[OSM-dev] Best and FASTEST Spatial DB for only calculating points near to X?

Ander Pijoan ander.pijoan at deusto.es
Sat Dec 15 13:18:06 GMT 2012


2012/12/14 Martin Schafran <martin at ampelmeter.com>

> what happens if X=X matches the half database?
> a lot of comparisons.
>
> filter nodes by bbox
>
> WHERE ST_MakeBox2D(ST_GeomFromText(?, 4326), ST_GeomFromText(?, 4326)) &&
> n.geom
>
> statement.setString(1, lowerLeft.toString());
> statement.setString(2, upperRight.toString());
>
> and then
> DISTANCE<5km
>
2012/12/14 Ian Dees <ian.dees at gmail.com>

> http://postgis.refractions.net/docs/ST_DWithin.html
>

2012/12/15 Derick Rethans <osm at derickrethans.nl>

> The 2D spatial stuff in MongoDB is simple, but it *is* really fast. What
> sort of operations are you thinking of running exactly?
>
>
This is a good point. I know more or less how PostGIS works and how I
should structure the query.

My question was more related to fastness. The type of queries will just be
some "geoNear" queries, just to get all the points within a bounding box
that match certain value.

I know that PostGIS is a better GIS and provides far more features, but
just for geoNear queries and taking into account that MongoDB (for what
I've been reading) is supposed to be more web-scalable and easier to export
to JSON I think I will try booth.

Thank you very much to all for all the advice. I will try booth of them
(and also to ) and hope some testing will determine which suits better.

-- 
Ander Pijoan Lamas
Research Assistant, Deustotech
Computer Science Engineer
University of Deusto
E-mail: ander.pijoan at deusto.es
Phone: +34 664471228
in: http://www.linkedin.com/profile/view?id=162888312
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/dev/attachments/20121215/f1f3734e/attachment.html>


More information about the dev mailing list