[Qgis-psc] Performance regressions (was QGIS 2.18 bug fixing effort)

[Matthias Kuhn]

Hi all,

I remember that some other projects (KDE?) implemented something like
this and it wasn't easy. As far as I remember they have put their own
infrastructure into place to have a stable environment where to execute
their tests.

There are a couple of different ways to measure performance (wall clock,
cpu time) and a couple of external constraints (foremost I/O based) that
have a strong effect on performance measurements.

It's gonna be a tough job to get this right and future proof, just not
to underestimate it :)

Matthias


On 5/24/17 2:07 PM, Nyall Dawson wrote:
> On 24 May 2017 at 22:04, Neumann, Andreas <a.neumann at carto.net> wrote:
>> Hi Nyall, Giovanni,
>>
>> I wonder if we could have/introduce an automated test suite for checking
>> performance regressions (e.g. in rendering, table loading, project loading).
>> Would this be feasible to have? Just wondering.
>>
>> Or is something like this already available and I am not aware of it?
> We don't have anything like this yet, but it's definitely doable. I'd
> say it's an ideal candidate for the next round of grant funding...
>
> Nyall
> _______________________________________________
> Qgis-psc mailing list
> Qgis-psc at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/qgis-psc
>