#mapnik log: Monday 23, February 2009

2009 | 02

previous | next
00:05:48 *** ninja_ (n=pankur@nat/yahoo/x-cd18ef5dae174ddf) has joined #mapnik
00:42:18 *** ninja_ has quit (Read error: 104 (Connection reset by peer))
00:45:01 *** ninja_ (n=pankur@nat/yahoo/x-6c5fc6050d118d80) has joined #mapnik
00:46:53 *** ninja_ has quit (Read error: 104 (Connection reset by peer))
00:47:01 <nikq> Mapnik Trac: Changeset [956]: scons: add option called 'LIB_DIR_NAME' to control the relative directory  ... | http://trac.mapnik.org/changeset/956
00:47:32 *** ninja_ (n=pankur@nat/yahoo/x-239790a277237788) has joined #mapnik
01:11:30 *** springmeyer has quit (Read error: 104 (Connection reset by peer))
01:12:00 <CIA-23> mapnik: dane * r956 /trunk/ (10 files in 10 dirs): scons: add option called 'LIB_DIR_NAME' to control the relative directory where input plugins are installed and where python looks to register them
01:13:57 *** springmeyer (n=dane@c-24-19-50-92.hsd1.wa.comcast.net) has joined #mapnik
03:54:00 *** cremaster (i=blotanic@vpn-wireless-client-pool-245-101.cpmc.columbia.edu) has joined #mapnik
04:02:13 *** VidTheKid (n=dexter@adsl-69-223-184-203.dsl.wotnoh.sbcglobal.net) has joined #mapnik
04:05:02 *** rcoup has quit ()
04:39:15 *** cremaster has quit (Read error: 104 (Connection reset by peer))
04:42:08 *** cremaster (i=blotanic@vpn-wireless-client-pool-245-32.cpmc.columbia.edu) has joined #mapnik
04:46:29 *** cremaster has quit (Read error: 104 (Connection reset by peer))
04:46:37 *** cremaster (i=blotanic@vpn-wireless-client-pool-245-6.cpmc.columbia.edu) has joined #mapnik
05:40:53 <dukeku> ugh. changed something in my python and now one of my layers isn't drawing and i have no idea why
05:43:36 <dukeku> http://dpaste.com/123845/ does anyone see anything wrong with that xml? i've removed everything unrelated and i still can't figure it out
05:47:41 <dukeku> heh, permissions error in postgis. didn't have permission to select from that particular table
06:08:43 <dukeku> what units does min_distance in textsymbolizer use?
06:59:47 *** VidTheKid has quit ("Going back to Dexter's laboratory...")
07:05:18 *** springmeyer has quit (brown.freenode.net irc.freenode.net)
07:05:18 *** jbglaw has quit (brown.freenode.net irc.freenode.net)
07:05:18 *** ser has quit (brown.freenode.net irc.freenode.net)
07:06:14 *** springmeyer (n=dane@c-24-19-50-92.hsd1.wa.comcast.net) has joined #mapnik
07:06:14 *** ser (n=ser@sergiusz.pawlowicz.name) has joined #mapnik
07:06:14 *** jbglaw (n=jbglaw@lug-owl.de) has joined #mapnik
07:12:17 *** weizhuo has quit ("ChatZilla 0.9.84 [Firefox 2.0.0.20/2008121709]")
07:48:19 *** rcoup (n=rcoup@ip-118-90-67-87.xdsl.xnet.co.nz) has joined #mapnik
07:56:13 *** cremaster has quit (Read error: 104 (Connection reset by peer))
07:56:25 *** xcacou (n=aga@AToulouse-157-1-94-155.w86-213.abo.wanadoo.fr) has joined #mapnik
09:23:02 *** rcoup has quit ()
11:19:55 *** kunitoki (n=kraken@host217-241-dynamic.44-79-r.retail.telecomitalia.it) has joined #mapnik
11:25:58 <kunitoki> springmeyer: i'm having problems with sqlite, i don't know why but as soon as i try to visualize data from spatial sqlite database... it say "no such module: rtree"
11:26:38 <kunitoki> have you got an idea ?
11:26:58 <kunitoki> i was able to make it work last week... and i changed nothing !
