#mapnik log: Friday 05, February 2010

2010 | 02

previous | next
00:10:35 *** ajashton has quit (Remote host closed the connection)
00:19:49 *** Ldp__ has quit ()
00:41:56 *** jctull has quit (Quit: jctull)
01:16:16 *** chad_burt has quit (Quit: Leaving...)
01:21:25 *** cgs_bob has quit (Read error: Connection reset by peer)
01:34:04 *** springmeyer has quit (Quit: springmeyer)
01:54:14 <nikq> Mapnik Trac: Ticket #511 (Change in alpha value in Mapnik color constructor between Mapnik 0.7 and  ...) created | http://trac.mapnik.org/ticket/511
02:14:09 *** tcarobruce has quit (Quit: tcarobruce)
02:51:21 *** ajturner (~ajturner@pool-72-66-109-70.washdc.fios.verizon.net) has joined #mapnik
02:53:12 *** cgs_bob (~bob@122.135-78-65.ftth.swbr.surewest.net) has joined #mapnik
03:14:40 *** ajturner has quit (Quit: ajturner)
03:32:32 *** cgs_bob has quit (Ping timeout: 240 seconds)
04:45:36 *** cgs_bob (~bob@122.135-78-65.ftth.swbr.surewest.net) has joined #mapnik
05:34:20 *** tcarobruce (~tim@c-98-210-194-147.hsd1.ca.comcast.net) has joined #mapnik
05:58:02 *** HounD (~HounD@unics1.grfc.ru) has joined #mapnik
06:54:34 *** jfreeman has quit (Read error: Connection reset by peer)
07:34:38 *** tcarobruce has quit (Quit: tcarobruce)
07:39:36 *** dodobas has quit (Changing host)
07:39:36 *** dodobas (~dodobas@unaffiliated/dodobas) has joined #mapnik
07:49:41 *** Ldp__ (~thid@osm.xs4all.nl) has joined #mapnik
08:12:35 *** jfreeman (~jfreeman@60-242-98-237.static.tpgi.com.au) has joined #mapnik
08:19:33 *** jfreeman has quit (Ping timeout: 260 seconds)
09:27:46 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
09:27:46 *** mperry has quit (Read error: Connection reset by peer)
09:27:47 *** mperry_ is now known as mperry
11:01:38 *** jfreeman (~jfreeman@60-242-98-237.static.tpgi.com.au) has joined #mapnik
11:17:27 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
11:17:27 *** mperry has quit (Read error: Connection reset by peer)
11:17:28 *** mperry_ is now known as mperry
11:32:58 *** D3f0 (~D3f0@www.transpa-sa.com.ar) has joined #mapnik
11:43:19 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
11:43:19 *** mperry has quit (Read error: Connection reset by peer)
11:43:20 *** mperry_ is now known as mperry
12:20:01 *** jfreeman has quit (Read error: Operation timed out)
12:24:45 *** jfxberns (~jfxberns@ppp-61-90-15-181.revip.asianet.co.th) has joined #mapnik
13:08:05 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
13:08:05 *** mperry has quit (Read error: Connection reset by peer)
13:08:06 *** mperry_ is now known as mperry
13:14:15 *** ajturner (~ajturner@209.155.228.129) has joined #mapnik
14:12:25 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
14:20:15 *** jfxberns_ (~jfxberns@ppp-58-8-117-48.revip2.asianet.co.th) has joined #mapnik
14:20:30 *** jfxberns has quit (Ping timeout: 240 seconds)
14:20:31 *** jfxberns_ is now known as jfxberns
14:30:56 *** artem has quit (Read error: No route to host)
14:31:15 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
14:37:10 *** avar has parted #mapnik (None)
14:42:18 *** ajashton (~aj@c-69-136-229-112.hsd1.dc.comcast.net) has joined #mapnik
14:51:13 *** HounD has parted #mapnik (None)
14:55:24 <nikq> Mapnik Trac: Ticket #511 (Change in alpha value in Mapnik color constructor between Mapnik 0.7 and  ...) updated | http://trac.mapnik.org/ticket/511#comment:1
15:34:27 *** willwhite (~diggersf@c-69-136-229-112.hsd1.dc.comcast.net) has joined #mapnik
16:07:04 *** springmeyer (~springmey@c-76-28-156-154.hsd1.wa.comcast.net) has joined #mapnik
16:31:30 *** jfxberns has quit (Quit: Going to hide someplace dark and quiet...)
