Re: [boost] Another GGL review

Subject: Re: [boost] Another GGL review
From: Mateusz Loskot (mateusz_at_[hidden])
Date: 2009-11-17 07:43:44

Jose wrote:
> Also, most mentions to indexes are memory indexes. Are there any
> solutions for persistent spatial indexes ? maybe by interfacing to an
> external library

Jose,

This is a very interesting subject indeed. I would be interested
in solving spatial indexing in generic way myself.

(I'm sorry for constantly referring to GIS domain,
an occupational disease, clearly.)
 From GIS domain perspective, regarding persistent spatial index, I've 
observed popularity of two solutions based on Open Source implementations:
- PostgreSQL GiST as foundation of spatial index for PostGIS [1]
- simple single file-based (.qix) quadtree index [2] and [3]

> Do you see state-of-the-art c++ libraries like mapnik potentially
> using GGL, instead of their abstractions ?

I am not a member of Mapnik team, however I have spoken to Artem 
Pavlenko long time ago about generic geometry library and he
was confirming high interest in this subject.
In those times, the only usable alternative for GIS was GEOS.
Artem was not very enthusiastic about using GEOS. Given that,
I would assume Mapnik project would be a potential user of GGL.

[1] http://postgis.refractions.net/docs/ch04.html
[2] http://trac.osgeo.org/mapserver/wiki/ShpTree
[3] http://www.mapserver.org/utilities/shptreevis.html

Best regards,
-- 
Mateusz Loskot, http://mateusz.loskot.net
Charter Member of OSGeo, http://osgeo.org