Thanks for your elaborate answer,, I will have a good think about it. Just a random thought I had: As I was doing some experiments with st_intersection, I noticed some strange thing though. On our prod. machine (windows, pg11, postgis 2.5.0 (not 100% sure), geos 3.7.0), it takes about a minute to calculate the st_intersection, on my machine (windows only using libgeos) it takes 11 seconds after my optimization (15seconds without) and according to Martin Davis at his end (java?) it takes about 4 seconds. I don't have any macros defined (I saw f.e. a USE_RECTANGLE_INTERSECTION), could that be the cause of the speed-differences? And can it be verified that all code/compiler optimizations are enabled for the windows builds of postgis/geos?
P.S. can you include me in the cc?
_______________________________________________ postgis-users mailing list [email protected] https://lists.osgeo.org/mailman/listinfo/postgis-users
