00:00:23 <springmeyer> k, cool
00:07:10 <springmeyer> okay, got it
00:08:30 <springmeyer> ajashton: hg pull -u #tilelite
00:09:57 *** drewby has parted #mapnik (None)
00:10:57 <springmeyer> ajashton: also svn up cascadenik
00:11:44 <CIA-31> mapnik-utils: dane.springmeyer * r901 /trunk/serverside/cascadenik/ (cascadenik-compile.py cascadenik/compile.py): by default don't move image files that are already local, but add an option, -m, to force moving them to the location specified by the -d flag - further addresses Issue #19
00:11:44 <nikq> Ticket #19: 'Shaping' support, http://trac.mapnik.org/ticket/19
00:12:17 <springmeyer> heh, I wish it were that number 19
00:13:40 <springmeyer> ajashton: not copying (and opening with PIL) local images should speed up the Cascadenik compile time remarkably
00:14:59 <ajashton> springmeyer: anything I need to do other than hg pull/svn up and 'sudo python setup.py install' each?
00:15:10 <springmeyer> nope
00:17:11 <springmeyer> ajashton:  not working?
00:17:45 <ajashton> no, same as before "RuntimeError: /tmp/shp_custom/hti-dom-border does not exist"
00:18:54 <springmeyer> hrm, I must be missing something in my test cases
00:19:42 <ajashton> I've been defining relative paths via './' ...that have anything to do with it?
00:20:38 <springmeyer> hmm, shouldn't
00:20:52 <springmeyer> so, lets recap: that's a shapefile that it is failing on?
00:21:00 <springmeyer> what is the actual path?
00:24:13 <ajashton> ~/devseed/maps/haiti/shp_custom/hti-dom-border.shp ..my working directory calling liteserv.py is ~/devseed/maps/haiti/
00:24:33 <ajashton> (where haiti.mml is)
00:24:42 <springmeyer> okay, I'll try to replicate that
00:25:03 <springmeyer> what is the exact line in the .mml ?
00:25:08 <springmeyer> (for the shape)
00:25:35 <ajashton> 519 ..its the first shapefile defined
00:25:51 <springmeyer> oh, I mean the text of the line...
00:25:58 <ajashton> lol of course
00:26:02 <springmeyer> :)
00:26:07 <ajashton> <Parameter name="file">./shp_custom/hti-dom-border</Parameter>
00:26:13 <springmeyer> k, great
00:28:26 <springmeyer> hmm, can't replicate
00:28:46 <springmeyer> ajashton: sorry, can you try to re-install tilelite?
00:28:54 <springmeyer> maybe the change did not get pulled?
00:30:11 <springmeyer> also, would help to have you go into the cascadenik/doc directory
00:30:27 <springmeyer> and see if those examples work for you
00:33:27 <springmeyer> ajashton: you are running on linux?
00:33:34 <ajashton> yes
00:33:47 <springmeyer> perhaps the detection of local paths is working differently then on mac (what I testing on)
00:33:58 * springmeyer goes to boot up karmic...
00:34:53 <ajashton> I recloned/installed tilelite & it didn't change anything
00:35:04 <springmeyer> okay, shoot
00:35:19 <springmeyer> I figure I'll see the problem once I can test on linux.
00:35:27 <springmeyer> I think we're close, thanks for helping
00:38:00 <ajashton> I also ran the mml files in doc/ ..my versions are getting cropped 40-80px wider on the top & bottom vs the examples
00:39:06 <springmeyer> ah crap, forgot that PIL could be used for that
00:40:08 <springmeyer> so just to confirm, you are setting the w,h in the mms which is different that the source image so that they get scaled?