12:15:38 *** ninja__ (n=pankur@cm122.psi133.maxonline.com.sg) has joined #mapnik
12:32:02 *** ninja_ has quit (Read error: 110 (Connection timed out))
12:59:02 *** ninja__ has quit ()
13:15:13 *** sanjiv (n=chatzill@59.180.186.57) has joined #mapnik
13:31:57 <nikq> Mapnik Trac: Changeset [957]: + use "the_geom" as a default geometry field | http://trac.mapnik.org/changeset/957
13:58:10 *** artem_ (n=artem@i-83-67-142-225.freedom2surf.net) has joined #mapnik
14:05:37 <CIA-23> mapnik: artem * r957 /trunk/plugins/input/sqlite/sqlite_datasource.cpp: + use "the_geom" as a default geometry field
14:07:27 <artem_> kunitoki: hey, I changed sqlite.input to use standard wkb by default, this can be the reason.
14:09:20 <kunitoki> artem_: standard wkb by default ?
14:09:42 <kunitoki> ah ok the wkb conversion
14:09:51 <artem_> yep, it should be optional I guess
14:10:03 <kunitoki> it should be selectable by a parameter
14:10:11 <artem_> yep
14:11:17 <artem_> kunitoki : i written pgsql2sqlite utility which creates sqlite file from any pgsql table (query)
14:12:04 <kunitoki> yes i saw in the last repo change
14:12:07 <kunitoki> +s
14:12:46 <kunitoki> so practically my sqlite database are full of crap now ;)
14:12:58 <artem_> it doesn't seem to make much sense to have an extra four double per geometry just to be compliant with spatialite which uses it's own non standard wkb
14:13:10 <artem_> kunitoki: not quite :)
14:13:13 <kunitoki> yes i think that is okay
14:13:29 <kunitoki> anyway would be cool if we can somewhat guess the blob
14:13:35 <kunitoki> automagically
14:13:56 <artem_> kunitoki:  i don't think it's possible
14:14:09 <kunitoki> pqsql2sqlite is cool, but i need oracle2sqlite or better shape2sqlite
14:14:30 <kunitoki> artem_: yeah probably a perfect guess is impossible
14:15:03 <kunitoki> at least i can guess if a blob is spatialite one or plain wkb
14:15:11 <kunitoki> anyway doesn't matter
14:15:12 <artem_> kunitoki: yes, I think it is OK to be able to specify blob format
14:15:56 <artem_> this way everything will work , including your database.
14:16:05 <kunitoki> ok
14:16:41 <artem_> I have some ideas to extent sqlite.input to support multiple files
14:17:05 <artem_> I have europe.sqlite which is 3.6 GB - quite big
14:17:43 <artem_> It'll be cool to be able to handle multiple smaller files
14:18:27 <artem_> through 'secondary' index - this way it'll be possible to render any size dataset
14:20:34 *** ninja (n=pankur@cm122.psi133.maxonline.com.sg) has joined #mapnik
14:22:12 <artem_> kunitoki:  re: anything2sqlite  - it would be fairly easy to write ogr2sqlite
14:22:40 <crschmidt> er
14:22:45 <crschmidt> ogr has sqlite support?
14:23:00 <crschmidt> directly
14:23:05 *** ninja_ (n=pankur@nat/yahoo/x-cb8f658e75179d47) has joined #mapnik
14:23:13 <artem_> I don't think it can do spatial index+blob but :)
14:23:21 <crschmidt> Define 'blob'?
14:23:35 <crschmidt> It stores geometry as wkt or wkb
14:23:52 <artem_> ok
14:24:03 * artem_ checking ..
14:24:09 <crschmidt> If Mapnik needs its own proprietary geometry format in the database... welll, I'd recommend against that :)
14:24:22 <crschmidt> ANd I think that creating a spatial index afterwards should just be a single sql comand, right?
14:25:00 <artem_> nope unless you use spatialite
14:25:13 <crschmidt> Hm
14:25:18 <crschmidt> I guess I don't understand what you mean?
14:25:29 <artem_> which is return uses slightly diff format for wkb
14:25:36 <artem_> is in return
14:25:52 <artem_> ok, to create spatial index :
14:26:12 <crschmidt> Anyway, I'd recommend that if there is a reason that you can't use the OGR support ifor sqlite tables, it would be worth mentioning as a bug to the gdal trac
14:26:38 <crschmidt> since having yet another sqlite format that ogr can't read would kinda suck
14:27:40 <artem_> let me check what ogr can and can't read from sqlite ..
14:33:44 <kunitoki> well it actually support plain wkb/wkt in the geometry column
14:34:25 <kunitoki> and tables metadata... nothing more than that
14:35:00 * artem_ failing to force ogr2ogr to use WKB instead of WKT
14:39:37 <artem_> doh , maybe I need to upgrade gdal stuff ..
