1239699 - print the bugid that is the cause of a regression, probably do more

Status: RESOLVED FIXED

(Testing :: mozregression, defect)

Description

[Julien Pagès (:parkouss)]

It looks like bugs that introduced a regression should be flagged with special keywords.

As someone suggested in a release quality thread (and this is an excellent idea!), mozregression could help in multiple ways here:

- print the bug id/url that probably caused the regression, and tell to the user to set the relevant flags in that bug
- we could even think about doing that in an automated way, more or less in a mozreview way to publish the review ("Comment and set regressions flags on bug XXX ?[Yn]").

So, I have some questions:

- which flags are needed ?
- I believe sometimes users are working on a bug to find the regression (let's call it "bug1") and then they use mozregression, and find the bug "bug2" which seems to be the cause. Do we need/want to do something about "bug1" ? (it's just that I'm thinking about it right now, but we can see that later)

Comment 1

[Henrik Skupin [:whimboo][⌚️UTC+2]]

The keyword we should set on bug1 is `regression`, and if `regressionwindow-wanted` is already set it needs to be removed. Further a dependency has to be added so bug1 blocks bug2. Also it would be great to ni? the assignee of the former bug, or at least cc the person to bug1.

Comment 2

[Julien Pagès (:parkouss)]

Thanks Henrik!

So bugsy currently need support for both CC and and NI,

https://github.com/AutomatedTester/Bugsy/issues/33
https://github.com/AutomatedTester/Bugsy/issues/34

So we'll need to wait for that. Meanwhile, It is useful to print the bug2 if we find it, let's start with that.

Comment 3

[Julien Pagès (:parkouss)]

Attachment: print the bug url that introduced a regression

Comment 4

[William Lachance (:wlach)]

Comment on attachment 8708284 [details] [review]
print the bug url that introduced a regression

I have to admit I'm a bit worried this will generate false positives if the person doesn't know what they're doing when they're using mozregression... but it's worth a try. Let's see how it goes!

Comment 5

[Julien Pagès (:parkouss)]

(In reply to William Lachance (:wlach) from comment #4)
> Comment on attachment 8708284 [details] [review]
> print the bug url that introduced a regression
> 
> I have to admit I'm a bit worried this will generate false positives if the
> person doesn't know what they're doing when they're using mozregression...
> but it's worth a try. Let's see how it goes!

Oh, as for now it is just a message to indicate the bug that may have introduced the regression, I believe it's ok. Though you are right, we may think twice before going further on this. :)

Landed in https://github.com/mozilla/mozregression/commit/efa02112c9fcbfa2a857f356cab7ac5c4e23db06.

Comment 6

[BugBot [:suhaib / :marco]]

The bug assignee didn't login in Bugzilla in the last 7 months.
:zeid, could you have a look please?
For more information, please visit auto_nag documentation.

Comment 7

[Zeid [:zeid]]

This PR appears to have been merged a long time ago.