#POSTGIS IRC Log - 2009-04-22

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
05:47:54 CIA-34: mloskot * r4016 /spike/wktraster/scripts/Makefile.rt.sample: scripts/Makefile.rt.sample: rtpostgis.sql lives in /contrib not /rt_pg/contrib.
08:30:54 strk: does anyone here work with ArcGIS/PostGIS ?
08:38:50 dassouki: like linking the two ?
08:39:00 dassouki: arcgis installs on my ubuntu but doesn't want to run :(
08:40:09 strk: well, yes, using arcgis to manage postgis databases
09:03:43 moseA: Has anyoen been following this thread on pgsql-performance? "[PERFORM] GiST index performance"
09:52:00 pramsey: moseA: does it get more interesting? the thread ends with the poster saying there was a big java process on his machine that might have been getting in the way of things
10:27:18 dassouki: is there a way to do a disolve based on zone_id field, while adding the values in number_of_residences_per_zone field?
10:29:27 CIA-34: robe * r4017 /trunk/doc/xsl/postgis_gardentest.sql.xsl: revise to include a version variable and get rid of limit ( think its causing different tests to be run non-consistently depending on ordering of postgresql so hard to compare outputs )
10:29:33 pramsey: select zone_id, st_union( the_geom ) as the_geom, sum( number_of_residences_per_zone ) as number_of_residences from thetable group by zone_id;
10:29:49 dassouki: thank you :D
14:36:27 moseA: pramsey: there was some discussion in the thread about a bug in the picksplit function in the cube and seg contrib modules, so I think it is more interesting than a runaway JVM... I have had problems with the R-tree via GiST in Postgis ( unexpectedly bad performance ), but um, the picksplit function in lwgeom_gist.c is clearly quite different so I apologize for mentioning it : )
23:52:17 sigq: geosfeed: Ticket #215 ( enhancement updated ): Single-sided buffering implementation <http://trac.osgeo.org/geos/ticket/215#comment:7>