14:40:03 *** ninja has quit (Read error: 110 (Connection timed out))
14:44:10 <artem_> crschmidt : I must be missing something but I can't find a link to download GDAL source on :http://trac.osgeo.org/gdal/
14:44:34 <artem_> got it
14:44:39 <artem_> it was hidden :)
14:45:14 <crschmidt> artem_: generally speaking, http://gdal.org/ is where you would go for source
14:45:30 <artem_> opps, thanks
14:45:34 <crschmidt> (in the same way I can't find a link for source from http://trac.mapnik.org/ )
14:45:53 <artem_> fair point
14:46:22 <artem_> must be to do with trac.*.org URLs
14:46:58 <crschmidt> well, http://trac.openlayers.org/ has 'how to get the code' and 'how to file a bug' as the first and second bulleted links
14:47:21 <artem_> ok, it looks like wkb was added in 1.6.0
14:47:40 <crschmidt> Yep
14:48:05 <crschmidt> http://trac.osgeo.org/gdal/changeset/14056
14:48:06 <artem_> yep. I like 'get the code'
14:48:18 * artem_ compiling gdal
14:48:55 * crschmidt doesn't ever compile gdal
14:48:59 <crschmidt> that's what KyngChaos is for
14:49:40 *** aub (n=aubrey@70.107.236.83) has joined #mapnik
14:49:49 <artem_> can KyngChaos compile on mac ?
14:51:23 <crschmidt> artem_: KyngChaos is the set of precompiled frameworks for mac
14:51:28 <crschmidt> So... "no"?
14:52:29 <crschmidt> Because it's already compiled :)
14:54:27 <kunitoki> artem_: regarding the "no module found: rtree" have you got some ideas ?
14:54:46 <kunitoki> if i open my sqlite database (created with spatialite)
14:55:02 <kunitoki> with sqlite3 and query the spatial index i get that message
14:55:02 * artem_ back
14:55:21 <artem_> kunitoki; yes
14:55:25 <kunitoki> shouldn't the rtree module be already built in ?
14:55:31 <artem_> nope
14:55:45 <artem_> are you on linux ?
14:55:48 <kunitoki> yeah
14:56:08 <artem_> r*tree needs enabling at compilation time
14:56:09 <kunitoki> last week i was able to use the spatial index with the sqlite.input
14:56:37 <artem_> you probably have multiple sqlite3 around I had similar probs
14:56:40 <kunitoki> mmmmh probably my version of sqlite has that not enabled ?
14:56:47 <kunitoki> yeah could be
14:57:04 <kunitoki> i'll recompile a new version
14:57:35 <artem_> you need : -DSQLITE_ENABLE_RTREE=1
14:58:20 <artem_> spatialite uses its own copy of quite outdated sqlite 3.6.6
15:00:30 <nikq> Mapnik Trac: Changeset [958]: - sqlite.input: added row_offset and row_limit to parameters -  ... | http://trac.mapnik.org/changeset/958
15:07:53 <artem_> ok, gdal 1.6.0 supports WKB in SQLiite but there's no easy way to create a spatial index
15:08:50 <CIA-23> mapnik: kunitoki * r958 /trunk/ (7 files in 3 dirs):
15:08:50 <CIA-23> mapnik: - sqlite.input: added row_offset and row_limit to parameters
15:08:50 <CIA-23> mapnik: - sqlite.input: added wkb_format parameter for selecting WKB format (generic/spatialite)
15:08:50 <CIA-23> mapnik: - sqlite.input: commented check of the spatial index
15:08:50 <CIA-23> mapnik: - wkb.hpp: removed wkqSQLite in favour of wkbSpatiaLite
15:11:24 <artem_> kunitoki: ogr2ogr should do the trick for shp2sqlite but I'm not sure about oci2sqlite
15:12:22 <kunitoki> well, it shouldn't be a problem as we have the occi.input plugin now
15:13:05 <kunitoki> things that mapnik now can load and browse, it can also output
15:13:17 <kunitoki> what about output plugins in mapnik ?
15:14:06 <crschmidt> artem_: how do you build an index if you've got the data in a database? I assumed there was a SQL 'create index' command
15:14:25 <kunitoki> in sqlite is create virtual X using rtree
15:14:45 <artem_> create virtial table (id, minx.maxx.miny.maxy);
15:14:50 <crschmidt> the OGR command line utils (ogrinfo) can execute raw sqlite sql
15:15:11 <crschmidt> so it's a two liner: ogr2ogr , then ogrinfo -sql "CREATE VIRTUAL foo (id, blah)" filename.sqlite
15:15:26 <artem_> you need to be able to extract envelope on geometry column
15:15:32 <artem_> from
15:15:45 <crschmidt> And you can't do that in SQL?
