Discussion:
[Development] Closing bugs as won't fix or out of scope
Tuukka Turunen
2018-11-08 12:05:45 UTC
Permalink
Hi,

I was looking a bit into our older bugs and found quite many that in my opinion could be closed. We do not have a QUIP to define the policy of closing bugs currently, so this is a bit of a matter of viewpoint. In my opinion we should be more active in closing a bug that no-one has an intention of fixing. If the bug later turns out to be relevant to fix, it can always be re-opened.

Some of the bugs I was looking at were

* reported to happen in some exotic device, sometimes also quite old one (by now, if bug has been hanging for a long time)
* reported to happen only in an old version of the operating system
* fixed in a later version of Qt, but left open as bug was not feasible to fix in the version it was reported from
* reported for old version of Qt, and no longer relevant functionality in the current versions
* caused by a bug in the operating system
* maybe possible to fix, but with a risk of causing a regression or a behavior change
* such that would require a major rewrite of the functionality to fix
* not reproduceable any more (or reported by the user not to be able to reproduce)
* marked as duplicate, but both instances of the bug were open/reported state
* not actually bugs, but suggestions for new functionality reported as a bug (should be moved to be a suggestion)

Different persons have unique viewpoints to what kind of bug should be closed and what not. Keeping everything open until the reported issue fixed, is also a valid view, but that leads to a large number of bugs and can hinder getting the right understanding of what is planned to be fixed or finding the important items to fix. Of course sometimes we also simply forget to close a bug or some bug gets fixed in conjunction of other changes (and bug report is not closed).

I am not saying that we should necessarily close immediately all and every bug that match criteria I mentioned in the bullet list. What would be good is a bit more of a common agreement of what to do with the bugs that are not indented to be fixed (or at least very likely not going to be fixed).

Yours,

Tuukka
Tuukka Turunen
2018-11-08 13:22:28 UTC
Permalink
Hi,

Would be great if you add that info as a comment to the bug reports.

Yours,

Tuukka

From: Massimo Callegari <***@yahoo.it>
Date: Thursday, 8 November 2018 at 15.05
To: Qt Project Development Mailing List <***@qt-project.org>, Tuukka Turunen <***@qt.io>
Subject: Re: [Development] Closing bugs as won't fix or out of scope

Hello Tuukka,
on my end you can close the following.

https://bugreports.qt.io/browse/QTBUG-58987
https://bugreports.qt.io/browse/QTBUG-46707
https://bugreports.qt.io/browse/QTBUG-49788
https://bugreports.qt.io/browse/QTBUG-53580
https://bugreports.qt.io/browse/QTBUG-44923

I do believe Qt3D issues need some more love by maintainers.
For example https://bugreports.qt.io/browse/QTBUG-69721 is a plain regression reported on August 1st and completely ignored.

This can also be closed: https://bugreports.qt.io/browse/QTBUG-62578 (not reproduced anymore with recent Qt versions)

Thanks,
Massimo

Il giovedì 8 novembre 2018, 13:05:53 CET, Tuukka Turunen <***@qt.io> ha scritto:
Hi,

I was looking a bit into our older bugs and found quite many that in my opinion could be closed. We do not have a QUIP to define the policy of closing bugs currently, so this is a bit of a matter of viewpoint. In my opinion we should be more active in closing a bug that no-one has an intention of fixing. If the bug later turns out to be relevant to fix, it can always be re-opened.



Some of the bugs I was looking at were

· reported to happen in some exotic device, sometimes also quite old one (by now, if bug has been hanging for a long time)

· reported to happen only in an old version of the operating system

· fixed in a later version of Qt, but left open as bug was not feasible to fix in the version it was reported from

· reported for old version of Qt, and no longer relevant functionality in the current versions

· caused by a bug in the operating system

· maybe possible to fix, but with a risk of causing a regression or a behavior change

· such that would require a major rewrite of the functionality to fix

· not reproduceable any more (or reported by the user not to be able to reproduce)

· marked as duplicate, but both instances of the bug were open/reported state

· not actually bugs, but suggestions for new functionality reported as a bug (should be moved to be a suggestion)



Different persons have unique viewpoints to what kind of bug should be closed and what not. Keeping everything open until the reported issue fixed, is also a valid view, but that leads to a large number of bugs and can hinder getting the right understanding of what is planned to be fixed or finding the important items to fix. Of course sometimes we also simply forget to close a bug or some bug gets fixed in conjunction of other changes (and bug report is not closed).



I am not saying that we should necessarily close immediately all and every bug that match criteria I mentioned in the bullet list. What would be good is a bit more of a common agreement of what to do with the bugs that are not indented to be fixed (or at least very likely not going to be fixed).



Yours,



Tuukka
_______________________________________________
Development mailing list
***@qt-project.org<mailto:***@qt-project.org>
http://lists.qt-project.org/mailman/listinfo/development

Loading...