00:42:58 <ajashton> I did not edit the files at all.. they don't seem to contain any width/height info
00:43:20 <ajashton> I am not very familiar with nik2img.. I just ran the command you pasted earlier
00:43:29 <springmeyer> oh, oh
00:43:40 <springmeyer> sorry I thought you were talking about symbolizer images
00:44:02 <ajashton> oh, no
00:44:04 <springmeyer> ya, the image dimensions of the examples will likely be different with nik2img
00:44:11 <ajashton> ok
00:44:12 <springmeyer> so does
00:44:27 <springmeyer> Cascadnik support scaling symbolizer images?
00:44:51 <springmeyer> I didn't notice that support but it seems like it would be desirable
00:45:06 <ajashton> not sure.. I've never done that
00:45:14 <springmeyer> ah, okay
00:46:39 <ajashton> there are point-width, point-height parameters.. but I tend to just make multiple sizes of pngs
00:47:08 <ajashton> or are you talking patterns?
00:47:49 <springmeyer> nope, talking symbols, so images used with Point/Shield symbolizers
00:51:40 * springmeyer fights with VM of lucid
01:04:28 <ajashton> springmeyer: I'm heading out now. Thanks for these fixes, going to save me a bunch of time :)
01:04:48 <springmeyer> ajashton: you're welcome. catch you later
01:04:59 *** ajashton has quit (Remote host closed the connection)
01:12:20 *** Ldp__ has quit ()
01:19:45 *** chad_burt has quit (Quit: Leaving...)
01:31:02 *** cgs_bob_ has quit (Ping timeout: 246 seconds)
01:45:58 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
01:48:10 *** mperry has quit (Ping timeout: 265 seconds)
01:48:10 *** mperry_ is now known as mperry
01:58:31 *** ajturner has quit (Quit: ajturner)
02:04:14 *** springmeyer has quit (Quit: springmeyer)
02:06:23 *** tcarobruce has quit (Quit: tcarobruce)
02:49:41 *** drewbert (~drewby@96.241.159.224) has joined #mapnik
03:45:50 *** cgs_bob (~bob@239.sub-75-208-141.myvzw.com) has joined #mapnik
04:12:17 *** cgs_bob_ (~bob@122.135-78-65.ftth.swbr.surewest.net) has joined #mapnik
04:13:39 *** cgs_bob has quit (Ping timeout: 265 seconds)
05:05:36 *** Belanchuk (~Belanchuk@89.204.23.129) has joined #mapnik
05:13:02 *** cgs_bob_ is now known as cgs_bob
06:15:45 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
06:15:45 *** mperry has quit (Read error: Connection reset by peer)
06:15:46 *** mperry_ is now known as mperry
06:20:33 *** HounD (~HounD@unics1.grfc.ru) has joined #mapnik
06:21:37 *** HounD has parted #mapnik (None)
06:22:48 <CIA-31> mapnik-utils: dane.springmeyer * r902 /trunk/nik2img/ (CHANGELOG.txt setup.py nik2img.py): setup for the nik2img 0.4.1 release
06:26:56 <CIA-31> mapnik-utils: dane.springmeyer * r903 /tags/nik2img/0.4.1: tagging nik2img 0.4.1 release
06:26:56 <CIA-31> mapnik-utils: dane.springmeyer * r904 /tags/nik2img/nik2img-0.4.1.tar.gz: adding nik2img 0.4.1 tagged tarball
06:26:57 <CIA-31> mapnik-utils: dane.springmeyer * r905 /trunk/nik2img/deploy.py: turn deploy script debug off
06:26:57 <CIA-31> mapnik-utils: dane.springmeyer * r906 /trunk/nik2img/CHANGELOG.txt: fix missing char
07:19:26 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
07:19:26 *** mperry has quit (Read error: Connection reset by peer)
07:19:27 *** mperry_ is now known as mperry
07:29:21 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
07:29:22 *** mperry has quit (Read error: Connection reset by peer)
07:29:23 *** mperry_ is now known as mperry
07:31:46 *** drewbert has parted #mapnik (None)