15:15:52 <artem_> nope
15:16:17 <artem_> it is a just blob from SQL point of view
15:16:27 <crschmidt> artem_: ah, but ogrinfo can also give you that information
15:16:39 <crschmidt> ogrinfo -so -al filename.sqlite
15:16:45 <crschmidt> Will give you an "Extent" line
15:17:00 <artem_> for each geometry ?
15:17:07 <crschmidt> No
15:17:10 <crschmidt> -so is 'summary'
15:17:15 <crschmidt> ('summary only')
15:17:19 <crschmidt> and it gives you an extent for the layer
15:17:44 <crschmidt> on my projected dataset, it looks like:
15:17:44 <crschmidt> Extent: (232092.630000, 829383.067800) - (248819.768300, 840903.220900)
15:17:53 <artem_> extent for the layer is good but not good enough to index individual geometries
15:17:59 <crschmidt> er
15:18:02 <crschmidt> I'm confused, I guess
15:18:20 <crschmidt> So when you create the index, you create it on each individual geometry?
15:18:53 <artem_> of course :  insert into spatial_index(id, lox,hix,loy,hiy);
15:19:01 <artem_> for each geometry
15:19:12 <kunitoki> if we have a global spatial index for all geometries... is of no use
15:19:16 <crschmidt> 'of course', like this actually makes sense in the ralm of anything :p
15:19:57 <kunitoki> we should have each geometry indexed. as you index an integer field in a database, you index each row in the table not the whole table
15:20:15 <crschmidt> No, I tell the database 'CREATE INDEX on COLUMN'
15:20:23 <artem_> :D. well you confused me , too. You create a spatial index on a table(layer) but you need to know extents of individual geoms
15:20:31 <crschmidt> And the database is smart enough to know what that means
15:21:01 <artem_> ha sqlite is not that smart
15:21:15 <artem_> and rightly so - it is light:)
15:21:18 <crschmidt> artem_: SQLite is that smart, for indexes on columns it understands.
15:22:00 <artem_> exactly, so spatialite has got custom functions and data types etc etc
15:22:11 <crschmidt> Okay.
15:22:52 <artem_> but from consumer point (mapnik just consumes data) all this OGC crap is not needed
15:22:53 <crschmidt> So yes, if you're not able to use spatiallite, You'll have to write a custom script
15:23:18 <crschmidt> That said, if you have ogr2ogr handy, you can get a non-geometry indexed sqlite file out easily
15:23:29 <crschmidt> So I guess the question that I have is whether that format is readable by mapnik
15:24:00 <artem_> just a min and I'll give the answer ..
15:27:22 * crschmidt doesn't think you *need* indexes for everything. Sequential scans of not-too-many rows aren't a problem
15:28:28 <artem_> sure
15:29:21 <artem_> kunitoki: ok. sqlite.input wants PK_UID while ogr produced something else  ..
15:29:37 <crschmidt> OGC_FID?
15:29:55 <artem_> OGC_FID indeed
15:30:01 <crschmidt> OGC_FID will be a unique integer, probably starting at 0 for most datasets
15:30:23 <crschmidt> Don't know if that maps to what PK_UID is expected to be
15:30:49 <artem_> kunitoki: I think we shouldn't be too restrict about unique key name
15:30:59 <artem_> PK_UID is spatialite art
15:31:51 <artem_> kunitoki : how about configurable key name ? default to OGC_FID ?
15:33:19 <kunitoki> artem_: yes key_name is already configurable
15:33:28 <kunitoki> defaulting to what you like best
15:33:39 <artem_> sounds good!
15:33:43 <kunitoki> for me is ok everything
15:33:48 <kunitoki> ip refer a plain ID
15:33:50 <kunitoki> :)
15:34:21 <kunitoki> i commented out the check for spatial index existence
15:34:33 <artem_> why ?
15:34:44 <kunitoki> cause i think it could slow little tables
15:34:51 <kunitoki> i do a double check
15:35:04 <kunitoki> check for the exietence of 'idx_table_geonmetry' in sqlite_master
15:35:25 <artem_> ok
15:35:40 <kunitoki> and check if spatial_index_enabled, but that is spatialite voodoo
15:36:16 <artem_> cool
15:36:29 <kunitoki> i can make check only if the idx_table_geonmetry exists, but it's one more query to do every layer
15:37:03 <kunitoki> anyway that will disable the spatial index. if you don't have it should exit with an error not bypass it
15:37:26 <artem_> I guess we need to strike a good balance but it's fine by me
15:37:55 <artem_> yep, great.
