[geos-devel] GEOS robustness question

[Martin Davis]

> It looks to me like fairly complex intersection/topology/graph structures are created

Um, yeah, it's complex processing

> the entire matrix is correctly populated for every predicate call.

Yep.  There is the chance to do some optimization if the predicate called for doesn't require all matrix entries.  For most preds however this won't make much time difference, I believe - the critical step is the intersection determination

> It also appears that intersection points are generated when one wants to know if two shapes cross or intersect.

Nope.  The intersection points are not generated explicitly - only the topology of the intersection is computed (e.g. proper or at endpoint of one or both

> It also appears that if I want to call a 
> predicate on A&B, followed by one on A&C, it is not possible 
> to cache any intermediate processing on A between calls, like 
> the monotone chains, spatial indexes, and the like.

Correct.  Would be nice to open this up so that intermediate results can be cached.  In the plans for JTS...

Thanks for the comments, Kevin.  

Martin Davis, Senior Technical Architect
Vivid Solutions Inc.
Suite #1A-2328 Government Street   Victoria, B.C.   V8T 5G5
Phone: (250) 385 6040    Fax: (250) 385 6046
EMail: mbdavis at vividsolutions.com  Web: www.vividsolutions.com