07:42:25 *** PhurlIpv4 has quit (Ping timeout: 260 seconds)
07:42:51 *** PhurlIpv4 (~mdupont@cl-1773.dus-01.de.sixxs.net) has joined #mapnik
08:38:30 *** HounD1 (~HounD@unics1.grfc.ru) has joined #mapnik
09:05:48 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
09:05:48 *** mperry has quit (Read error: Connection reset by peer)
09:05:49 *** mperry_ is now known as mperry
09:33:29 *** PhurlIpv4 has quit (Ping timeout: 260 seconds)
09:44:34 *** Ldp__ (~thid@osm.xs4all.nl) has joined #mapnik
10:53:16 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
10:53:16 *** mperry has quit (Read error: Connection reset by peer)
10:53:17 *** mperry_ is now known as mperry
11:01:54 *** Belanchuk has quit (Ping timeout: 248 seconds)
11:26:01 *** D3f0 (~D3f0@www.transpa-sa.com.ar) has joined #mapnik
11:38:42 *** HounD1 has quit (Ping timeout: 240 seconds)
11:47:17 *** HounD (~HounD@unics1.grfc.ru) has joined #mapnik
11:56:44 *** Phurl (~mdupont@ip-81-210-228-126.unitymediagroup.de) has joined #mapnik
13:18:54 *** HounD1 (~HounD@unics1.grfc.ru) has joined #mapnik
13:20:58 *** HounD has quit (Ping timeout: 240 seconds)
13:54:00 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
14:13:49 <nikq> Mapnik Trac: Ticket #509 (mapnik::query resolution needs both x and y components) created | http://trac.mapnik.org/ticket/509
14:17:10 *** artem has quit (Quit: artem)
14:18:48 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
14:37:23 *** ajturner (~ajturner@209.155.228.129) has joined #mapnik
14:39:31 *** D3f0 has quit (Quit: Saliendo)
14:53:16 *** chad_burt (~chad_burt@mm-01.msi.ucsb.edu) has joined #mapnik
14:57:06 *** mperry has quit (Read error: Connection reset by peer)
15:00:31 *** mperry (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
15:00:42 *** chad_burt has quit (Quit: Leaving...)
15:00:45 *** chad_burt (~chad_burt@mm-01.msi.ucsb.edu) has joined #mapnik
15:04:35 *** ajashton (~aj@c-69-136-229-112.hsd1.dc.comcast.net) has joined #mapnik
15:06:45 *** HounD1 has parted #mapnik (None)
15:48:17 *** artem has quit (Quit: artem)
15:50:03 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
15:55:15 <Phurl> ok, now got the osm layer to link statically and have the plugin registered statically
15:55:58 <Phurl> now working on refactoring the ::get_symbolizers to use a base class and virtual interface instead of funny boost stuff
15:56:07 *** springmeyer (~springmey@c-76-28-156-154.hsd1.wa.comcast.net) has joined #mapnik
15:56:48 <Phurl> springmeyer, ok, now got the osm layer to link statically and have the plugin registered statically. now working on refactoring the ::get_symbolizers to use a base class and virtual interface instead of funny boost stuff
16:00:49 * springmeyer wonders if he'll recognize mapnik when you are done ;)
16:00:56 * springmeyer needs coffee
16:01:28 <artem> hey springmeyer
16:03:18 <Phurl> :)
16:03:21 <Phurl> springmeyer, you will
16:03:29 <Phurl> my changes are not much...
16:12:25 <Phurl> typedef boost::variant< point_symbolizer,
16:12:25 <Phurl> line_symbolizer, line_pattern_symbolizer,
16:12:25 <Phurl> polygon_symbolizer, polygon_pattern_symbolizer,
16:12:25 <Phurl> raster_symbolizer, shield_symbolizer,
16:12:25 <Phurl> text_symbolizer >
16:12:31 <Phurl> that variant type
16:12:36 <Phurl> is a very bad design idea
16:12:46 <Phurl> it exposes the details to the users
16:12:53 <Phurl> and makes it very hard to change anything
16:13:11 <Phurl> that is what virtual base classes are for...