15:38:13 *** matt_c (n=mcroydon@gateway.sunflowerbroadband.com) has joined #mapnik
15:44:21 <artem_> Okey-dokey, mapnik can read OGR-made *.sqlite with following gotchas :
15:44:27 <artem_> 1. no spatial index support
15:44:39 <artem_> 2. <Parameter name="key_field">OGC_FID</Parameter>
15:45:17 <crschmidt> artem_: Cool. And can ogr read the sqlite files made by mapnik utils? (there's a csv2sqlite, right?)
15:45:43 *** shantz (n=jak@umma193.dn68.umontana.edu) has joined #mapnik
15:45:44 <artem_> you mean pgsql2sqlite ?
15:46:06 <artem_> let me try
15:46:47 <artem_> ok, nope because ogr expects geometry_column table. I guess I can add this at some point
15:47:34 <artem_> but binaries (WKBs) are the same, which is good
15:48:46 <artem_> pgsql2slite can accept _any_ SQL statement to select data from pgsql - ogr2ogr only where clause
15:50:42 <artem_> what I'm really after is to be able to 'cooke cut' any polygon from postgresql - ogr2pgr can't do this.
15:51:05 <crschmidt> artem_: if binaries are the same, all else is a wash
15:51:17 <crschmidt> artem_: mostly, I'm trying to fight against the pgsql EWKB mess all over again :)
15:51:45 <artem_> good point
15:52:13 <artem_> it might well be that EWKB == "spatialite WKB"
15:52:17 <artem_> ??
15:52:36 <crschmidt> Dunno
15:52:39 <artem_> EWKB = "first byte" + Envelope + "the rest"
15:52:44 <kunitoki> similar
15:53:11 <artem_> s/EWKB/SpatialiteWKB/ sorry
16:02:18 <kunitoki> at some point we could make a small tool, based on mapnik which reads all the spatial tables inside a sqlite database and builds the indexes
16:02:56 <kunitoki> or keep the indexes up to date if already present
16:03:54 <artem_> yep, good idea
16:05:03 <nikq> Mapnik Trac: Changeset [959]: - sqlite.input: revert back the check for spatial index virtual table in  ... | http://trac.mapnik.org/changeset/959
16:06:08 *** ninja_ has quit ()
16:15:07 <CIA-23> mapnik: kunitoki * r959 /trunk/plugins/input/sqlite/sqlite_datasource.cpp:
16:15:07 <CIA-23> mapnik: - sqlite.input: revert back the check for spatial index virtual table in sqlite_master
16:15:07 <CIA-23> mapnik: - sqlite.input: default of key_field is now OGC_FID (artem)
16:22:12 *** cremaster (i=blotanic@vpn-wireless-client-pool-245-56.cpmc.columbia.edu) has joined #mapnik
16:38:02 *** sanjiv has quit (Remote closed the connection)
16:39:08 *** sanjiv (n=chatzill@59.180.186.57) has joined #mapnik
16:42:31 *** xcacou has quit (Remote closed the connection)
16:52:48 <cremaster> can someone point me to an example or just the syntax of using a gdal datasource
16:55:01 <cremaster> layer.datasource = gdal(file='path') ?
16:55:31 <artem_> cremaster : I don't remember all details but looks about right
16:58:26 <springmeyer> cremaster: yes. but 'Gdal(...' not 'gdal(..'
