#POSTGIS IRC Log - 2009-04-17

For logs after Feb 3, 2007, all times are GMT-8. Prior logs are GMT-9.
Back to Logs
00:01:59 bbgeos: build #52 of telascience-quick started including [2379]
00:25:45 mloskot: bbgeos: status
00:25:45 bbgeos: telascience-full: idle, last build 9m56s ago: failed test-xml
00:25:45 bbgeos: telascience-quick: building( compiling )
00:25:45 bbgeos: telascience-stable: idle
00:26:00 mloskot: bbgeos: notify
00:26:00 bbgeos: try 'notify on|off|list <EVENT>'
00:26:03 mloskot: bbgeos: notify on
00:26:03 bbgeos: The following events are being notified: ['started', 'finished']
00:26:54 strk: mloskot: could you try upgrading libtool in there ?
00:27:24 mloskot: strk: I don't have access to ssh
00:27:28 mloskot: and no time ATM
00:27:31 mloskot: I can do it tomorrow
00:27:42 strk: thanks
00:37:11 strk: now, this is weird
00:37:23 strk: right before the flags checking I get:
00:37:33 strk: ----NOTE----- ac_compile=$CC -c $CFLAGS $CPPFLAGS conftest.$ac_ext >&5
00:37:33 strk: ----NOTE----- CC=gcc
00:37:38 strk: which looks all fine
00:41:51 strk: now this is interesting
00:41:52 strk: ++ echo 'configure:23713: gcc -c -g -O2 -ffloat-store conftest.c >&5'
00:41:59 strk: configure *DOES* use gcc
00:42:04 strk: only conftest.c contains fortran code !!
00:42:10 strk: + printf ' subroutine t\n return\n end\n'
00:42:12 strk: doh
00:44:38 : * mloskot has no idea
00:51:49 strk: no dice, really looks like a newer libtool is needed
00:51:57 strk: or autoconf or I've no idea
00:52:42 mloskot: strk: f**k libtool and forget it
00:52:57 strk: I'll try an hack
00:53:03 mloskot: we've been wasting tons of hours recently, to fix libtool issues for postgis, geos and even gdal/ogr
00:53:16 mloskot: this is ridiculous or we all are members of S/M club or something
00:53:25 mloskot: jiz, for god sake
00:53:38 mloskot: FOSS need to go and have a therapy, they like to waste their time
00:53:54 mloskot: listen to Glynn from GRASS, really
00:54:15 mloskot: he is very reasonable saying that if GRASS start uses libtool, he will run away
00:54:40 mloskot: emotionally, I care because geos is important to many of us, but practically I don't care
00:55:06 mloskot: CMake rulez :- )
00:57:20 mloskot: What I know is that it's impossible to update libtool on buildbot.osgeo.org unless we are 101% sure it won't break other software which is being built on it
00:57:43 mloskot: I'm not sure, so I'm not going to take responsibility for potentially breaking them.
01:10:21 jef: mloskot: does Glynn agree on CMake?
01:10:44 mloskot: jef: no idea
01:11:32 mloskot: jef: http://grass.itc.it/pipermail/grass5/2002-April/005065.html
01:11:33 sigq: Title: [GRASS5] Shared Libraries ( at grass.itc.it )
01:11:39 jef: mloskot: I doubt that - at least not enough to switch GRASS to cmake.
01:12:42 mloskot: jef: it's not about switching to anything, it's about using plain makefiles and forget the troublemaking libtool
01:12:53 mloskot: cmake is just a meta-builder that generates makefiles
01:13:04 mloskot: instead of using cmake, you can craft them manually, matter of choice
01:13:44 jef: mloskot: or have cmake create nmake files...
01:13:56 mloskot: jef: rather vcproj files
01:14:11 jef: mloskot: sure
01:14:36 mloskot: another option is bjam from boost build, does the same, but has terrible syntax
01:21:27 jef: mloskot: oh, I just realize, that I didn't follow up on http://lists.osgeo.org/pipermail/grass-dev/2009-March/043099.html
01:21:28 sigq: Title: [GRASS-dev] Re: [GRASS-windows] windows qgis-grass: make grass build with vc or keep a separate qgis-grass mingw build? ( at lists.osgeo.org )
01:52:27 CIA-34: strk * r2380 /trunk/configure.in: Add debugging output of variables used by AC_LIBTOOL_COMPILER_OPTION. Surprisingly, the buildbots try to use a fortran compiler there
01:55:08 CIA-34: mloskot * r4007 /spike/wktraster/scripts/gdal2wktraster.py: gdal2wktraster.py: Print WARNING messages only in verbose mode.