16:27:06 <artem> Phurl: do you mind elaborate about your design ideas :) ?
16:27:18 <Phurl> no problem.
16:27:31 <Phurl> I would like to have a base class
16:27:34 <Phurl> symbolizer_base
16:27:44 <Phurl> that is the parent class for all symbolizers
16:27:54 <Phurl> that has only a virtual interface
16:28:13 <Phurl> and provides a consistent way to call the functions in the various symbolizers
16:28:26 <Phurl> that way, you can write your code to use that interface
16:28:41 <Phurl> and it will be dispatched to the right method via a virtual table
16:28:44 <Phurl> in c++
16:29:20 <Phurl> that is at least my idea
16:29:45 <Phurl> it is still in progress... but variant types are a bad idea
16:29:52 <artem> why?
16:29:56 <Phurl> at least in c++ there should never be a need
16:30:06 <Phurl> because it exposes the internals
16:30:15 <artem> ??
16:30:19 <Phurl> and creates a strong coupling betweent he components
16:30:36 <Phurl> well you need to have all the classes used in the variant defined
16:30:46 <Phurl> and then you have compile dependancies on all of them
16:30:57 <Phurl> they can be isolated
16:31:01 <Phurl> and factored out
16:32:46 <artem> I suggest you read : http://www.boost.org/doc/libs/1_42_0/doc/html/variant.html#variant.motivation
16:32:54 <artem> for starters :)
16:33:03 <Phurl> yes, i saw that
16:33:09 <Phurl> but it is not convincing me
16:34:12 <Phurl> I am going to work it through
16:34:20 <Phurl> and see if I can get it working
16:34:20 <artem> Phurl: sure, good luck
16:34:35 <Phurl> i see that mapnik is too interdependant
16:34:50 <Phurl> there are too many dependancies
16:34:56 <Phurl> and that creates larger compile times
16:35:49 <artem> i agree code base can be refactored to reduce compile times etc it just we didn't have a time and also It wasn't a problem for me
16:35:54 <Phurl> of course
16:36:15 <Phurl> i will see if I can get it to work
16:36:29 <artem> ok
16:36:31 <Phurl> but, the good news is that I have now something that can run
16:36:48 <Phurl> and i can work productivly
16:36:55 <Phurl> and I can change the implemention of a method
16:36:59 <Phurl> without recompiling everythign
16:37:10 <Phurl> because I moved some code out into the source modules
16:37:21 <Phurl> that is another thing with the variant
16:37:28 <Phurl> if you change one line in a symbolizer
16:37:33 <Phurl> then you need to recompile everythign
16:37:43 <Phurl> because everything uses the variant
16:37:50 <Phurl> and that uses everythign...
16:38:37 <Phurl> i mean if you do all your work in python, you dont care
16:38:47 <Phurl> but if you are working in c++, it is very painful
16:40:53 <artem> Phurl: well, if it works for you then great.
16:41:36 <Phurl> well, i am working on it
16:50:20 <StormTide> btw springmeyer, just rebuilt trunk, and it builds properly for me.
16:52:13 <artem> Phurl: fiy I'm planning to re-implement symbolisers to be key=tag containers - some symbolizers e.g text_symbolizer got too blobby for my liking.
16:52:42 <Phurl> ok
16:53:12 <Phurl> it will take me months to get up to speed on this code.
16:57:01 <nikq> Mapnik Trac: Ticket #509 (mapnik::query resolution needs both x and y components) closed | http://trac.mapnik.org/ticket/509#comment:1
17:04:40 <StormTide> springmeyer, actually, scratch that, my text rendering died
17:04:49 <StormTide> the depths numbers dont render anymore
17:05:46 * springmeyer coffee in hand, omelets finished for wife
17:05:50 <springmeyer> hey artem!
