02:01:25 *** weizhuo (n=chatzill@nat/yahoo/x-9c54d31734f97d01) has joined #mapnik
03:47:37 *** stamf_ has quit (Read error: 104 (Connection reset by peer))
03:51:42 *** stamf (n=stamf@116.73.233.220.exetel.com.au) has joined #mapnik
03:59:15 *** aura is now known as w0lfie_
04:12:56 *** rcoup has quit (Read error: 60 (Operation timed out))
04:17:31 *** rcoup (n=rcoup@ip-118-90-12-207.xdsl.xnet.co.nz) has joined #mapnik
05:57:54 *** rcoup has quit ()
06:55:28 *** xcacou (n=aga@vel78-2-82-243-106-36.fbx.proxad.net) has joined #mapnik
07:23:18 *** weizhuo has quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
07:50:25 *** darragh (n=darragh@83.70.173.25) has joined #mapnik
08:15:01 *** darragh_ (n=darragh@83.70.173.25) has joined #mapnik
08:16:16 *** darragh_ has quit (Remote closed the connection)
08:24:23 *** darragh has quit (Read error: 110 (Connection timed out))
09:43:07 *** logion (n=florisvd@ahmed.ugent.be) has joined #mapnik
09:43:44 <logion> hi, i'm trying to setup mapnik as a WMS server through ogcserver but I can't get it to work
09:43:55 <logion> using mapnik 0.60
09:44:22 <logion> I keep getting "The factory module could not be loaded."
09:44:48 <logion> (been using this guide : http://code.google.com/p/mapnik-utils/wiki/WmsInstallGuide )
09:44:56 <logion> any ideas?
09:47:31 <logion> http://www.paste-it.net/public/z0baceb/
09:47:36 <logion> this is the complete traceback
09:50:59 <logion> it seems like the map_factory.py can't be found, it's next to my ogcserver.conf
09:54:02 <logion> anyone?
12:11:23 *** w0lfie_ has quit (brown.freenode.net irc.freenode.net)
12:11:49 *** w0lfie_ (n=wolf@cpe-67-49-133-78.hawaii.res.rr.com) has joined #mapnik
12:19:19 <springmeyer> logion: make sure it is on your PYTHONPATH too
12:36:29 <ser> hello, what can i get from OSM using GetFeatureInfo (via ogcserver)?
12:44:40 <logion> it's working now springmeyer, thx anyway
12:49:34 *** mib_5k27af (i=9dc14d64@gateway/web/ajax/mibbit.com/x-907835604f397bcb) has joined #mapnik
12:50:22 <mib_5k27af> floris is een mietje!
13:28:24 *** mib_5k27af has quit ("http://www.mibbit.com ajax IRC Client")
13:35:42 <springmeyer> ah good logion: what got it working?
13:35:52 <springmeyer> ser: hello :)
13:36:16 <springmeyer> GetFeatureInfo returns a list of all feature attributes for a layer
13:37:19 <springmeyer> for OSM you'll get a lot of empty columns but you'll be able to pull out street names, tags, etc
13:37:29 <springmeyer> whatever is in the postgis db
13:46:29 <springmeyer> ser: here is an example of text based output of GetFeatureInfo: http://dpaste.com/30713/
13:47:10 <logion> ah
13:47:19 <logion> i had an error in my python path indeed
13:51:59 <logion> I was wondering if there are alternatives to OGC server, because for this project there'll have to be caching to make load times acceptable
13:54:05 <springmeyer> yes, there are a variety of options for caching
13:55:01 <springmeyer> if you are planning to serve/cache tiles in the OSM/Google scheme of x/y/z.png then you should check out mod_tile
13:55:08 <springmeyer> which is what OSM uses
13:56:37 <springmeyer> if you need to support a different tiling scheme/projection or want to accept WMS bbox queries and still cache tiles then TileCache is excellent
13:57:08 <logion> I was just reading up on TileCache
13:58:46 <logion> What we'd like to do is serve our own map(in shape format) through a WMS server(which we're running)
13:59:16 <logion> But in order to limit access times there would have to be some kind of caching
14:00:11 <springmeyer> sure, and if your data is not changing frequently and you have the disk space there is no good reason not to cache
14:00:36 <springmeyer> logion: so just a single shapefile, and in what projection?
