Fwd: Re: Are crashes automatically confirmed

Matteo Sisti Sette matteosistisette at gmail.com
Thu Oct 31 23:16:42 UTC 2013




-------- Original Message --------
Subject: Re: Are crashes automatically confirmed
Date: Fri, 01 Nov 2013 00:16:21 +0100
From: Matteo Sisti Sette <matteosistisette at gmail.com>
To: Alberto Salvia Novella <es20490446e at gmail.com>

Crashes should be automatically marked as confirmed for the very fact
they are crashes.
At least when the bug report is generated with apport and contains
enough automatically collected information to ensure it actually was a
crash (i.e. a core dump or something), so it can't have been mistakenly
taken as a crash (or invented, or dreamed of) by the user.
The fact that an application has crashed is in itself by definition the
proof that some bug exists, no matter how badly the user can have used
the program or how wrong the user's report can be. A program that
crashes, no matter under what conditions or how rarely, has a bug. So if
a crash has happened, the bug report should be marked as confirmed (the
existence of the bug in itself confirms the bug).

When a user reports some other kind of bug it is reasonable that
confirmation is required by some other person that the described wrong
behavior actually exists and is not a misperception of the user
reporting it. But when a crash is reported using whatever automatic tool
that collects and uploads information about the crash, we already know
that the crash exists without any possibility of mistake, and if a crash
has happened, a bug exists.

On 31/10/13 23:38, Alberto Salvia Novella wrote:
> Thanks for the response.
>
>
> El 31/10/13 21:47, Brian Murray escribió:
>> On Thu, Oct 31, 2013 at 09:39:44PM +0100, Alberto Salvia Novella wrote:
>>> I think that crashes in Launchpad are usually automatically marked
>>> as "confirmed" if other user's crash has the same fingerprint. Is
>>> this true?
>>>
>> Bugs are automatically set to confirmed if another person clicks "This
>> bug affects me" or if another bug is marked as a duplicate of it.  The
>> apport retracer marks bug as duplicates of other bugs so that falls into
>> the second case.
>>
>> --
>> Brian Murray
>> Ubuntu Bug Master
>>
>
>
>
>






More information about the Ubuntu-quality mailing list