17:05:54 <artem> hey
17:05:55 <springmeyer> nice stuff :)
17:06:19 <springmeyer> that raster thing had been bugging me, dodobas ran into it but I had not been able to replicate
17:06:19 <artem> sure, which one ?
17:06:25 <springmeyer> ha, ya
17:06:27 <artem> ah, yes
17:06:33 <springmeyer> the sliver/mem problem
17:06:42 * springmeyer just looking through commits now...
17:06:57 <springmeyer> which certainly could screw up gdal plugin
17:07:05 <springmeyer> .startrss
17:07:06 <nikq> Okay, I'll re-start rss...
17:07:35 <StormTide> hrm, wtf, it works via qgis loading the same xml file, but not from mod_tile
17:07:35 <artem> there are some alignments fixes - it works great with raster.input but gdal.input needs some more attention
17:07:44 *** nikq (~nikq@li21-121.members.linode.com) has joined #mapnik
17:08:06 <springmeyer> artem: okay. yes and the gdal + map.buffer_size needs attention
17:08:11 <springmeyer> .startrss
17:08:11 <nikq> Okay, I'll start rss fetching...
17:08:13 <nikq> Mapnik Trac: Ticket #509 (mapnik::query resolution needs both x and y components) closed | http://trac.mapnik.org/ticket/509#comment:1
17:08:29 *** cgs_bob has quit (Ping timeout: 260 seconds)
17:09:12 * artem generating large amount of raster from srtm for testing
17:09:20 <springmeyer> awesome
17:09:45 <artem> any limitation on number of datasets in *.vrt ?
17:09:57 <artem> (gdal that is)
17:10:19 <springmeyer> not that I know of
17:10:38 <artem> ok, we'll find out :)
17:10:45 <springmeyer> #502 <-- bug in gdal when buffer_size > 0
17:10:45 <nikq> Ticket #502: Map buffer_size not accounted for within GDAL plugin scaling logic, http://trac.mapnik.org/ticket/502
17:11:12 <springmeyer> I figure we need to send the non buffered extent in query object, but I'm not sure yet
17:12:05 <artem> ok, I missed that one - should read it thourgh
17:13:58 <nikq> Mapnik Trac: Ticket #502 (Map buffer_size not accounted for within GDAL plugin scaling logic) updated | http://trac.mapnik.org/ticket/502#comment:1
17:20:33 *** twain47 has quit (*.net *.split)
17:22:16 *** twain47 (~twain47@cpc8-shef7-0-0-cust85.barn.cable.virginmedia.com) has joined #mapnik
17:27:41 <StormTide> springmeyer, any idea on why recompiling mapnik would nuke labels?
17:30:38 <springmeyer> StormTide: do they show up if you remove the regex stuff?
17:30:48 <StormTide> lemme try
17:30:48 <StormTide> sec
17:31:26 <Ldp__> springmeyer: I added the #485 change to the backwards incompatibility list for 0.7.0, but I notice it never made the Changelog too (see mapnik-users earlier today)
17:31:27 <nikq> Ticket #485: TextSymbolizer vertical_align should have varying default, http://trac.mapnik.org/ticket/485
17:31:55 <springmeyer> Ldp__: sorry about that, I will add it to the CHANGELOG in svn now...
17:31:58 <StormTide> nope
17:32:05 <StormTide> regexp makes no dif
17:32:35 <StormTide> whats weird is it is rendering in quantumnik, but not mod_tile now
17:33:05 <StormTide> not after the regexp... but before that even it was running in quantumnik and not mod_tile
17:33:13 <Ldp__> springmeyer: no worries. And jburgess spotted the difference in our NL osm.xml tiles already :)
17:33:35 <springmeyer> Ldp__: :)
17:33:36 <StormTide> feels like a font issue, but mod_tile says its loaded the font, and theres no missing font errors
17:33:40 <springmeyer> StormTide: have you restarted apache?