14:01:15 <logion> Horizontaal referentiesysteem:  EPSG:  31370  (Belge 1972 / Belgian Lambert 72)
14:01:15 <logion> Verticaal referentiesysteem: NGI: TAW (Tweede Algemene Waterpassing)
14:01:33 <logion> is this what u meant?(i'm pretty new in all this)
14:02:13 <springmeyer> ya, exactly
14:02:20 <springmeyer> http://spatialreference.org/ref/epsg/31370/
14:02:43 <logion> jup
14:03:08 <springmeyer> so, thats great. It will just take some work to contruct your TileCache parameters
14:03:11 <logion> allthough we are limiting our map to a specific area of belgium, city Gent
14:03:23 <springmeyer> for example, max extents, scales, etc
14:03:27 <springmeyer> okay
14:03:46 <springmeyer> you'll control that with the data_extent parameter
14:03:52 <logion> is there a way to extract a new shape file only containing the city we want to extract?
14:04:55 <logion> Because working with a smaller shape file might require less load
14:04:58 <springmeyer> ya, not with mapnik but with something like QuantumGIS you could do that
14:05:22 <springmeyer> with mapnik you can define styles and filters to only 'show' certain data
14:05:46 <springmeyer> but ya, actually reducing the size of the file would be good for speed and I'd use QGis for that
14:06:13 <springmeyer> or ogr2ogr on the command line with the --sql argument (to select out data via an attribute like city name)
14:07:45 <logion> quantumgis looks intereseting, we're actualy waiting for the shape file though so can't work on that
14:08:36 <logion> what we want to achieve on this project is making our own 'google maps', so that we can serve our own maps that we can style at wish
14:09:28 <springmeyer> sounds good
14:09:59 <springmeyer> generally that desire is what brings many to use Mapnik with OpenStreetMap data
14:10:17 <logion> well
14:10:29 <logion> i downloaded some OSM data last night for belgium
14:10:46 <logion> http://download.geofabrik.de/osm/europe/
14:11:23 <logion> and put it in a postgresql database using osm2pgsql, but I don't know what to do with it now
14:11:41 <springmeyer> ah, cool
14:11:44 <logion> I also downloaded the shape file hosted @ that website, but it doesn't look all that 'fancy'
14:12:20 <springmeyer> the shapefile is may be just coastlines?
14:12:41 <springmeyer> the stuff you imported into postgresql is all the good stuff
14:12:54 <logion> oh, well there are files in the downloaded zip like roads & natural
14:13:01 <springmeyer> okay
14:13:03 <logion> oh rly?
14:13:23 <logion> but how do i access it from my WMS server, does it also work with bbox'es etc. ?
14:13:34 <springmeyer> logion: next step is to grab a set of XML styles that OSM uses to style the data in postgres
14:13:38 <logion> cause now i just load a shape file in my python mapnik map factory
14:14:21 <springmeyer> ya, you just need to load the XML in your map factory rather than writing stuff in python
14:14:48 <logion> well yeah, that's how i'm working now
14:15:19 <springmeyer> so you saw the 'loadXML()' example here ?
14:15:20 <springmeyer> http://trac.mapnik.org/wiki/OgcServerSvn
14:15:22 <logion> yeah
14:15:28 <logion> I got that running locally actualy
14:15:39 <springmeyer> .g generate_tiles.py
14:15:40 <nikq> springmeyer: http://svn.openstreetmap.org/applications/rendering/mapnik/
14:15:58 <springmeyer> so that is the SVN repository that contains the huge XML that works with OSM DATA
14:16:42 <springmeyer> see the README there
14:16:44 <springmeyer> http://svn.openstreetmap.org/applications/rendering/mapnik/README
14:17:01 <logion> ah
14:17:30 <springmeyer> or the 'Rendering with Mapnik' section here:
14:17:31 <springmeyer> http://wiki.openstreetmap.org/index.php/Mapnik
14:18:15 <logion> that was the guide I used yesterday, but I got stuck after creating the db
14:19:38 <logion> the database is named gis
14:20:34 <springmeyer> sure
14:20:49 <logion> any idea's on postgres command to check its content?
