[QGIS-Developer] [Qgis-psc] QGIS Grant: Expand unit test coverage for geometry classes

[Enrico Ferreguti]

Beyond the report on an accomplished task, I see the joy and the passion on
making something good. I can't explain how much your efforts on taking care
of QGIS make my everyday work better.

Thank you very much.

Il 26/set/2017 06:40, "Tim Sutton" <tim at kartoza.com> ha scritto:

> Hi Nyall
>
>
> Thank you so much for doing this work and for your report back! I hope
> this is just one of many times that we will be able to fund our community
> members to make substantial improvements to the quality of QGIS!
>
> Regards
>
> Tim
>
> > On 26 Sep 2017, at 9:31 AM, Nyall Dawson <nyall.dawson at gmail.com> wrote:
> >
> > Hi PSC,
> >
> > Just reporting in that with PR #5254 [1], the grant work has now been
> > completed and we're sitting as close as possible to 100% unit test
> > coverage for the geometry primitive classes (i.e. the classes which
> > represent the raw geometry types, such as points/multipoints/circular
> > strings/ etc).
> >
> > While the purpose of this work wasn't directly to fix bugs, writing
> > the tests has allowed me to identify and fix dozens of issues in these
> > geometry classes. The fixes range from pedantic (e.g. crashes when
> > requesting parts of the geometry with invalid indexes such as negative
> > numbers), through to severe (a geometry corruption caused after
> > converting a MultiLineString to GML which would definitely cause QGIS
> > server to crash on certain requests).
> >
> > Just to ensure that there's no misconceptions here: having 100% test
> > coverage doesn't mean the geometry classes are now bug free - rather
> > it means that the current logic is protected so that future features,
> > fixes, or optimisations will not cause regressions from the current
> > behaviour.
> >
> > I'd like the thank the PSC once again for accepting this funding
> > request (and the sponsors who made it possible!). In my opinion this
> > was a crucial piece of work required to help make future QGIS releases
> > much less likely to suffer serious regressions, so I'm very thankful
> > that you allowed it to happen!
> >
> > Regards,
> > Nyall
> >
> >
> > [1] https://github.com/qgis/QGIS/pull/5254
> > _______________________________________________
> > Qgis-psc mailing list
> > Qgis-psc at lists.osgeo.org
> > https://lists.osgeo.org/mailman/listinfo/qgis-psc
>
> _______________________________________________
> QGIS-Developer mailing list
> QGIS-Developer at lists.osgeo.org
> List info: https://lists.osgeo.org/mailman/listinfo/qgis-developer
> Unsubscribe: https://lists.osgeo.org/mailman/listinfo/qgis-developer
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/qgis-developer/attachments/20170926/31efdd2f/attachment-0001.html>