[postgis-users] caching distance calculations

[Matthew Rushton]

Great thanks, that makes sense. We're a startup building a mobile LBS and are dealing with a bunch of fixed points (as well as some moving ones!).
-Matt


--- On Mon, 12/22/08, Paul Ramsey <pramsey at cleverelephant.ca> wrote:

> From: Paul Ramsey <pramsey at cleverelephant.ca>
> Subject: Re: [postgis-users] caching distance calculations
> To: mrushton7 at yahoo.com, "PostGIS Users Discussion" <postgis-users at postgis.refractions.net>
> Date: Monday, December 22, 2008, 11:58 AM
> Totally up to you... depends on how it fits your
> application. It's not
> a bad design pattern, by any means, and it does provide
> guaranteed
> performance, assuming your data doesn't change too
> much.
> 
> What are you doing this distance calculation set in aid of?
> 
> P
> 
> On Mon, Dec 22, 2008 at 8:52 AM, Matthew Rushton
> <mrushton7 at yahoo.com> wrote:
> > Hi,
> >  We recently switched from mysql to postgres to get
> the great spatial support and are looking to leverage it a
> bit more. One area we are looking at would require a great
> number of distance calculations (from point to polygon) to
> be made frequently which we think would be a huge
> bottleneck. I imagine this a common performance issue others
> have hit. What is the recommended course of action? Does the
> database do any caching for example on ST_Distance
> calculations? My assumption is no. If that's the case
> does it make sense to do the calculations once and store
> them in the db (in our case it's the same large set of
> distance calculations that are made)? I'm just wondering
> if this is a common practice or discouraged? Thanks.
> > -Matt
> >
> >
> >
> > _______________________________________________
> > postgis-users mailing list
> > postgis-users at postgis.refractions.net
> >
> http://postgis.refractions.net/mailman/listinfo/postgis-users
> >