14:21:06 <springmeyer> $ psql -U postgres gis
14:21:20 <springmeyer> \d # lists tables
14:21:41 <springmeyer> \d <table name> # lists fields for table
14:22:05 <springmeyer> in the case of osm data if your import did not work you will not have many columns in your tables
14:22:13 <springmeyer> so that is a good first check
14:22:16 <springmeyer> then...
14:23:30 <springmeyer> select count(*) from planet_osm_roads;
14:23:36 <springmeyer> ( for example)
14:23:38 <logion> the tables have columns
14:24:11 <springmeyer> lots ? >15-20  for roads?
14:24:48 <logion>   Table "public.planet_osm_roads"
14:24:48 <logion>    Column    |   Type   | Modifiers
14:24:48 <logion> -------------+----------+-----------
14:24:48 <logion>  osm_id      | integer  |
14:24:48 <logion>  access      | text     |
14:24:49 <logion>  admin_level | text     |
14:24:51 <logion>  aeroway     | text     |
14:24:53 <logion>  amenity     | text     |
14:24:55 <logion> etc.
14:25:04 <springmeyer> sure, looks good
14:25:14 <logion> 19060 rows in planet_osm_roads (only belgium)
14:25:25 <springmeyer> nice
14:25:29 * springmeyer is jealous :)
14:25:37 <logion> :o
14:25:43 <logion> how come? :p
14:26:19 <springmeyer> I'm in the United States were we've got some catching up to do :)
14:26:48 <logion> :o
14:27:21 <logion> well that database looks like it's in order
14:27:30 <logion> the guide mentions : "Unpack the lowzoom and coastline data you downloaded earlier"
14:30:44 <logion> is this also required when working with the belgium osm files?
14:31:38 <springmeyer> well, if you are just rendering belgium, then perhaps you can grab a different coastlines shapefile from geofabric
14:32:05 <springmeyer> for now I would just comment out that layer in the osm.xml
14:32:17 <springmeyer> <!-- xml code -->
14:32:22 <logion> ah k
14:34:24 <logion> checking out svn http://svn.openstreetmap.org/applications/rendering/mapnik now
14:39:09 <logion> ./customize-mapnik-map >$MAPNIK_MAP_FILE
14:39:12 <logion> errors though
14:41:59 <logion> i just edited the set-mapnik-env to take notice of the database name and user
14:44:01 <logion> does set-mapnik-env expect the files to be in your home directory?(all the paths are beginning with a ~)
14:45:20 <springmeyer> I've rarely used it, sorry
14:45:34 <springmeyer> that does does like a home directory thing
14:46:17 <logion> well generate_image.py doesn't work,
14:46:39 <logion> prolly because it expects the file to be in my home directoy while they are in /tmp
14:46:42 <logion> i'll move the folder
14:46:49 <springmeyer> you might try installing nik2img.py
14:46:57 <springmeyer> $ easy_install nik2img
14:47:03 <springmeyer> then just do:
14:47:27 <springmeyer> $ nik2img.py -m /path/to/mapfile.xml -o image_test.png
14:48:17 <logion> well, I moved the cheeckout osm folder to my home but now it works
14:48:20 <logion> and*
14:48:27 <springmeyer> cool
14:49:30 <logion> but it complains when i execute generate_tiles
14:49:49 <logion> i'll prolly have to comment some stuff out in the xml file
14:50:16 <logion> comment out the shoreline style in the world layeR?
