[geos-devel] CascadedPolygonUnion XML testing

[G. Allegri]

> For the record, using SnapToGrid gives you more control over the
> roundup. 1e-6 "gridding" works in the exp1_red2 case too:

Well, this is like using a FIXED precision model with scale 1e6, if
I'm not wrong. Anyway I would expect a FLOATING model to work well
too, as it does in JTS. I've tested HEXWKB from exp1_red2, and it
works fine, so I think the result in JTS doesn't depend on the lossy
wkt representation.