16:58:35 <springmeyer> tiny sample you can try to run here: http://mapnik-utils.googlecode.com/svn/example_code/gdal/
17:11:14 *** artem_ has quit ()
17:11:39 *** sanjiv has quit (Read error: 110 (Connection timed out))
17:14:28 *** sanjiv (n=chatzill@59.180.157.206) has joined #mapnik
17:14:48 *** artem_ (n=artem@i-83-67-142-225.freedom2surf.net) has joined #mapnik
17:16:17 <nikq> Mapnik Trac: Changeset [960]: - occi.input: improved determination of layer envelope by querying the  ... | http://trac.mapnik.org/changeset/960
17:16:40 *** cremaster has quit (Read error: 110 (Connection timed out))
17:18:50 <nikq> Mapnik Trac: Ticket #226 (OGCServer Style and Layer ordering with load_map() support) closed | http://trac.mapnik.org/ticket/226#comment:2
17:21:58 <kunitoki> bye bye people
17:22:02 <kunitoki> laterz
17:22:06 *** kunitoki has quit ("Lost terminal")
17:22:18 <CIA-23> mapnik: kunitoki * r960 /trunk/plugins/input/occi/ (occi_featureset.cpp occi_types.hpp occi_datasource.cpp):
17:22:18 <CIA-23> mapnik: - occi.input: improved determination of layer envelope by querying the spatial index metadata
17:22:18 <CIA-23> mapnik: - occi.input: added optional estimate_extent parameter (much like pgsql.input)
17:22:18 <CIA-23> mapnik: - occi.input: added prefetch of 100 rows in spatial queries (must be tunable)
17:22:18 <CIA-23> mapnik: - occi.input: commented a console debug clog output that was flooding my bash
17:53:10 *** artem_ has quit ()
18:22:38 *** artem_ (n=artem@i-83-67-142-225.freedom2surf.net) has joined #mapnik
18:23:50 *** cremaster (i=blotanic@vpn-wireless-client-pool-245-58.cpmc.columbia.edu) has joined #mapnik
18:25:20 *** artem_ has quit (Client Quit)
18:35:28 *** kunitoki (n=kraken@host217-241-dynamic.44-79-r.retail.telecomitalia.it) has joined #mapnik
18:50:36 *** sanjiv has quit ("ChatZilla 0.9.84 [Firefox 3.0.6/2009020911]")
19:07:17 *** artem_ (n=artem@i-83-67-142-225.freedom2surf.net) has joined #mapnik
20:02:01 *** matt_c_ (n=mcroydon@gateway.sunflowerbroadband.com) has joined #mapnik
20:03:17 *** artem_ has quit ()
20:06:26 *** matt_c_ has quit ()
20:16:42 *** matt_c has quit (Connection timed out)
20:22:38 *** dsfds (i=blotanic@vpn-wireless-client-pool-245-101.cpmc.columbia.edu) has joined #mapnik
20:22:54 *** cremaster has quit (Nick collision from services.)
20:22:58 *** dsfds is now known as cremaster
20:26:29 *** matt_c (n=mcroydon@gateway.sunflowerbroadband.com) has joined #mapnik
20:27:51 *** matt_c has quit (Client Quit)
20:57:50 <nikq> Mapnik Trac: Ticket #244 (Support interleaved ShieldSymbolizers on a way) closed | http://trac.mapnik.org/ticket/244#comment:1
20:59:41 *** matt_c (n=mcroydon@gateway.sunflowerbroadband.com) has joined #mapnik
21:02:16 <nikq> Mapnik Trac: ShieldSymbolizer edited | http://trac.mapnik.org/wiki/ShieldSymbolizer?version=12
21:19:28 *** cremaster has quit (Read error: 104 (Connection reset by peer))
21:19:37 *** cremaster (i=blotanic@vpn-wireless-client-pool-245-3.cpmc.columbia.edu) has joined #mapnik
21:36:57 *** matt_c has quit (Remote closed the connection)
21:38:24 *** matt_c (n=mcroydon@gateway.sunflowerbroadband.com) has joined #mapnik
21:52:55 *** ChanServ has quit (brown.freenode.net irc.freenode.net)
22:04:09 *** Ruffiano (n=ruffiano@static-70-21-119-183.res.east.verizon.net) has joined #mapnik
22:04:44 *** ChanServ (ChanServ@services.) has joined #mapnik
22:05:05 <Ruffiano> getting a scons build error on the trunk build:
22:05:12 <Ruffiano> /usr/include/boost/python/converter/pyobject_type.hpp:21: warning: dereferencing type-punned pointer will break strict-aliasing rules
22:05:12 <Ruffiano> scons: *** [bindings/python/mapnik_rule.os] Error 1
22:05:12 <Ruffiano> scons: building terminated because of errors.
22:05:52 <springmeyer> Ruffiano: what version of boost does Scons print out that it finds?
22:06:02 <Ruffiano> all the dependencies are found, only optionals that show up missing is  boost_system-mt and cario,,-1.0
22:06:05 <Ruffiano> 1_33!
22:06:09 <Ruffiano> 1_33_1 ***
22:06:12 <springmeyer> okay
22:06:38 <springmeyer> what os? linux version...?
22:06:43 <Ruffiano> should I try 1.38 ?
22:06:47 <Ruffiano> CentOS
22:07:02 <springmeyer> okay
22:07:18 <springmeyer> well, it sure would be nice to know what's wrong before you upgrade :)
22:07:54 <Ruffiano> =)
22:08:08 <Ruffiano> CentOS 5 x86 64bit
22:08:15 <Ruffiano> to be a bit more precise
22:08:23 <springmeyer> so 1_33_1 from emerge or source?