14:50:28 <logion> error => /root/mapnik/world_boundaries/shoreline_300 does not exist in layer 'world'
14:51:24 <springmeyer> yep
14:51:40 <springmeyer> or download your coastlines now and set the path to them there
14:52:25 <logion> well, that geofabrik website doesn't seem to offer any coastlines
14:52:42 <springmeyer> not in the shapefile set?
14:53:04 <springmeyer> downloads.cloudmade.com might have them as well, clipped for belgium
14:53:43 <logion> the shapefile zip contains waterways
14:53:53 <springmeyer> ah
14:53:55 <logion> nothing about coastlines though
14:54:36 <logion> http://downloads.cloudmade.com/europe/belgium#breadcrumbs looks super
14:54:55 <logion> will have to reimport those osm files to pgsql though
14:56:08 <springmeyer> well, for mapnik all you need is the osm.bz2 and coastlines shapefile
14:56:32 <springmeyer> I'm pretty sure that cloudmade distributes the coastlines.shp inside their shapefile bundle
14:57:06 <logion> well there is a coastline.osm file
14:57:20 <logion> there are also shapefiles though, i'll have a look
14:57:39 <springmeyer> ya, that might work but at least the default OSM styles (as you know) are looking for a shapefile format
14:58:07 <logion> but i'll have to recreate my database no?
14:58:15 <springmeyer> no, why?
14:58:40 <logion> are the files on cloudmade the same as those on geofabrik ?
14:58:47 <logion> (the osm files I mean)
14:58:50 <springmeyer> their source is the same yes
14:58:53 <logion> ah
14:59:24 <springmeyer> they are both great distributors of a variety of data, and for mapnik both distribute the planet files for osm2pgsql
14:59:26 <logion> so the coastline shape file on cloudmade will work with the osm's of geofabrik?
14:59:41 <springmeyer> one may be a week or two more updated at any time but I think they are both great
14:59:47 <springmeyer> yes
14:59:55 <logion> i'll try it out then :)
14:59:57 <springmeyer> its just a shapefile that is used for lines
15:00:02 <springmeyer> nothing fancy
15:00:21 <springmeyer> later you may care more about the source to make sure you get the highest resolution data
15:00:28 <springmeyer> but for now? anything should be fine :)
15:00:57 <logion> i'll first try to get it working without coastlines
15:02:19 <logion> allthough there might be something wrong with my db :
15:02:27 <logion>  Schema |        Name        | Type  |  Owner
15:02:27 <logion> --------+--------------------+-------+----------
15:02:27 <logion>  public | geometry_columns   | table | gis
15:02:27 <logion>  public | planet_osm_line    | table | postgres
15:02:27 <logion>  public | planet_osm_point   | table | postgres
15:02:27 <logion>  public | planet_osm_polygon | table | postgres
15:02:29 <logion>  public | planet_osm_roads   | table | postgres
15:02:31 <logion>  public | spatial_ref_sys    | table | gis
15:02:46 <logion> the gis user won't be able to acces to planet_ tables right?
15:04:14 <springmeyer> won't know till you try
15:04:32 <springmeyer> GRANT is the command in postgres to switch access if you need it
15:05:53 <logion> well i switched over to the root user, postgres
15:05:54 <logion> FATAL:  Ident authentication failed for user "postgres" in layer 'leisure'
15:09:34 <ser> springmeyer: thanks for an example, i am studying it now
15:11:51 <ser> but what is an associated query with these results?
15:15:09 <logion> meh, I can't change my postgres user passwd
15:16:37 <ser> logion: more postgres docs reading required :)
15:18:39 <springmeyer> ser: logion: postgres does not have a password :)
15:19:05 <springmeyer> you likely need to edit pg.conf to TRUST all local connections
15:19:24 <springmeyer> ser: you are running Mapnik 0.6.0 or trunk? I can't recall...
15:19:52 <ser> 0.6 but i can switch to trunk if it is needed
15:20:04 <ser> yes, postgres does have passwords
15:20:53 <springmeyer> :) depends
15:20:57 <ser> hehe
15:21:03 <logion> passwd postgres
15:21:07 <logion> :(
15:21:27 <springmeyer> careful though - follow ser's advice and read about 'the postgres user'
15:21:27 <ser> logion: why do you want to change the password?