17:33:58 <StormTide> i restarted renderd
17:34:04 <StormTide> lemme try apache
17:34:36 <StormTide> restarting apache == no change
17:35:21 <StormTide> the test harness is http
17:35:28 <StormTide> http://dev.stormtide.ca/test.htm
17:35:51 <StormTide> the xml is mapnikpointswithtext.xml
17:36:12 <StormTide> the paths linked should all be available from the webserver too
17:36:18 <nikq> Mapnik Trac: Release0.7.0 edited | http://trac.mapnik.org/wiki/Release0.7.0?version=4
17:37:04 <StormTide> this was an instant failure after i reinstalled mapnik... lemme try reverting to the previous svn revision and see if rebuilding fixes er
17:37:39 <nikq> Mapnik Trac: Changeset [1616]: add missed vertical_alignment change in default behavior | http://trac.mapnik.org/changeset/1616
17:42:57 *** cgs_bob (~bob@115.sub-75-210-8.myvzw.com) has joined #mapnik
17:46:48 <StormTide> ok, confirmed, reverting svn fixes
17:47:07 <StormTide> -r 1599 tests as working
17:48:39 *** springmeyer has quit (Remote host closed the connection)
17:48:44 *** springmeyer (~springmey@c-76-28-156-154.hsd1.wa.comcast.net) has joined #mapnik
17:48:56 <StormTide> <StormTide> ok, confirmed, reverting svn fixes
17:48:56 <StormTide> <StormTide> -r 1599 tests as working
17:51:14 *** tcarobruce (~tcarobruc@adsl-75-10-247-30.dsl.pltn13.sbcglobal.net) has joined #mapnik
17:51:30 <springmeyer> StormTide: hrm, so svn up -r 1600 and it breaks?
17:51:42 <StormTide> to head it did.... trying to 1602 now
17:52:11 <StormTide> just waiting on the recompile
17:54:44 <StormTide> as of 1602, still working
17:55:06 <StormTide> which rev do you think might have broken it, i can try building that rev
17:58:47 <StormTide> working up to 1605..
17:58:55 <springmeyer> I'm not sure, but 1600 didn't look like the one...
18:00:00 <springmeyer> StormTide: and you are making sure to `install` each time you re-compile?
18:00:07 <StormTide> yes
18:00:10 <springmeyer> k
18:00:31 <StormTide> 1606 works
18:00:33 * springmeyer looking through commits post 1600...
18:00:35 <springmeyer> k
18:00:49 <StormTide> trying 07
18:01:15 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
18:01:33 <springmeyer> hmm, 07 could have done it
18:01:56 <StormTide> let you know just as soon as it finnishes building
18:02:18 *** mperry has quit (Ping timeout: 240 seconds)
18:02:18 *** mperry_ is now known as mperry
18:03:21 <nikq> Mapnik Trac: Changeset [1617]: add missed vertical_alignment change in default behavior and  ... | http://trac.mapnik.org/changeset/1617
18:04:21 <springmeyer> ya, its gotta be r1607
18:04:22 <nikq> http://trac.mapnik.org/changeset/1607, at , by dane: reenamble python access to TextSymbolizer 'name' expression
18:05:08 <StormTide> yes
18:05:11 <StormTide> confirmed
18:05:17 <StormTide> 1607 breaks
18:05:24 <springmeyer> okay, nice catch. looking into it...
18:07:19 <StormTide> just lemme know if you need me to try anything out
18:09:46 *** ajturner has quit (Quit: ajturner)
18:15:39 <springmeyer> hmm, artem - how should expression_ptr be returned in python? something other than `return_value_policy<copy_const_reference>())` ?