16:31:54 *** jctull (~jctull@ppp-71-142-138-235.dsl.renocs.pacbell.net) has joined #mapnik
16:59:45 *** D3f0 has quit (Ping timeout: 256 seconds)
17:13:52 <springmeyer> artem: cool, thanks for feedback on colors
17:14:36 <springmeyer> I rarely create colors as rgba, just hit that running a script I have which does use that approach
17:15:06 * springmeyer honestly does not fully understand the interplay of using rgba + opacity attribute in symbolizers
17:17:47 *** jctull has quit (Quit: jctull)
17:39:33 *** D3f0 (~D3f0@190.177.84.42) has joined #mapnik
17:56:39 *** chad_burt (~chad_burt@mm-01.msi.ucsb.edu) has joined #mapnik
18:09:06 *** mperry has quit (Ping timeout: 276 seconds)
18:09:58 *** tomhughes has quit (Quit: Coyote finally caught me)
18:10:05 *** tomhughes (~tom@gosford.compton.nu) has joined #mapnik
18:17:35 *** tcarobruce (~tcarobruc@adsl-75-10-247-30.dsl.pltn13.sbcglobal.net) has joined #mapnik
18:22:06 <artem> springmeyer: solved gdal issue - turned out to be subtle bug/feature when creating *.vrt from images with diff resolution.
18:22:49 <springmeyer> ah, artem good news
18:24:28 <artem> nice write up about nikweb : http://www.google.com/url?sa=X&q=http://www.alpstein.com/blog/outdoor-geo-technology/2010/02/05/nikweb-mapnik-geojson-web-service-for-nice-static-maps/&ct=ga&cd=d9yg00R_Qx8&usg=AFQjCNH23HFjAPljjn2JJPIQfEL8anZkOw
18:25:57 *** mperry (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
18:36:37 <springmeyer> nice!
19:05:02 *** willwhite has quit (Ping timeout: 272 seconds)
19:05:13 *** ajashton has quit (Ping timeout: 264 seconds)
19:05:17 *** aj_ashton (~aj@c-69-136-229-112.hsd1.dc.comcast.net) has joined #mapnik
19:06:02 *** willwhite (~diggersf@c-69-136-229-112.hsd1.dc.comcast.net) has joined #mapnik
19:10:12 *** artem has quit (Quit: artem)
19:10:25 *** cmarqu has quit (Remote host closed the connection)
19:36:28 *** aj_ashton has quit (Remote host closed the connection)
19:36:50 *** ajashton (~aj@c-69-136-229-112.hsd1.dc.comcast.net) has joined #mapnik
19:45:58 *** mak (~mak@vobster.nepharia.org) has joined #mapnik
19:54:07 <mak> I have a problem with the ShieldSymbolizer. The problem is that many of my tiles (http://i.imgur.com/vmSD3.png) have few or no shields, while others have what I'd expect given my settings.
19:54:38 <mak> Any ideas as to why I have such a difference between the number of shields in my tiles?
19:55:26 <mak> (buffer=128, tiles=256x256, spacing=60)
19:59:17 *** ajturner has quit (Quit: ajturner)
19:59:41 *** ajturner (~ajturner@209.155.228.129) has joined #mapnik
20:06:36 *** AndrewOrr (~andrew@vobster.nepharia.org) has joined #mapnik
20:06:53 <AndrewOrr> Sounds questionable to me.
