qa-tagger

qa metadata for commits is write-once and cannot be updated

Bug #659629 reported by Edwin Grubbs
26
This bug affects 4 people
Affects Status Importance Assigned to Milestone
qa-tagger
Triaged
High
Unassigned

Bug Description

When a revision has a pqm commit message with a bug number that is wrong, qatagger will consider it a blocking revision if the bug does not have a qa-ok/qa-untestable tag, but it often shouldn't have that tag.

For example, revision 9864 has bug 644824 in its commit message, but that bug won't be fixed until the next cycle.

We need to be able to unlink that bug from the qa taggers view of the revision (and add the right one)

See original description
Revision history for this message
Robert Collins (lifeless) wrote : Re: [Bug 659629] [NEW] mistakes in pqm commit message should be ignorable

More broadly: any metadata we use *will* be wrong from time to time,
so we need to accomodate that.

This may be a good reason to have a separate revisions-status DB at
some future point.

-Rob

Revision history for this message
Ursula Junque (ursinha) wrote :

On Wed, Oct 13, 2010 at 12:40 AM, Robert Collins
<email address hidden>wrote:

> More broadly: any metadata we use *will* be wrong from time to time,
> so we need to accomodate that.
>

Robert, is this the "partial branches that cannot be landed alone" problem?
Or just misconnections between bugs and commits? If it's the latter, we can
handle today using the qa-untestable tag, but I agree we should handle such
cases more appropriately.

>
> This may be a good reason to have a separate revisions-status DB at
> some future point.
>
> -Rob
>
> --
> mistakes in pqm commit message should be ignorable
> https://bugs.launchpad.net/bugs/659629
> You received this bug notification because you are subscribed to
> Launchpad Suite.
>

Revision history for this message
Robert Collins (lifeless) wrote :

On Thu, Oct 14, 2010 at 8:36 AM, Ursula Junque <email address hidden> wrote:
> On Wed, Oct 13, 2010 at 12:40 AM, Robert Collins
> <email address hidden>wrote:
>
>> More broadly: any metadata we use *will* be wrong from time to time,
>> so we need to accomodate that.
>>
>
>
> Robert, is this the "partial branches that cannot be landed alone" problem?
> Or just misconnections between bugs and commits? If it's the latter, we can
> handle today using the qa-untestable tag, but I agree we should handle such
> cases more appropriately.

We can use this bug for 'misconnections between bugs and commits', but
the topic is a broad one: any data has to be editable to permit
correcting mistakes. Some sources of data are harder to edit than
others, so in qatagger we may want to mirror the source data and
permit editing the mirror.
/handwave

Ursula Junque (ursinha)
summary: - mistakes in pqm commit message should be ignorable
+ mistakes in commit message should be ignorable
Robert Collins (lifeless)
summary: - mistakes in commit message should be ignorable
+ qa metadata for commits is write-once and cannot be updated
description: updated
Changed in qa-tagger:
status: New → Triaged
importance: Undecided → High
Revision history for this message
Aaron Bentley (abentley) wrote :

See also https://dev.launchpad.net/LEP/RevisionBugLinks

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Subscribing...

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.