15:21:38 <springmeyer> it is not the same as a unix user :)
15:21:56 <logion> cause i'm getting "FATAL:  Ident authentication failed for user "postgres" in layer 'leisure'"
15:22:22 <springmeyer> ser: to get GetFeatureInfo to work you have to 1) mark all layers queryable="true" in your XML
15:22:28 <ser> logion: unix passwords are independent from database passwords
15:22:41 <ser> logion: please start from postgres docs
15:22:47 <springmeyer> and 2) generate the right GetCapabilities to get the list of all your layers
15:23:13 <springmeyer> ser: or wait a second and I'll post a patch to trunk for accepting GetFeatureInfo on all layers
15:23:54 <ser> springmeyer: i will happily wait and get trunk
15:24:13 <springmeyer> okay, ya I'm able to do something like this with the patch:
15:24:16 <springmeyer> http://localhost:8000/?LAYERS=__all__&FORMAT=image%2Fpng&SERVICE=WMS&VERSION=1.1.1&REQUEST=GetFeatureInfo&STYLES=&EXCEPTIONS=application%2Fvnd.ogc.se_inimage&SRS=EPSG%3A4326&BBOX=-124.954365%2C47.530893%2C-121.369909%2C48.697518&X=542&Y=80&INFO_FORMAT=text%2Fplain&QUERY_LAYERS=__all__&WIDTH=1229&HEIGHT=400
15:25:14 <ser> ok, everything pasted into my learning book, thanks! i will ask more after i get enough knowledge
15:25:19 <ser> thanks!
15:27:43 <ser> how do you advice me to trace why ogcserver is leaking?
15:28:02 <ser> what toolset should i use?
15:28:48 <springmeyer> hmmm, good question
15:29:21 <springmeyer> I'm not sure about tools, but I'm sure there could be some helpful ones
15:29:44 <logion> one would expect ALTER USER postgres WITH ENCRYPTED PASSWORD 'password'; but no
15:30:01 <springmeyer> overall, it could either be a memory leak in mapnik c++, mapnik boost::python bindings, or the OGCServer code
15:30:13 <ser> logion: /j postgres ;)
15:30:43 <springmeyer> I would guess is it most likely ogcserver or python bindings
15:30:45 <ser> it is rather hard to trace fcgi ap
15:31:05 <ser> so maybe i should run it as independent server first?
15:31:13 <springmeyer> I bet there are tools for the wsgi adapter
15:31:16 <ser> without any "proxies"
15:31:25 <springmeyer> wsgi middleware perhaps
15:31:59 <springmeyer> ser: how are you tracking mem usage for the process?
15:32:15 <ser> simple htop
15:32:34 <springmeyer> k
15:33:00 <ser> i discovered it is leaking when 6GB of memory was taken
15:33:10 <ser> it is growing very fastly
15:35:14 <ser> gotta go, thanks again & see you later
15:35:19 <springmeyer> see ya
15:36:45 *** sanjiv (n=sanjiv@59.180.137.181) has joined #mapnik
16:06:45 <logion> generate_titles.py is taking its time :)
16:10:25 <springmeyer> logion: ya, you need to edit the default bbox's in the script to avoid it trying to render the whole world
16:10:39 <logion> apparently, cause all the tiles are empty
16:11:14 <springmeyer> logion: TileLite is a script I wrote that will allow you to render on demand. sample OpenLayers file to test your setup
16:11:32 <springmeyer> .g springmeyer/tilelite
16:11:33 <nikq> springmeyer: http://bitbucket.org/springmeyer/tilelite/
16:13:40 <logion> hmm
16:15:09 <logion> i'd like to get this to work first
16:17:04 <logion> so editing bbox in the generate_tiles.py should change the rendering area?
