<HTML><HEAD><TITLE>[mapguide-users] OGR / GEOS selection problem</TITLE>
<META http-equiv=Content-Type content="text/html; charset=utf-8">
<META content="MSHTML 6.00.6000.16441" name=GENERATOR></HEAD>
<BODY dir=ltr>
<DIV>Hi Kenneth,</DIV>
<DIV> </DIV>
<DIV>Thanks for tracking this down. I would try to fix this by bypassing Geos. FDO has utility code that can check for an intersection. We could take the BBOX result and pass that through a real intersection check on the provider side. This way the intersection check would be performed only on geometries that are close enough to the query polygon so the overhead should not be too bad.</DIV>
<DIV> </DIV>
<DIV> </DIV>
<DIV>Traian</DIV>
<DIV> </DIV>
<BLOCKQUOTE dir=ltr style="MARGIN-RIGHT: 0px">
<DIV><FONT size=2>-----Original Message----- <BR><B>From:</B> mapguide-users-bounces@lists.osgeo.org on behalf of Kenneth, GEOGRAF A/S <BR><B>Sent:</B> Thu 6/7/2007 2:54 AM <BR><B>To:</B> MapGuide Users Mail List <BR><B>Cc:</B> <BR><B>Subject:</B> [mapguide-users] OGR / GEOS selection problem<BR><BR></FONT></DIV>
<P><FONT size=2>Maciej previously reported problems with the OGR provider using PostGIS.<BR><BR>The problem is with the OGR provider.<BR>Se the Trac ticket for details:<BR><A href="http://trac.osgeo.org/mapguide/ticket/172">http://trac.osgeo.org/mapguide/ticket/172</A><BR><BR>I have posted some follow up on the issue, but in hope someone knows a<BR>lot about GEOS, I'm also posting it here:<BR><BR><BR>I have been examining the problem, using a custom build of the OGR<BR>provider. It is true that without GEOS support in OGR, the Intersect<BR>only does a BBOX check.<BR><BR>However, merely enabling GEOS will not work, as the viewer sends a<BR>square that is orthogonal to the axes (like a BBOX). The GEOS library<BR>will interpret this as another BBOX, and only perform the BBOX check<BR>anyway.<BR><BR>When I change the polgyon sent from the viewer to a triangle, it takes<BR>22 seconds to calculate the Intersects result with a polygon with 8000<BR>linesegments. I have a 2+ GHz machine.<BR><BR>It seems that GEOS always builds a full intersection matrix, before<BR>returning the result. This is obviously not optimal for a simple hit<BR>test. Even with this knowledge, I still don't see how it can take 22<BR>seconds to calculate 8000 * 3 intersections.<BR><BR>Can someone explain this, or how to fix it?<BR><BR><BR><BR>--<BR>Regards, Kenneth, GEOGRAF A/S<BR><BR><BR>_______________________________________________<BR>mapguide-users mailing list<BR>mapguide-users@lists.osgeo.org<BR><A href="http://lists.osgeo.org/mailman/listinfo/mapguide-users">http://lists.osgeo.org/mailman/listinfo/mapguide-users</A><BR></FONT></P></BLOCKQUOTE></BODY></HTML>