01:55:41 bbgeos: build #31 of telascience-full started including []
01:55:56 strk: no clue
01:56:05 strk: I kind of found an hack, but I'm not sure I want that in
01:58:34 strk: http://buildbot.osgeo.org:8506/builders/telascience-full/builds/31/steps/configure/logs/stdio
01:58:35 sigq: Title: Log File contents ( at buildbot.osgeo.org:8506 )
01:58:42 strk: NOTE ---- lt_simple_compile_test_code= subroutine t\n return\n end\n
01:58:46 strk: ^^ that's the culprit
01:58:51 strk: NOTE ---- compiler=g77
01:59:01 strk: NO clue on the reason for that braindamage
01:59:38 strk: forcing lt_simple_compile_test_code to a C code fixes it ..
01:59:58 strk: but very fragile hack
02:07:27 bbgeos: build #53 of telascience-quick started including [2380]
03:20:49 CIA-34: mloskot * r4008 /spike/wktraster/scripts/gdal2wktraster.py: gdal2wktraster.py: Fixed warnings formatting.
03:26:29 CIA-34: mloskot * r4009 /spike/wktraster/ ( rt_core/rt_api.h rt_pg/pgsql_compat.h rt_pg/rt_pg.h ): A bit of uniqueness of inclusion guards.
04:48:37 mloskot: L'ds and Gents, bands truncation bug has been confirmed as fixed
04:48:38 mloskot: http://osgeo.pastebin.com/m1a77953f
04:48:39 sigq: Title: osgeo private pastebin - collaborative debugging tool ( at osgeo.pastebin.com )
05:49:37 gzur: Hi guys, I've been using ST_Boundary( geom ) on Linestrings to exctract a bounding box
05:49:38 gzur: and it works fine
05:49:57 gzur: but when I try the same with a polygon I get the whole polygon ( sometimes hundreds of points )
05:50:11 gzur: is there some other function I should be using on polygons?
05:52:28 bbgeos: build #32 of telascience-full started including [2379, 2380]
05:55:21 CIA-34: robe * r4010 /trunk/doc/xsl/postgis_gardentest.sql.xsl: add ST_MinimumBoundingCircle to 1.3 exclude
05:55:42 CIA-34: strk * r2381 /trunk/ ( 3 files in 3 dirs ): renamed intersection constants
06:10:43 bbgeos: build #54 of telascience-quick started including [2381]
06:31:56 bbgeos: build #33 of telascience-full started including []
07:31:12 CIA-34: strk * r2382 /trunk/source/headers/geos/geom/BinaryOp.h: Header guard
07:33:47 CIA-34: strk * r2383 /trunk/ ( 3 files in 2 dirs ): proper inline check_valid source/headers/geos/geom/BinaryOp.h
07:43:22 CIA-34: strk * r2384 /trunk/tests/xmltester/ ( BufferResultMatcher.cpp BufferResultMatcher.h XMLTester.cpp ): Port BufferResultMatcher. Gives 31 new failures !!
07:58:22 bbgeos: build #55 of telascience-quick started including [2382, 2383, 2384]
08:22:35 CIA-34: strk * r2385 /trunk/source/ ( 2 files in 2 dirs ): Add missing implementation bits
08:37:35 bbgeos: build #56 of telascience-quick started including [2385]
08:42:27 CIA-34: strk * r2386 /trunk/source/headers/geos/algorithm/distance/PointPairDistance.h: Fix typo
08:43:21 CIA-34: strk * r2387 /trunk/source/ ( 2 files in 2 dirs ): Offline some more
08:45:40 CIA-34: strk * r2388 /trunk/tests/unit/ ( 3 files in 2 dirs ): Port DiscreteHausdorffDistance unit test
08:46:31 CIA-34: strk * r2389 /trunk/tests/unit/algorithm/distance/DiscreteHausdorffDistanceTest.cpp: comment out debugging lines
08:58:53 strk: phwee, failures get back to 2
08:59:29 strk: hole_from_shell and hole_red
09:01:35 bbgeos: build #57 of telascience-quick started including [2386, 2387, 2388, 2389]
09:06:35 strk: uhm
09:06:41 strk: isn't equals( ) topological equality ?
09:15:16 strk: yes
12:46:34 bbgeos: build #34 of telascience-full started including [2381, 2382, 2383, 2384, 2385, 2386, 2387, 2388, 2389]
13:09:15 jgeerdes: Good afternoon; can someone tell me how to convert a hexadecimal string to a PG geometry value?
13:09:31 jgeerdes: The hex was exported from PG into a csv, and now I need to put it back.
13:14:26 jgeerdes: Anybody?
14:13:12 jgeerdes: Hello?
14:18:36 mloskot: bbgeos: status
14:18:36 bbgeos: telascience-full: idle, last build 1h11m35s ago: failed test-xml
14:18:36 bbgeos: telascience-quick: idle, last build 5h13m28s ago: failed test-xml
14:18:36 bbgeos: telascience-stable: idle
14:18:48 mloskot: bbgeos: force build telascience-stable
14:18:48 bbgeos: build #0 of telascience-stable started including []
14:18:48 bbgeos: build #0 forced
14:18:48 bbgeos: I'll give a shout when the build finishes
14:21:20 jgeerdes: Can someone tell me how to convert a hexadecimal string to a PG geometry value?