16:17:28 <springmeyer> yes, open the script and see some examples at the bottom
16:18:04 <logion> when i run the script I do get alot of : /bin/sh: convert: command not found
16:18:04 <logion> unknown [ 10 - 16 ]:  11 1021 691 p: (-0.52734375, 50.289339253291779) (-0.3515625, 50.401515322782366)
16:20:00 <springmeyer> ya, the script is trying to make the tiles smaller using imagemagick
16:20:19 <springmeyer> so either comment out that line or install imagemagick
16:22:09 <logion> imagemagick did the trick
16:24:42 *** wonxly (i=9e931426@gateway/web/ajax/mibbit.com/x-3f7e5f7495b71d66) has joined #mapnik
16:25:41 <logion> the tiles are still empty though
16:25:44 <wonxly> hi everyone, can anyone help me out with a mapnik problem?  i think i have everything i need setup but when i get tiles rendered it is not showing world boundaries
16:26:21 <wonxly> in my debug output i was getting that query size=0 and features=0
16:26:42 <wonxly> i found a post somewhere that mentioned a problem with the index files...so i regenerated them with shapeindex
16:26:58 <wonxly> now i get query size=83886080 but 0 features still
16:58:36 <logion> any idea's why the generate_image.py reduces a image with just a bg color but nothing else springmeyer?
17:02:30 <logion> owla piola, nvm ^^
17:02:55 *** ortelius (n=chatzill@ip72-197-145-158.sd.sd.cox.net) has joined #mapnik
17:07:26 *** jburgess (n=jburgess@bb-87-80-234-70.ukonline.co.uk) has joined #mapnik
17:13:36 *** xcacou has quit (Remote closed the connection)
17:13:53 *** synax (n=synax@24.222.57.182) has joined #mapnik
17:14:30 <synax> how does mapnik know how to handle PostGIS tables? what I mean is, how does it know what column to pull geometry from?
17:14:43 <synax> w.r.t. http://mapnik-utils.googlecode.com/svn/example_code/postgis/postgis_geometry.py
17:16:29 <springmeyer> geometry columns table created by AddGeometryColumn
17:16:52 <springmeyer> logion: you bbox is likely not right
17:17:07 <springmeyer> wonxly: does it work if you take the .index file away?
17:17:35 <wonxly> i will give it a try now...
17:17:41 <synax> uh, what if your table has multiple geometry columns?
17:18:10 <springmeyer> then you have to give an argument of 'geometry_columns=<alternate_name>'
17:18:23 <springmeyer> only available in 0.6.0 >
17:19:19 <synax> and where do you supply that argument?
17:20:15 <springmeyer> as another keyword argument in Python and another <Parameter name="geometry_column"> in xml
17:20:37 <synax> sorry, I don't follow
17:20:46 <logion> any idea why this isn't giving any output ?(something to do with that output prolly) :
17:20:47 <logion> render_tiles( (3.7029999999999998, 51.050199999999997, 3.7467000000000001, 51.031100000000002) /root/mapnik/osm.xml /root/mapnik/tiles/ 18 21 World )
17:21:08 <springmeyer> whoops, it is actually 'geometry_field'
17:21:14 <synax> I'm specifying the layer datasource to use a PostGIS connection
17:22:13 <springmeyer> in XML or python?
17:22:17 <synax> Python
17:22:24 <springmeyer> what mapnik version?
17:22:31 <synax> whatever's in SVN
17:22:35 <springmeyer> okay
17:22:47 <springmeyer> >>> from mapnik import PostGIS
17:22:54 <springmeyer> >>> help(PostGIS)
17:24:12 <synax> ah, super
17:24:15 <synax> thanks
17:24:17 <springmeyer> logion: I'm not sure
17:24:56 *** aba has quit (brown.freenode.net irc.freenode.net)
17:24:56 *** jlivni has quit (brown.freenode.net irc.freenode.net)
17:26:42 <springmeyer> logion: did you install osm2pgsql from svn source? or from packages?
17:27:02 <logion> hmm, can't remember, i think from packages
17:27:15 <logion> it works though, just not when I raise the zoomlevels
17:28:11 <synax> what takes precedence? The SRS defined on the layer, or the SRID used when the data was inserted?
