[postgis] idea for postgis application

Nedjo Rogers nedjo at miningwatch.org
Mon Mar 18 10:21:23 PST 2002


Alternately or to supplement some sort of vertex filtering (e.g., an
implementation of the Douglas-Peucker line simplification algorithm), a
simple clent-side approach would be to define minimum and maximum display
scales for large data sets and load them selectively (i.e., load only those
features intersecting bounding box), perhaps using a server-side session log
to avoid sending any one feature twice in a session.

If there aren't good reasons otherwise, it would be good to see client
software built around the OpenGIS web feature server specification, i.e.,
querying and loading according to the xml specs given there, e.g., using
Geoserver.

Native browser support for SVG is seemingly still a ways off, but in the
meantime using the Adobe SVG plugin does offer possibilities.  I'm currently
looking at an approach that would use DOM and the Adobe functions parseXML
and postURL to post queries to a WFS and parse the results, drawing on
approaches I've sketched in in my opensvgmapserver.  It turns out that
parsing GML into SVG using DOM is relatively straighforward, as SVG is so
much simpler (e.g., concatenates all path rings into one string,
irrespective of whether they're "inner" or "outer" rings or what polygon
they belong to).  E.g., for polygon or linestring features:

coords=node.getElementsByTagNameNS("gml","coordinates");

then cycle through the coords nodes putting them in path format and
concatenating them into a string to be the "d" attribute of a new path.

Nedjo Rogers

> -----Original Message-----
> From: Tyler Mitchell [mailto:tmitchell at lignum.com]
> Sent: March 13, 2002 4:48 PM
> To: postgis at yahoogroups.com
> Subject: Re: [postgis] idea for postgis application
>
>
>
> Paul wrote:
> >Any interactive viewer would be sweeeeet, although Mac-only might be a
> >bummer for linux'ers like myself. The main issue I forsee w/ interactive
> >viewers is avoiding the "crushing load" problem when people are looking
> >at large extents on large large tables. (All streams in British
> >Columbia, for example.)
>
> Is it realistic to think on doing some sort of on-the-fly filtering of
> features to generalize them depending on scale of the query?  Just a
> though/question for your Paul.  Perhaps a stored function (right
> terminology?) that would filter out various vertex is desired?  I was
> originally thinking this would reduce the "load" at the client end but I
> guess there would be a server-side processing load increase eh?
>
> Tyler
>
>
>
>
> To unsubscribe from this group, send an email to:
> postgis-unsubscribe at yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>


------------------------ Yahoo! Groups Sponsor ---------------------~-->
Tiny Wireless Camera under $80!
Order Now! FREE VCR Commander!
Click Here - Only 1 Day Left!
http://us.click.yahoo.com/nuyOHD/7.PDAA/yigFAA/PhFolB/TM
---------------------------------------------------------------------~->

To unsubscribe from this group, send an email to:
postgis-unsubscribe at yahoogroups.com

 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/ 





More information about the postgis-users mailing list