Re: [QGIS-Developer] Bringing back a *highly curated* blocker tag on redmine?

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: [QGIS-Developer] Bringing back a *highly curated* blocker tag on redmine?

Alessandro Pasotti-2
On Tue, Jan 30, 2018 at 12:22 AM, Nyall Dawson <[hidden email]> wrote:
Hi all,

I'd like to raise the idea of resurrecting the "blocker" tag on
redmine. I understand the reasons behind its original removal, and
fully agree that the new process is working much better.

BUT...

We have two extremely serious issues open against QGIS 3.0, which I
think SHOULD block the release. These are:

https://issues.qgis.org/issues/17985 : Huge memory leak with NULL
attributes in Python
https://issues.qgis.org/issues/17916 : Crashes and instability with Qt 5.10

I don't think we can proceed with any 3.0 release while these two
issues are unsolved.

That's why I'd like to see the blocker tag brought back, but with it
actually BLOCKING the release (vs the previous definition, which
ultimately just meant "hey please fix this because it impacts me a
lot!"). And I'd like to see it highly curated to avoid users tagging
their own pet bugs with it.

Thoughts?


I think that a carefully used BLOCKER tag makes a lot of sense.

There are some bugs that we should definitely not be part of a release or this will seriously impact user's trust in QGIS both as a software and as a community.

I also suggest that the ones who take care of the bug queue (kind of QA managers) should be in charge of managing and lifting the blocker tag.

--
Alessandro Pasotti
w3:   www.itopen.it

_______________________________________________
Qgis-psc mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/qgis-psc
Reply | Threaded
Open this post in threaded view
|

Re: [QGIS-Developer] Bringing back a *highly curated* blocker tag on redmine?

Giovanni Manghi
Hi Jürgen,


> On Tue, 30. Jan 2018 at 09:09:52 +0000, Giovanni Manghi wrote:
>> * a regression that causes data corruption
>> * a regression that causes qgis to crash
>
> I still don't understand this obsession with regressions ;)

:)


> Any bug that is severe should block the release.  Whether it's a regression or
> not doesn't matter.

I can't agree more :)


> If it's an unavoidable bug in some heavy used thing that is new, it's just as
> blocking as a big bug in known territory.
>
> IMHO a bug in some remote, hardly used function shouldn't block a release -
> even if it's a regression that corrupts data and causes qgis to crash in some
> edge cases.  We'll probably have plenty of those that we don't know of anyway.
>
> To me it's just a matter of the impact on usablity a bug has.  I'd opt for
> common sense instead of setting strict rules.


and again... I agree. In fact the example that I brought here (once
more) was about a basic GIS functionality (that unfortunately despite
being broke hasn't stopped us to make new releases).

For such cases I like the idea of having a mini panel (2 people, maybe
better 3 like 1 developer, a power user and the release manager?) to
untie this cases.


-- Giovanni --
_______________________________________________
Qgis-psc mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/qgis-psc
Reply | Threaded
Open this post in threaded view
|

Re: [QGIS-Developer] Bringing back a *highly curated* blocker tag on redmine?

Giovanni Manghi
Hi Nyall!


>> and again... I agree. In fact the example that I brought here (once
>> more) was about a basic GIS functionality (that unfortunately despite
>> being broke hasn't stopped us to make new releases).
>
> What bug is this?

was (hopefully is not "is" anymore) the $area one (which at some point
started to compute bogus areas).

>
>> For such cases I like the idea of having a mini panel (2 people, maybe
>> better 3 like 1 developer, a power user and the release manager?) to
>> untie this cases.
>
> Sounds good to me - just make sure you're included in the panel! :)

if the PSC wants I'm available :)

cheers!

-- g --
_______________________________________________
Qgis-psc mailing list
[hidden email]
https://lists.osgeo.org/mailman/listinfo/qgis-psc