[Qgis-psc] Rethink using github discussions?

[Nyall Dawson]

On Mon, 8 Feb 2021 at 09:25, Giovanni Manghi <giovanni.manghi at gmail.com> wrote:
>
> Hi Nyall,
>
> discussions were activated recently by Tim, if I'm not wrong.
>
> I have no strong opinions about it,  one good thing is that we can move tickets (from the bug tracker, for example any time a user wrongly post a question instead of a report) to discussions. We cannot move questions wrongly posted in the tracker to SE. Not saying that this alone justify keeping discussions, just pointing that is a nice feature.

Yeah, I can definitely concede that point.

But on the other hand a question/issue like
https://github.com/qgis/QGIS/discussions/41400 should really just be
closed. That user will NEVER get the answer they are looking for on
either a QGIS ticket or the github discussion. There's a (small)
chance they'll get an answer on the dev list, but in my experience
with questions of this nature it's extremely unlikely. So I think
really the best response to a ticket like this would be closing with a
reply like:

"This tracker is for QGIS issues only. For questions of this nature
please use gis.stackexchange, the QGIS dev mailing list or QGIS user
mailing. If you're still searching for answers, we suggest engaging a
QGIS Commercial Support provider to assist with your solution".

That's my 2c at least :)

Nyall

>
> cheers
>
> -- G --
>
>
>
>
> On Sun, Feb 7, 2021 at 11:20 PM Nyall Dawson <nyall.dawson at gmail.com> wrote:
>>
>> Hi PSC!
>>
>> I'd like to raise discussions about whether or not it's a good move to
>> be allowing the github discussions feature on the QGIS repo:
>>
>> https://github.com/qgis/QGIS/discussions
>>
>> I may be pessimistic, but I just can't see this Stackexchange clone
>> gathering the same momentum and expose as gis.stackexchange enjoys. So
>> my concern is that by allowing users to ask questions here we are just
>> fragmenting the support base and will end up with a whole set of
>> unanswered questions on github, simply because there's no-one there
>> who is motivated to answer these questions.
>>
>> In contrast, there's LOTS of informed users answering all the QGIS
>> questions on gis.stackexchange.
>>
>> My summary:
>>
>> Stackexchange: lots of users, including many from outside the regular
>> community. Almost all QGIS questions end up with a knowledgeable
>> answer (and if not, there's enough knowledgeable users to vote down
>> bad answers). Already a proven working system for end-user QGIS
>> support. QGIS is one of the most heavily used tags on the site, and
>> the good answers present QGIS in a very positive light.
>>
>> Github discussions: basically no users, and the only ones looking are
>> a subset of the very small, overworked usual crowd. Not a single
>> question has an answer. It just looks bad and unmaintained, and
>> presents the QGIS project in the poor light of a stagnant open-source
>> project which no-one cares about.
>>
>> Do we REALLY need this feature? Is there any compelling reason to keep
>> it active?
>>
>> Nyall
>> _______________________________________________
>> Qgis-psc mailing list
>> Qgis-psc at lists.osgeo.org
>> https://lists.osgeo.org/mailman/listinfo/qgis-psc