17:28:14 <springmeyer> if I remember right the osm2pgsql from packages is old enough that long/lat coordinates may be default vs importing in mercator (which is now the default)
17:28:24 <springmeyer> ie. the -m flag
17:29:07 <springmeyer> the default osm.xml uses extents that are in mercator, so I __always__ build osm2pgsql from svn
17:29:26 *** cmarqu_ (i=colin@195.49.138.42) has joined #mapnik
17:29:47 <logion> hmm
17:29:48 <springmeyer> synax: what do you mean 'when the data was inserted' ?
17:29:56 <cmarqu_> Hi guys
17:30:05 <springmeyer> hello cmarqu_ :)
17:30:09 <synax> into the PostgreSQL table
17:30:33 <springmeyer> oh. the layer srs must match the data
17:30:41 <synax> I see...
17:30:49 <synax> my SRID is -1 :/
17:31:36 <springmeyer> well, I'm not sure the SRID in postgis matters to mapnik
17:31:42 <springmeyer> synax: how did you insert the data?
17:31:55 <synax> I actually didn't insert it, so I can't answer that question
17:33:05 <cmarqu_> springmeyer: I just saw (for the 2nd time I think) something strange with tilelite: when I let it monitor and reload the map file, it didn't render two layers with shield symbolizers. When I kill liteserv.py and start it again, they appear.
17:33:31 <wonxly> without the index file the processor is staying at 100% and it is taking forever...
17:34:25 <synax> what *does* mapnik care about in terms of PostGIS geometry? I don't even know how to debug my problems
17:35:11 <springmeyer> cmarqu_: interesting
17:35:18 <springmeyer> cmarqu_: did you see this: http://trac.mapnik.org/ticket/294 ?
17:36:49 <springmeyer> so in relation to that
17:37:01 <cmarqu_> springmeyer: I didn't see the bugreport until now, and I also didn't see the effect of it. But maybe jburgess had seen it a while ago, he talked about something like this when I tried the hillshading patches for the first time.
17:37:24 <springmeyer> hearing that symbols dissapear seems perhaps related to overall thread-safety issue
17:37:43 <cmarqu_> Oh, ok, that may be.
17:38:05 <logion> any idea why mapnik won't generate any tiles for zoom levels > 11 ?
17:38:23 <logion> ideas*
17:38:31 <springmeyer> in other words, I'm not surprised. trying to reload the map objects but get the process (and other objects in memory) hits these issues
17:38:41 <springmeyer> get/keeping
17:39:05 <cmarqu_> Sounds convincing
17:39:17 <springmeyer> cmarqu_: so a better solution may be to ditch the mapfile watching as something inside TileLite
17:39:41 <springmeyer> and rather have it simply be something that starts/restarts the server process (ie 'liteserv.py')
17:40:30 <springmeyer> the reason I did not go that route is that I figured if it were internal to tilelite.py you could benefit from reloading when deployed with apache too
17:40:58 <springmeyer> but I think otherwise now on that benefit :)
17:41:01 <nikq> Mapnik Trac: Ticket #300 (Improve tests) updated | http://trac.mapnik.org/ticket/300#comment:1
17:41:03 <nikq> Mapnik Trac: test_output.txt attached to Ticket #300 | http://trac.mapnik.org/attachment/ticket/300/test_output.txt
17:41:12 <nikq> Mapnik Trac: Ticket #300 (Improve tests) updated | http://trac.mapnik.org/ticket/300#comment:1
17:41:57 <springmeyer> synax: what are your problems? what does not work?
17:41:58 <cmarqu_> springmeyer: I wouldn't mind such a change.
17:42:04 <springmeyer> cmarqu_: k
17:42:40 <synax> springmeyer: I think the missing projection information may be the issue.
