<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7653.38">
<TITLE>RE: [geos-devel] 3.1.0rc2</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<P><FONT SIZE=2>This is just an observation. I've seen a lot of these type failures in my garden tests and just ignored them.<BR>
<BR>
This for example happens often if you in PostGIS create a MULTIPOLYGON by collecting valid polygons that happen to intersect in some way<BR>
and then pass this to ST_Union. Those are invalid anyway so of course its not that big of a surprise they fail.<BR>
<BR>
I'm just wandering if maybe as part of intermediary process, the union algorithm is creating such invalid multipolygons<BR>
<BR>
-----Original Message-----<BR>
From: geos-devel-bounces@lists.osgeo.org on behalf of Hartmut Kaiser<BR>
Sent: Fri 1/30/2009 6:35 PM<BR>
To: 'GEOS Development List'<BR>
Subject: RE: [geos-devel] 3.1.0rc2<BR>
<BR>
The union bug seems to be longstanding. The old process also failed.<BR>
It is a mystery to me why we can't produce pairwise failures though.<BR>
<BR>
It might be, that you're seeing an error from unioning geometries being<BR>
intermediary results from preceding union operations. The cascaded union<BR>
algorithm doesn't 'grow' the result sequentially, but builds several<BR>
sub-geometries based on proximity criteria of the initial geometries and<BR>
'grows' those. As a result the sequence of pair-wise union operations is<BR>
completely different from the naïve union algorithm.<BR>
<BR>
Regards Hartmut<BR>
<BR>
><BR>
> P<BR>
><BR>
> On 30-Jan-09, at 1:52 PM, Sean Gillies <sgillies@frii.com> wrote:<BR>
><BR>
> > Fine here. That cascaded union problem remains.<BR>
> ><BR>
> > On Jan 30, 2009, at 1:18 PM, Paul Ramsey wrote:<BR>
> ><BR>
> >> I've placed RC2 on the download site, with Matz changes and the<BR>
> >> removal of stoopid debugging noise in the last one.<BR>
> >><BR>
> >> <A HREF="http://download.osgeo.org/geos/geos-3.1.0rc2.tar.bz2">http://download.osgeo.org/geos/geos-3.1.0rc2.tar.bz2</A><BR>
> >><BR>
> >> PSC, if you don't howl about this one, I'm going to make the link<BR>
> >> public on Monday morning.<BR>
> >><BR>
> >> --<BR>
> >> Paul Ramsey<BR>
> >> OpenGeo - <A HREF="http://opengeo.org">http://opengeo.org</A><BR>
> >> PostGIS is Love.<BR>
> >><BR>
> >> _______________________________________________<BR>
> >> geos-devel mailing list<BR>
> >> geos-devel@lists.osgeo.org<BR>
> >> <A HREF="http://lists.osgeo.org/mailman/listinfo/geos-devel">http://lists.osgeo.org/mailman/listinfo/geos-devel</A><BR>
> ><BR>
> ><BR>
> ><BR>
> > --<BR>
> > Sean Gillies<BR>
> > sean.gillies@gmail.com<BR>
> > <A HREF="http://sgillies.net">http://sgillies.net</A><BR>
> ><BR>
> > _______________________________________________<BR>
> > geos-devel mailing list<BR>
> > geos-devel@lists.osgeo.org<BR>
> > <A HREF="http://lists.osgeo.org/mailman/listinfo/geos-devel">http://lists.osgeo.org/mailman/listinfo/geos-devel</A><BR>
> _______________________________________________<BR>
> geos-devel mailing list<BR>
> geos-devel@lists.osgeo.org<BR>
> <A HREF="http://lists.osgeo.org/mailman/listinfo/geos-devel">http://lists.osgeo.org/mailman/listinfo/geos-devel</A><BR>
<BR>
_______________________________________________<BR>
geos-devel mailing list<BR>
geos-devel@lists.osgeo.org<BR>
<A HREF="http://lists.osgeo.org/mailman/listinfo/geos-devel">http://lists.osgeo.org/mailman/listinfo/geos-devel</A><BR>
<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>
<HTML><BODY><P><hr size=1></P>
<P><STRONG>
The substance of this message, including any attachments, may be confidential, legally privileged and/or exempt from disclosure pursuant to Massachusetts law. It is intended solely for the addressee. If you received this in error, please contact the sender and delete the material from any computer.
</STRONG></P></BODY></HTML>
<P><hr size=1></P>
<P><STRONG><font size="2" color="339900"> Help make the earth a greener place. If at all possible resist printing this email and join us in saving paper. </p> <p> </font></STRONG></P>