never responds because he moved on and no longer cares,
- we have to keep some balance between the efforts required from people who
  report bugs and those who attempt to reproduce them.

For this ticket, if the reporter were paying my consulting rate — and that's
not far fetched, he can afford Oracle licenses after all — he would certainly
provide a more thorough description of his problem. Giving away our time
doesn't make it less respectable or valuable.

> On the other hand -- especially, in the wake of the "Perception of attitude 
> in 
> tickets" brouhaha -- should we really be closing tickets on needsinfo? In 
> other communities, I have usually seen "needsinfo" as a ticket state, rather 
> than a reason for closing; such tickets are then closed later, if enough time 
> has passed and no further info is received. I realise that, when the person 
> closing the ticket specifically says "when you have more info, please 
> re-open" 
> (as Aymeric did in this case), then the process is essentially the same, only 
> a little more efficient; but I worry that this efficiency may carry a price 
> in, as 
> Simon put it, "perception of attitude in tickets". 


When a ticket is closed a "needsinfo", it also carries the message that the
ticket wasn't reported properly, and thus wasted a bit of everyone's time.
Closing the ticket reflects this, for better or worse.

> For what it's worth, I, personally, would feel much more at ease setting 
> tickets as "needsinfo" without closing them -- and if it's a ticket state, 
> there should be no problem taking such tickets out of the reports developers 
> use, so they're almost as far out of the way as closed tickets.

I don't think it's appropriate to ignore open tickets. If we're going to ignore
them, let's close them.

Plus, everything Carl said, and which I removed from my draft :)

-- 
Aymeric.



-- 
You received this message because you are subscribed to the Google Groups 
"Django developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to django-developers+unsubscr...@googlegroups.com.
To post to this group, send email to django-developers@googlegroups.com.
Visit this group at http://groups.google.com/group/django-developers?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

Re: Triaging: Close as needsinfo

[Carl Meyer]

On 05/11/2013 06:01 PM, Shai Berger wrote:
> On Sunday 12 May 2013, Łukasz Rekucki wrote:
>> Hi,
>>
>> On 11 May 2013 22:58, Shai Berger  wrote:
>>> In
>>> other communities, I have usually seen "needsinfo" as a ticket state,
>>> rather
>>> than a reason for closing; such tickets are then closed later, if enough
>>> time
>>> has passed and no further info is received.
>>
>> To me that's just giving false hope for people viewing that ticket. If the
>> ticket is open, there's a reasonable expectation that there is some action
>> you can take to make it progress, but you most likely you can't as only the
>> reporter has the needed info. 
> 
> I disagree; every person who encounters a problem consistent with the 
> existing 
> description, can provide more info. This is, IMO, the only real (rather than 
> perceived) difference between closing as "needsinfo" and keeping the ticket 
> in 
> some sort of open limbo -- if someone else runs into the same problem, I'd 
> rather they enhance an existing report than start a new one; and nobody looks 
> at closed tickets.

It seems this is the only argument for not having "needsinfo" as a
closed state, and it rests on a premise ("nobody looks at closed
tickets") that is contradicted by the significant activity we regularly
see on closed tickets. Most people find tickets via Google or links on
e.g. Stack Overflow, and these do not discriminate between closed and
open tickets. I also have not observed a pattern of needsinfo tickets
collecting more dupes than any other type of ticket.

>> Of course, there are people who will take everything personal. Even if it's
>> getting your ticket closed by a stranger on a public bug tracker.
> 
> When people have already taken the trouble to file a ticket, if it is closed 
> and the closure seems arbitrary and off-handed, they will be offended. I 
> would 
> be. Then again, I would actually read the comments on the ticket...

In every case where I've seen a ticket closed as needsinfo, it came with
an explicit "please reopen if you can provide more info." This seems
good enough to me; I haven't observed a problem of people responding
poorly to a "needsinfo" closing (the issues are always around "wontfix"
closings). Making the change you are suggesting means introducing a new
tracker-gardening task, "closing old needsinfo tickets." Needsinfo was
added as a "closed" state precisely in order to get rid of the need for
this.

(And FWIW, community triagers like yourself should feel entirely free to
close tickets needsinfo. I think the only reason that isn't clearer in
the contributing guide is that the needsinfo status was added relatively
recently.)

Carl



signature.asc
Description: OpenPGP digital signature

Re: Triaging: Close as needsinfo

[Shai Berger]

On Sunday 12 May 2013, Łukasz Rekucki wrote:
> Hi,
> 
> On 11 May 2013 22:58, Shai Berger  wrote:
> > In
> > other communities, I have usually seen "needsinfo" as a ticket state,
> > rather
> > than a reason for closing; such tickets are then closed later, if enough
> > time
> > has passed and no further info is received.
> 
> To me that's just giving false hope for people viewing that ticket. If the
> ticket is open, there's a reasonable expectation that there is some action
> you can take to make it progress, but you most likely you can't as only the
> reporter has the needed info. 

I disagree; every person who encounters a problem consistent with the existing 
description, can provide more info. This is, IMO, the only real (rather than 
perceived) difference between closing as "needsinfo" and keeping the ticket in 
some sort of open limbo -- if someone else runs into the same problem, I'd 
rather they enhance an existing report than start a new one; and nobody looks 
at closed tickets.

> For bugs, "needsinfo" is almost identical to
> "cannot reproduce" (aka "worksforme") and I can't think of a reason to keep
> such a bug open.
> 

I hope I've given you one.

> Of course, there are people who will take everything personal. Even if it's
> getting your ticket closed by a stranger on a public bug tracker.

When people have already taken the trouble to file a ticket, if it is closed 
and the closure seems arbitrary and off-handed, they will be offended. I would 
be. Then again, I would actually read the comments on the ticket...

Shai.

-- 
You received this message because you are subscribed to the Google Groups 
"Django developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to django-developers+unsubscr...@googlegroups.com.
To post to this group, send email to django-developers@googlegroups.com.
Visit this group at http://groups.google.com/group/django-developers?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.

Re: Triaging: Close as needsinfo

[Łukasz Rekucki]

Hi,

On 11 May 2013 22:58, Shai Berger  wrote:

> In
> other communities, I have usually seen "needsinfo" as a ticket state,
> rather
> than a reason for closing; such tickets are then closed later, if enough
> time
> has passed and no further info is received.
>

To me that's just giving false hope for people viewing that ticket. If the
ticket is open, there's a reasonable expectation that there is some action
you can take to make it progress, but you most likely you can't as only the
reporter has the needed info. For bugs, "needsinfo" is almost identical to
"cannot reproduce" (aka "worksforme") and I can't think of a reason to keep
such a bug open.

Of course, there are people who will take everything personal. Even if it's
getting your ticket closed by a stranger on a public bug tracker.

-- 
Łukasz Rekucki

-- 
You received this message because you are subscribed to the Google Groups 
"Django developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to django-developers+unsubscr...@googlegroups.com.
To post to this group, send email to django-developers@googlegroups.com.
Visit this group at http://groups.google.com/group/django-developers?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.