20:06:54 *** AndrewOrr has parted #mapnik (None)
20:11:11 <springmeyer> mak: hmm, thats odd
20:11:29 <springmeyer> so its the yellow lines in other tiles that should have shields?
20:11:43 <springmeyer> not the black/brown as well? (just to be sure)
20:12:05 <springmeyer> first thing I would try is placement="vertex"
20:12:14 <mak> The bloak lines that look like a grid are the tile borders, put there for debugging.
20:12:24 <mak> ^black
20:12:28 <springmeyer> sure
20:12:46 <mak> All other lines have shield symbolizers attached.
20:13:11 <springmeyer> ruh roh
20:13:37 <springmeyer> how are these lines made up, lots of small segments or single large segments?
20:13:43 <springmeyer> single or multi linestrings?
20:14:36 <mak> They're multilinestrings that usually contain a single linestring.
20:15:06 <springmeyer> hmm, then ya, try placement="vertex" first
20:15:18 <mak> Okay, I'll try that now.
20:15:48 <springmeyer> and at the layer level try: clear_label_cache="true"
20:24:27 *** jfreeman (~jfreeman@60-242-98-237.static.tpgi.com.au) has joined #mapnik
20:25:05 <mak> I just managed to find the bug that has plagued me for the last week. A bunch of raods were filtered out because of a typo in a rule.
20:25:16 <mak> I'm going to kick off another render and hope this fixes everything.
20:26:12 <Phurl> ok guys
20:26:35 <Phurl> so what about these two renderers, cairo and agg
20:26:46 <Phurl> which is better? which do you use?
20:28:23 <Phurl> http://trac.mapnik.org/wiki/MapnikRenderers
20:28:30 <Phurl> what is faster however?
20:29:19 <springmeyer> mak, okay, hope so :)
20:29:39 <springmeyer> Phurl: AGG is faster
20:29:47 <Phurl> springmeyer, ok.
20:29:58 <Phurl> that is what I thought.
20:30:05 <Phurl> thanks
20:37:23 <Phurl> springmeyer, i am looking at a generic rendering interface
20:37:30 <Phurl> was wondering what api looks cleaner
20:37:49 <Phurl> i mean, ideally we would have a single interface to both renders..
20:38:03 <Phurl> the cairo implementation looks cleaner.
20:38:30 <springmeyer> yes, it is a bit more oo, as it was written second
20:39:51 <mak> Fixing that typo in my XML file fixed the bug that I've been stuck on for the past week...
20:40:09 <Phurl> so springmeyer if I abstract that cairo interface a bit, wrapping it.
20:40:20 <Phurl> then we can hid agg behind it in theory
20:40:30 <springmeyer> mak: that caused the missing shields?
20:40:35 <mak> But now I'm back to cut-off shields, again (http://i.imgur.com/4Wcbo.png). Any ideas?
20:40:43 <springmeyer> oh shoot
20:41:15 <dodobas> mak: you a bigger buffer
20:41:28 <mak> How big?
20:41:55 <springmeyer> mak: please confirm how you fixed the missing shields?
20:42:05 <Phurl> ok dinner time
20:42:10 <dodobas> mak how big are you using now?
20:42:13 <mak> I had a filter rule that was missing a clause, basically.
20:42:18 <springmeyer> k
20:42:20 <mak> dodobas: 128 on 256 tiles.
20:43:48 <dodobas> hmm...are you using tilecache?
20:44:11 <mak> Yes and no...
20:44:53 <dodobas> elaborate...
20:44:55 <mak> I was using TileCache for everything before, but I'm serving on an 8-core, so I wrote my own simple renderer that puts PNGs in the right spots and then TileCache serves them.
20:45:09 <mak> (So I can have 8 rendering threads.)
20:46:06 <mak> I am using TileCache's definitions for the bounding boxes of the tiles, though.