17:42:46 <springmeyer> cmarqu_: we may need to switch to something like cherrypy server that support code reloading
17:43:13 <springmeyer> and try to tailor that to support watching the mapfile instead of just the python code
17:43:43 *** logion has quit ()
17:45:37 <cmarqu_> springmeyer: Just now, I see that the shield symbols may disappear just with the presence of a gdal layer, indeed. I had commented that out, but it didn't make the shields appear again until I restarted liteserv.
17:46:06 <springmeyer> interesting
17:47:09 <cmarqu_> I saw tilelite give me new tiles without the gdal layer stuff so the reloading worked in principle, but also no shields (until the restart)
17:47:54 <cmarqu_> Now I got the gdal layer back in, restarted liteserv, but no shields still.
18:09:01 *** jlivni (n=josh@76.14.74.234) has joined #mapnik
18:09:01 *** aba (i=sam@libya.turmzimmer.net) has joined #mapnik
18:20:06 <nikq> Mapnik Trac: Ticket #301 (Mapnik does not issue a UserWarning when a bad (or no) datasource type is  ...) created | http://trac.mapnik.org/ticket/301
18:24:33 *** Ldp_ (n=thid@osm.xs4all.nl) has joined #mapnik
18:25:32 <nikq> Mapnik Trac: Ticket #302 (load_map is silent when a PostGIS layer is used if the postgis input  ...) created | http://trac.mapnik.org/ticket/302
18:35:14 <nikq> Mapnik Trac: Ticket #303 (ogcserver: Support built in xml fallback if lxml is not available) created | http://trac.mapnik.org/ticket/303
18:37:17 <nikq> Mapnik Trac: Ticket #304 (ogcserver: support GetFeatureInfo on all layers with LAYERS=__all__   ...) created | http://trac.mapnik.org/ticket/304
18:39:55 *** synax has quit ()
18:40:40 <nikq> Mapnik Trac: Changeset [1085]: Support xml lib fallback and GetFeatureInfo on all layers - closes #303  ... | http://trac.mapnik.org/changeset/1085
18:42:32 <nikq> Mapnik Trac: Changeset [1086]: formatting | http://trac.mapnik.org/changeset/1086
18:44:03 <nikq> Mapnik Trac: Ticket #304 (ogcserver: support GetFeatureInfo on all layers with LAYERS=__all__   ...) closed | http://trac.mapnik.org/ticket/304#comment:1
18:44:14 <nikq> Mapnik Trac: Ticket #303 (ogcserver: Support built in xml fallback if lxml is not available) updated | http://trac.mapnik.org/ticket/303#comment:1
18:47:57 <nikq> Mapnik Trac: Ticket #305 (ogcserver: Support map buffer_size) created | http://trac.mapnik.org/ticket/305
18:51:02 *** sanjiv has quit (Connection timed out)
18:52:20 *** sanjiv (n=sanjiv@59.180.133.147) has joined #mapnik
18:56:18 *** Ldp__ (n=thid@osm.xs4all.nl) has joined #mapnik
18:56:58 *** Ldp_ has quit (Read error: 60 (Operation timed out))
19:02:36 *** Ldp__ is now known as Ldp_
19:04:51 *** wonxly has parted #mapnik ()
20:02:36 *** sanjiv has quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032711]")
20:17:34 *** ortelius has quit ("ChatZilla 0.9.84 [Firefox 3.0.8/2009032609]")
20:35:53 *** springmeyer has quit ()
21:07:24 *** blodulv (n=beau@fw099rv1.costco.com) has joined #mapnik
21:10:59 *** blodulv has quit (Read error: 104 (Connection reset by peer))
21:11:12 *** blodulv (n=beau@fw099rv1.costco.com) has joined #mapnik
21:14:58 *** blodulv has quit (Read error: 104 (Connection reset by peer))
21:15:11 *** blodulv (n=beau@fw099rv1.costco.com) has joined #mapnik
21:15:18 *** blodulv has parted #mapnik ()
23:40:27 <nikq> Mapnik Trac: Ticket #242 (Drops the whole style or layer when a filter references a non-existent  ...) updated | http://trac.mapnik.org/ticket/242#comment:2