22:08:40 <springmeyer> wait or whatever packing centos uses...
22:09:18 <Ruffiano> right, not sure I am installing it elsewhere, I'll check
22:10:13 <Ruffiano> yeah yum is giving me 1.33 from the boost_devel package
22:10:29 <Ruffiano> and boost
22:13:29 <nikq> Mapnik Trac: Changeset [961]: + initial support for 'SELECT' statements in  param. + use limit 0  ... | http://trac.mapnik.org/changeset/961
22:17:28 <kunitoki> springmeyer: how to use nikq to let it spit out comments for when people will come online ?
22:17:57 <kunitoki> springmeyer: i need to make a comment for when artem is going online
22:19:00 * springmeyer on sec...
22:19:11 <CIA-23> mapnik: artem * r961 /trunk/plugins/input/sqlite/sqlite_datasource.cpp:
22:19:11 <CIA-23> mapnik: + initial support for 'SELECT' statements in <table> param.
22:19:11 <CIA-23> mapnik: + use limit 0 to avoid lengthy queries (does it work ?)
22:19:41 * springmeyer back
22:19:48 <kunitoki> his "use limit 0" does not work :) and i put already a message in the code
22:19:58 <kunitoki> but he clearly didn't read it
22:20:01 <springmeyer> kunitoki: you can say...
22:20:12 <springmeyer> nikq: tell artem 'this is a test'
22:20:12 <nikq> springmeyer: I'll pass that on when artem is around.
22:20:37 <kunitoki> springmeyer: wicked !
22:21:23 <springmeyer> ya :)
22:21:30 <springmeyer> other basic commands are here:
22:21:33 <springmeyer> .g phenny
22:21:34 <nikq> springmeyer: http://inamidst.com/phenny/
22:22:17 <nikq> Mapnik Trac: Changeset [962]: + small fix | http://trac.mapnik.org/changeset/962
22:24:01 <kunitoki> springmeyer: thanx :)
22:24:49 <kunitoki> nikq: .seen artem
22:24:57 <kunitoki> mmmmh
22:25:01 <kunitoki> :D
22:25:09 <springmeyer> .seen artem
22:25:09 <nikq> springmeyer: I last saw artem at 2009-02-22 23:47:15 UTC on #mapnik
22:25:14 <kunitoki> ah ok
22:25:18 <kunitoki> .seen artem
22:25:18 <nikq> kunitoki: I last saw artem at 2009-02-22 23:47:15 UTC on #mapnik
22:25:26 <kunitoki> ok now i got it
22:25:34 <springmeyer> ya I think anyting prefixed by a dot works that way...
22:25:40 <kunitoki> and tell ?
22:25:51 <kunitoki> i need to put nikq: before ?
22:26:01 <springmeyer> .tell kunitoki 'testing without addressing nikq'
22:26:12 <springmeyer> ya, needs nikq: i guess
22:26:21 *** cremaster has quit (Read error: 60 (Operation timed out))
22:26:26 <springmeyer> this bot is all new to me too :)
22:28:03 *** weizhuo (n=chatzill@nat/yahoo/x-5d3a0c43c086cca5) has joined #mapnik
22:28:29 <kunitoki> nikq: tell artem 'Ragarding your limit 0 fix in r961 you should have read the XXX comment before doing that: without a row, we cannot iterate over the fields'
22:28:29 <nikq> kunitoki: I'll pass that on when artem is around.
22:28:32 <nikq> http://trac.mapnik.org/changeset/961, at , by artem: + initial support for 'SELECT' statements in &lt;table&gt; param.+ use limit 0 to avoid lengthy queries (does it work ?)