14:27:06 mloskot: use postgresql functions to convert hex string to raw binary
14:27:40 jgeerdes: Thank you!
14:28:14 mloskot: check postgresql manual for encode/decode
14:28:32 mloskot: here you have sample
14:28:33 mloskot: http://mateusz.loskot.net/2008/08/21/wkb-hex-decoder-in-c/
14:28:34 sigq: Title: WKB hex decoder in C++ « Mateusz Loskot ( at mateusz.loskot.net )
14:28:56 jgeerdes: Thank you immensely!!!
14:32:24 bbgeos: Hey! build telascience-stable #0 is complete: Failure [failed test-xml]
14:32:24 bbgeos: Build details are at http://buildbot.osgeo.org:8506/builders/telascience-stable/builds/0
14:32:26 sigq: Title: Buildbot: telascience-stable Build #0 ( at buildbot.osgeo.org:8506 )
15:05:09 mloskot: jgeerdes: in case you need to remember :- )
15:05:10 mloskot: http://trac.osgeo.org/postgis/wiki/FAQ
15:05:11 sigq: Title: FAQ - PostGIS - Trac ( at trac.osgeo.org )
15:13:43 jgeerdes: Thank you ^ 3
15:14:46 jgeerdes: I'm getting an error that ST_GeomFromWKB does not exist
15:14:54 jgeerdes: Any thoughts on that?
15:20:43 mloskot: jgeerdes: I've missed some lessons at my last wizard course probably, sorry, no way to guess what version of postgis you are using
15:21:09 jgeerdes: Um, I want to say 8.3.4. Checking...
15:21:13 mloskot: I use SVN trunk
15:21:29 mloskot: I use PostGIS form SVN trunk
15:21:56 jgeerdes: Oh, geesh. I read PostGIS as Posgres.
15:22:10 jgeerdes: I just dl'ed PostGIS today. 1.3.5
15:23:07 mloskot: but you have installed it, haven't you?
15:23:13 mloskot: you know, downloading is not enough
15:23:18 jgeerdes: Yeah.
15:23:21 jgeerdes: I installed : )
15:23:38 mloskot: so, you are using PostGIS 1.3.5
15:23:52 jgeerdes: Yeah; on PostgreSQL8.3.4
15:23:55 mloskot: Cool
15:24:10 jgeerdes: Hmm. Let me check something.
15:24:19 mloskot: Then check PostGIS manual for your version and find geometry constructor that accepts WKB
15:25:45 jgeerdes: I've only been tinkering with PG, etc., for a couple of months in my spare time.
15:25:54 mloskot: Oh man!
15:25:57 jgeerdes: Generally ( I know this is a bad word in here ) a MySQL fan.
15:26:04 mloskot: Where did you found that PostgreSQL == PostGIS
15:26:16 jgeerdes: Oh, I know that it doesn't.
15:26:17 mloskot: You have downloaded postgis so you've been on its website
15:26:24 jgeerdes: Yes.
15:26:29 mloskot: re fan, I don't care
15:26:34 mloskot: we are talking about postgis
15:26:38 jgeerdes: Yeah.
15:26:45 mloskot: I'm trying to help you to help yourself :- )
15:26:47 jgeerdes: Too much on my brain.
15:27:12 mloskot: Shortly, for geometry functions -> PostGIS manual for X.Y.Z where XYZ is your version of PostGIS
15:27:13 jgeerdes: This morning, I was doing math that had anywhere from one to five instances of sin, tan, cos, etc.
15:27:17 jgeerdes: It fried me : )
15:27:26 mloskot: for all other things, like encode, decode, and other SQL functions -> PostgreSQL manual
15:27:33 jgeerdes: Right.
15:27:34 mloskot: :- )
15:27:46 mloskot: http://postgis.refractions.net/documentation/manual-1.3/ch06.html#id2747038
15:27:47 sigq: Title: Chapter 6. PostGIS Reference ( at postgis.refractions.net )
23:42:18 CIA-34: strk * r2390 /trunk/tests/xmltester/BufferResultMatcher.cpp: Don't densify coordinates if buffer distance is 0
23:45:53 CIA-34: strk * r2391 /trunk/tests/xmltester/tests/TestBigNastyBuffer.xml: Imported cleaned-up version of fme.xml -- this one fails ( the one we have is actually expecting a wrong result )
23:51:17 CIA-34: strk * r2392 /trunk/tests/xmltester/XMLTester.cpp: Use BufferResultMatcher for buffer operations. No failures.