[Qgis-developer] Failing tests consider blockers
madmanwoo at gmail.com
Wed Feb 19 00:45:00 PST 2014
One issue I see here is we are mixing unit tests with integration tests.
The tests for QgsExpression are examples of unit tests and should flag a
regression but stuff like rendering tests are more if a integration test
and may or may not flag a regression.
On 19/02/2014 6:42 pm, "Martin Dobias" <wonder.sk at gmail.com> wrote:
> On Wed, Feb 19, 2014 at 3:14 PM, Radim Blazek <radim.blazek at gmail.com>
> > On Sun, Feb 16, 2014 at 11:41 PM, Jürgen E. <jef at norbit.de> wrote:
> >> What I wanted to say is that there are a bunch of tests that fail
> because they
> >> are not up to date or fail for other reasons that are not related to
> >> bugs in qgis itself. For example render tests, which fail because of
> fonts or
> >> other differences (eg. removed renderers). I think that are the
> majority of
> >> tests that currently fail.
> > Platform dependent tests (fonts) don't make sense for me and should be
> > disabled until get fixed (e.g. using font set shipped with QGIS).
> > Renderer test should be updated and only real bugs will remain and
> > those should be blockers, I think.
> Let me add that in case of tests it will most likely not even help to
> ship a common font with QGIS: on different platforms the text is
> rendered by different libraries (e.g. FreeType on linux, native
> backends on windows/mac).
> Also, renderer tests should IMHO have some tolerances instead a set of
> anomalies. For example, in multi-threading branch some tests started
> to fail because of some microscopic differences - this will require
> another set of reference images (and anomaly images) although there is
> no real difference. This looks like an interesting tool for comparing
> Qgis-developer mailing list
> Qgis-developer at lists.osgeo.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Qgis-developer