22:29:17 <springmeyer> Ruffiano: all google gives me is: http://lists.gforge.inria.fr/pipermail/cgal-python-users/2007-November/000043.html
22:29:39 <springmeyer> seems it may be a boost python bug (note the error is coming from boost python src not mapniks)
22:29:46 <Ruffiano> aye
22:30:13 <Ruffiano> ok, I'll give updating it a go tomorrow, getting sleepy ;-)
22:30:21 <Ruffiano> as always, many thanks
22:30:27 *** matt_c has quit ()
22:31:34 <cmarqu> Behold the power of the angry FruitShieldSymbolizer: http://oerks.net/~colin/hike_n_bike_symbols_2009-02-23.png
22:31:43 *** ChanServ has quit (brown.freenode.net irc.freenode.net)
22:32:51 *** ChanServ (ChanServ@services.) has joined #mapnik
22:32:51 <springmeyer> Ruffiano: okay. cool
22:34:01 <springmeyer> Ruffiano: a quick ticket loggin the error would be nice if you are not too sleepy :)
22:35:02 <springmeyer> cmarqu: rockin' :)
22:35:05 <Ruffiano> lol I think I can manage tgat
22:38:20 <springmeyer> Ruffiano: just guessing this might be the change that fixes that error: https://svn.boost.org/trac/boost/changeset/39191/trunk/boost/python/converter/pyobject_type.hpp
22:39:21 <Ruffiano> I'll include that in the ticket too
22:40:04 <nikq> Mapnik Trac: Ticket #247 (CentOS 5 x86 64bit Python/Scons Error1 on build from boost) created | http://trac.mapnik.org/ticket/247
22:41:46 <springmeyer> thx
22:43:55 <Ruffiano> thanks again, and have a good evening
22:45:59 *** Ruffiano has parted #mapnik ()
22:49:39 <CIA-23> mapnik: artem * r962 /trunk/plugins/input/sqlite/sqlite_datasource.cpp: + small fix
22:52:39 *** racicot has quit (Connection reset by peer)
22:53:37 *** racicot (n=chatzill@dsl-209-166-85-189.whidbey.net) has joined #mapnik
22:56:43 *** kunitoki has quit ("Lost terminal")
22:56:53 *** artem_ (n=artem@i-83-67-142-225.freedom2surf.net) has joined #mapnik
22:58:41 <artem_> kunitoki: ping
22:59:32 <artem_> "limit 1 " is causing problem with table="select * from gig_table order by z;"
22:59:42 <artem_> gig->big
23:00:20 <artem_> there must be a better way ..
23:00:38 *** racicot has quit (Read error: 104 (Connection reset by peer))
23:01:37 *** racicot (n=chatzill@dsl-209-166-85-189.whidbey.net) has joined #mapnik
23:05:42 <artem_> .seen kunitoki
23:05:42 <nikq> artem_: I last saw kunitoki at 2009-02-23 22:28:29 UTC on #mapnik
23:26:58 <nikq> Mapnik Trac: Changeset [963]: + revert to "limit 1" logic + discard everything after table name when  ... | http://trac.mapnik.org/changeset/963
23:31:20 <artem_> springmeyer : osm.xml should be convertible to use sqlite.input now. looks good over here
23:31:54 <artem_> springmeyer : requires slightly diff SQL syntax
23:32:26 <springmeyer> hey artem_: very very cool
23:32:42 <springmeyer> whats an example of the syntax diff?
23:33:16 <artem_> pgsql : "(select blah,blah,blah) as roads"  - >
23:33:37 <artem_> sqlite: "(select blah,blah,blah) "
23:33:59 <artem_> char_length(..) as name  -->   length(name) as length
23:34:24 <artem_> sqlite doesn't like "as <name>" syntax
23:34:29 <springmeyer> great
23:34:49 * springmeyer wonders whether that syntax works for pgsql too?
23:34:56 <springmeyer> or could...
23:35:10 <artem_> nope it doesn't i'm afraid
23:35:34 <springmeyer> okay
23:36:18 <artem_> i'm pretty sure pgsql always wants "as " in sub-selects
23:44:46 <nikq> Mapnik Trac: Changeset [964]: + create table if not exists + use OGC_FID instead of PK_UID | http://trac.mapnik.org/changeset/964
23:48:06 * springmeyer gets... 'near "limit": syntax error'
23:48:21 <crschmidt> springmeyer: that's r963, right?
23:48:22 <nikq> http://trac.mapnik.org/changeset/963, at , by artem: + revert to "limit 1" logic+ discard everything after table name when building
23:48:32 <springmeyer> yes
23:48:53 <springmeyer> whoops, forgot to change the char length...
23:49:53 <springmeyer> ah, made no difference
23:50:17 <artem_> can i see your sql ?
23:53:15 <springmeyer> ya, its just some old osm.xml from a few months back.
23:53:20 <springmeyer> one moment
23:53:53 <artem_> i have : http://dpaste.com/371/
23:54:11 <artem_> and :(select * from planet_osm_line order by z_order)
23:56:12 <artem_> this might help: http://dpaste.com/372/
23:56:30 <springmeyer> ah, looks like it messed by my sql with sloppy find and replace :)
23:56:46 <artem_> aha, cool :)