[postgis-devel] GEOS changes breaking PostGIS testcases

[Sandro Santilli]

On Thu, May 30, 2013 at 03:51:39PM -0700, Paul Ramsey wrote:
> So I get to regress behavior if my change is "better"? 

It really depends on the quality of the test.
I checked all failing tests (they were only in the test for
TopoGeo_addLinestring) and all of them originated by tickets
that complained about being unable to build a topology.

The test checks for the actual topology being built, accepting
debatable arrangements, rather than simply testing for being
able to build a topology.

Now that I write it down I realize that my change didn't fix
that (ie: it keeps checking for output arrangement, while it
shouldn't).

--strk;