<!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 6.5.7654.12">
<TITLE>RE: [gdal-dev] Layer operations, a proposal</TITLE>
</HEAD>
<BODY>
<!-- Converted from text/plain format -->
<BR>
<BR>
<P><FONT SIZE=2>Jason Roberts wrote:<BR>
...<BR>
> For scenarios involving large numbers of features, I suspect it is much<BR>
> harder to do it fast and within available memory. It is probably necessary<BR>
> to do a multi-pass approach, where the first step operates only on the<BR>
> spatial indexes of the layers involved. It is probably too slow--or at least<BR>
> very suboptimal--to even read all of the features into memory, much less<BR>
> call GEOS on them; all of that would happen in a later pass, when a subset<BR>
> of features was identified via the spatial index pass....<BR>
<BR>
I was intrigued to read the earlier thread you pointed out Jason, and to see also<BR>
Jan's comment about spatial index handling. I would really love to see a common<BR>
approach to adding spatial index files to my files (and even database connections<BR>
done via a VRT) - and then having OGR understand how to use those indexes.<BR>
<BR>
That'd be pretty awesome in my opinion... not that I'm able to implement anything<BR>
myself, but thought I'd raise a cheer around the idea ;-)<BR>
<BR>
Tyler<BR>
<BR>
</FONT>
</P>
</BODY>
</HTML>