20:46:27 <dodobas> well, i just partition area of interest in 8 parts, and start 8 tilecache proceses... :)
20:46:57 <dodobas> and use 5,5 metatile with 256 buffer size
20:49:09 <Ldp__> mak: we have a mod for renderd that also makes it spit out png files (while still rendering metatiles)
20:50:35 <mak> Could you point me to a URL for that?
20:50:55 <Ldp__> http://trac.openstreetmap.org/ticket/2653
20:51:04 <mak> (I tried to get mod_tile and renderd working once, but never managed it.)
20:51:36 <Ldp__> we also have one to be able to set the output format per tileset (http://trac.openstreetmap.org/ticket/2654) but don't try anything else but png or png256 for now
20:53:08 <Ldp__> mak: we're using the png save ability together with a handler_tile for Cherokee. I don't know how well this patch will play with mod_tile in Apache.
20:53:52 *** jctull (~jctull@adsl-75-0-13-129.dsl.renocs.sbcglobal.net) has joined #mapnik
20:57:32 <mak> Here's an opinion question: What is the nicest way to serve tiles through OpenLayers? TileCache? mod_tile? Something else? The only requirements are that it use Mapnik, pre-render everything, and be able to render in parallel.
21:22:31 <springmeyer> my opinion is mod_tile, as long as you want spherical_merc tiles, and tilecache if you don't
21:22:59 <springmeyer> render in parallel you mean when generating a tile on-demand via a web request?
21:24:22 <mak> No, when pre-rendering things to fill a cache of tiles.
21:25:07 <mak> My bosses require that nothing be generated on-demand, basically.
21:26:39 <springmeyer> hmm, then you don't need a server, eh?
21:27:32 <springmeyer> e.g. if you cache your tiles in the x/y/x scheme you can just point OL at those tiles and serve them statically?
21:27:51 <springmeyer> ... don't need a tile server/renderer I mean
21:28:19 <springmeyer> you can just use generate_tiles.py, for example
21:28:30 <mak> I didn't know OL could handle just being pointed at a directory like that.
21:29:09 <springmeyer> sure, just need to be an directory exposed by apache or something like that
21:29:10 <mak> But yes, that would be ideal.
21:31:01 <mak> I cannot even begin to express how useful you lot have been today.
21:31:20 <mak> Thanks for all your help, I'm going to look into the things you've mentioned.
21:45:13 <springmeyer> mak: good
21:45:21 <springmeyer> mak: one question for you
21:45:49 <springmeyer> I'm interested in writing some sort of xml validator for mapnik xml and cascadenik mml
21:46:08 <springmeyer> to both catch potential errors and inefficiencies
21:46:41 <springmeyer> could you paste the mistake you made somewhere? I need to start some testcases
21:47:19 *** jfreeman has quit (Read error: Operation timed out)
21:47:33 <mak> My mistake wasn't invalid XML...
21:48:00 <mak> It was purely a braindead mistake of mine. Copy-and-paste gone wrong, effectively.
21:53:20 <springmeyer> ah, okay, nm
21:55:54 <Phurl> http://github.com/h4ck3rm1k3/MapNickAutotools/blob/master/src/agg_renderer.cpp see the class TRenderWrapper on line 973 for the start of the wrapper
22:02:35 <Phurl> ok more tomorrow. laters
23:06:43 *** chad_burt has quit (Quit: Leaving...)
23:09:28 *** jctull has quit (Quit: jctull)
23:13:26 *** jfreeman (~jfreeman@150.203.164.44) has joined #mapnik
23:19:30 *** willwhite has quit (Quit: willwhite)
23:31:52 *** drewbert (~drewby@96.241.159.224) has joined #mapnik
23:43:04 *** ajashton has quit (Remote host closed the connection)
23:51:41 *** jfreeman has quit (Remote host closed the connection)
23:52:29 *** jfreeman (~jfreeman@150.203.164.44) has joined #mapnik
23:55:30 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
23:55:42 *** artem has quit (Client Quit)