Thanks for the replies. Re: the order of the intersection points. the only difference in the output from the example they shared was that there were 3 multilinestrings, each of which contained 2 linestrings which were ordered differently.
Let's see whether normalize solves this for them. Andrew On Wed, Jan 6, 2021 at 7:36 PM Paul Ramsey <pram...@cleverelephant.ca> wrote: > > > > On Jan 6, 2021, at 5:31 PM, Andrew Hershberger < > andrew.d.hershber...@gmail.com> wrote: > > > > A GEOSwift consumer noticed some differences in intersection behavior > between geos 3.7.1 and 3.9.0. I'm guessing this is related to OverlayNG? > https://github.com/GEOSwift/GEOSwift/issues/212 > > Yes, component ordering in collections, ring start/end points, interior > ring ordering, it's all potentially changed in 3.9 with the NG overlay. You > can retain tests that work across new/old versions by applying the > GEOSNormalize() function to everything, to get a deterministic ordering of > rings/components/start points etc. > > > I'm looking for guidance on how to advise them w.r.t. their statement > that they're relying on "the order of the intersection points". Is that > part of the contract such that changes should be considered breaking? > > Not sure what that means, maybe they are intersecting lines and getting > points? Would need to know the precise use case to know if it's something > we used to try and preserve. > > P. > > > > > > > Andrew > > _______________________________________________ > > geos-devel mailing list > > geos-devel@lists.osgeo.org > > https://lists.osgeo.org/mailman/listinfo/geos-devel > > _______________________________________________ > geos-devel mailing list > geos-devel@lists.osgeo.org > https://lists.osgeo.org/mailman/listinfo/geos-devel >
_______________________________________________ geos-devel mailing list geos-devel@lists.osgeo.org https://lists.osgeo.org/mailman/listinfo/geos-devel
