| 00:48:49 | CIA-34: | strk * r2348 /trunk/source/ ( 2 files in 2 dirs ): Port OffsetCurveBuilder from JTS-1.9. Twenty time faster completion of fme.xml testcase ! |
| 01:10:53 | CIA-34: | strk * r2349 /trunk/source/ ( 2 files in 2 dirs ): Sync BufferSubgraph to JTS-1.9 |
| 01:15:52 | CIA-34: | strk * r2350 /trunk/source/ ( 2 files in 2 dirs ): Update port info for RightmostEdgeFinder ( checked ) |
| 01:18:40 | CIA-34: | strk * r2351 /trunk/source/ ( 2 files in 2 dirs ): Update port info of SubgraphDepthLocater |
| 01:50:20 | CIA-34: | strk * r2352 /trunk/tests/xmltester/ ( 10 files ): Fix XML for Buffer testing to match jts layout ( arg2 for distance, arg3 for quadrant segments ); copy buffer-related tests from jts: they succeed w/out editing. |
| 02:23:59 | CIA-34: | strk * r2353 /trunk/tests/xmltester/XMLTester.cpp: Add support for bufferMitredJoin tests |
| 02:46:36 | CIA-34: | strk * r2354 /trunk/source/operation/buffer/BufferBuilder.cpp: Fix memory leak in BufferBuilder ( due to just-ported short-circuit ) |
| 02:59:33 | CIA-34: | strk * r2355 /trunk/tests/xmltester/ ( 50 files in 2 dirs ): Move XML tests under their own directory |
| 03:01:27 | vik00: | ciao ^_^ |
| 03:02:05 | vik00: | somebody can help me ? |
| 04:39:01 | CIA-34: | strk * r2356 /trunk/source/ ( 2 files in 2 dirs ): Update port info, checked against JTS-1.9 |
| 04:51:56 | CIA-34: | strk * r2357 /trunk/source/ ( 3 files in 3 dirs ): Fix memory leak in RayCrossingCounter; update signatures to avoid pointers when unneeded; add a locatePointInRing taking a vector of coordinate pointers, for use by CGAlgorithms |
| 05:19:26 | CIA-34: | strk * r2358 /trunk/source/ ( 2 files in 2 dirs ): Sync CGAlgorithms with JTS-1.9 |
| 05:31:40 | CIA-34: | strk * r2359 /trunk/source/ ( 2 files in 2 dirs ): Sync ( Robust )LineIntersector to JTS-1.9 |
| 06:21:56 | CIA-34: | mcayland * r3999 /trunk/loader/shp2pgsql-core.c: Try and fix shp2pgsql-core for bug #124. Need to change sprintf( ) to pgis_exec( ) within the loader. |
| 07:22:16 | CIA-34: | strk * r2360 /trunk/source/ ( 13 files in 2 dirs ): Add CoordinateSequenceFilter support, fix default GeometryComponentFilter moving the logic to the correct place ( a Geometry private class ). |
| 08:39:58 | CIA-34: | strk * r2361 /trunk/ ( 13 files in 6 dirs ): Port the algorithm::distance package from JTS 1.9 |
| 11:07:47 | CIA-34: | strk * r2362 /trunk/ ( 3 files in 3 dirs ): Temporarly revert the unrolled computation in HCoordinate::intersection. Added mitred join buffer test from JTS ( failed with the unrolled comp, to be further inspected ). |
| 15:02:29 | strk: | pramsey: numerical stability strikes back ( remember the infamous HCoordinate results changing based on compiler flags ? ) |
| 15:02:53 | pramsey: | somewhat. does numerical stability ever go away? : ) |
| 15:17:39 | strk: | anyway, avoiding the "unrolled" computations in HCoordinates seems to "fix" the new testcases |
| 15:17:58 | strk: | where can I find the dataset you've used in Toronto for performance tests ? |
| 15:18:15 | strk: | I'd like to run some buffering on those : ) |
| 15:18:39 | strk: | have you read about the x20 speed improvement ? |
| 16:10:23 | CIA-34: | strk * r2363 /trunk/source/operation/buffer/BufferInputLineSimplifier.cpp: Fix typo in computing angleOrientation |
| 16:24:07 | CIA-34: | strk * r2364 /trunk/source/algorithm/HCoordinate.cpp: Explain why unrolled computation is turned off, and keep it off |
| 17:07:27 | CIA-34: | robe * r4000 /trunk/doc/reference.xml: amend ST_ContainsProperly to summarize its speed advantages |
| 17:18:29 | CIA-34: | robe * r4001 /trunk/doc/reference.xml: add link to st_relate |
| 20:27:44 | robink: | I'm having difficulty running the lwpostgis.sql init script |
| 20:27:56 | robink: | psql:lwpostgis.sql:4093: ERROR: could not find function "coveredby" in file "/usr/lib/postgresql//liblwgeom.so.1.3" |
| 20:31:30 | robink: | /usr/lib/postgresql//liblwgeom.so.1.3 exists. |
| 21:37:20 | robink: | Oh, I think I know what's going on. |
| 22:59:38 | robink: | Hello strk_off. |
| 23:06:32 | strk_off: | good morning |
| 23:10:59 | robink: | Yup, fixed |
| 23:11:01 | robink: | hooray |
| 23:34:09 | rudenstam: | when creating tables, one should create the table without geometry and then add the geometry coloumn later ? As postgres seems unaware of the datatype LINESTRING.. |
| 23:39:39 | strk: | w/out the geometry |
| 23:39:47 | strk: | and use AddGeometryColumn |
| 23:40:05 | strk: | anyway, the datatype is GEOMETRY in any case |
| 23:40:30 | strk: | for what concerns postgresql |
| 23:40:52 | strk: | only the AddGeometryColumn adds constraints to make sure the internal types are the ones you specify ( if you speciy one ) |
| 23:43:20 | rudenstam: | as I thought then, but was a bit unsure so thought I'd get it 100% right and not waste hours doing something dodgy |
| 23:44:54 | rudenstam: | thanks at least |
| 23:48:29 | CIA-34: | strk * r2365 /trunk/source/operation/valid/IndexedNestedRingTester.cpp: Fix memory leaks in IsValid operation |