18:30:46 *** mperry_ (~mperry@c-67-164-175-65.hsd1.co.comcast.net) has joined #mapnik
18:33:25 *** mperry has quit (Ping timeout: 260 seconds)
18:33:25 *** mperry_ is now known as mperry
18:36:04 <nikq> Mapnik Trac: Changeset [1618]: don't pass expression_ptr as const | http://trac.mapnik.org/changeset/1618
18:36:18 <springmeyer> StormTide:  ^^
18:37:53 <StormTide> *tries*
18:43:45 <springmeyer> artem: do you think it would be possible to enable the use of " (double quotes) in expressions?
18:43:59 <springmeyer> currently it appears that the parser only accepts single quotes
18:44:29 <StormTide> springmeyer, segfault
18:44:40 <StormTide> renderd[16277]: DEBUG: Got command RenderPrio fd(6) xml(default), z(11), x(323), y(709)
18:44:41 <StormTide> Segmentation fault
18:44:44 <StormTide> from mod_tile
18:44:56 <StormTide> [1302816.491610] renderd[16281]: segfault at 11 ip b5c144a3 sp b3405530 error 4 in shape.input[b5bea000+3c000]
18:44:58 <StormTide> is the dmesg
18:45:08 <StormTide> i can load it in gdb if you need more info....
18:45:16 <springmeyer> right. you will likely need to re-compile mod_tile
18:45:22 <springmeyer> since a const changed
18:45:24 <StormTide> oh?
18:45:26 <StormTide> ok
18:45:27 <StormTide> sec
18:45:34 <springmeyer> can you first try to see if Qnik renders labels now?
18:46:01 <StormTide> qnik always did
18:46:11 <springmeyer> oh?
18:46:14 <StormTide> will just take me a sec to recompile mod_tile
18:46:18 <springmeyer> gah
18:46:36 <springmeyer> well then it was likely a need to recompile mod_tile all along...
18:47:16 <StormTide> i did mention that before.... <StormTide> whats weird is it is rendering in quantumnik, but not mod_tile now
18:47:23 <StormTide> anyway, appears to be working now
18:47:29 <StormTide> no segfault, and labels are showing
18:49:21 *** myselfhimself has quit (Ping timeout: 245 seconds)
18:52:38 <StormTide> kew, thanks, running head now
18:52:43 <StormTide> working like a charm
19:01:28 <nikq> Mapnik Trac: TextSymbolizer edited | http://trac.mapnik.org/wiki/TextSymbolizer?version=24
19:02:08 <nikq> Mapnik Trac: ShieldSymbolizer edited | http://trac.mapnik.org/wiki/ShieldSymbolizer?version=23
19:02:19 <springmeyer> StormTide: sorry missed that
19:02:24 <springmeyer> glad it is working now
19:02:50 <springmeyer> C++ is pretty unforgiving in terms of ABI changes, which that was
19:03:20 <springmeyer> the other thing that can trip you up when running a development version of mapnik
19:03:40 <springmeyer> is that plugins can get installed, then the ABI changes, and when they are not rebuilt you'll get segfaults
19:05:15 <StormTide> what does ABI mean
19:05:24 <springmeyer> Application Binary Interface
19:05:47 <StormTide> ah, for some reason i thought mod_tile was talking to renderd and mapnik via a socket
19:05:53 <StormTide> and they were all encapsulated
19:06:02 <StormTide> didnt realize they were that interdependent
19:06:12 * springmeyer not sure
19:06:21 <Ldp__> mod_tile talks to renderd through a socket. renderd uses the C++ bindings
19:06:45 <StormTide> that makes sense then
19:06:48 <artem> StormTide: mod_tile talks to renderd over unix sock
19:06:52 <Ldp__> always rebuild mod_tile as a whole when you change your mapnik version, just to be sure
19:06:57 <StormTide> i'll be sure to recompile renderd with mapnik upgrades in the future
19:07:59 <artem> Ldp__: one thing I notice with mod_tile that just 'make' doesn't rebuild everything -> make clean ; make
19:08:45 <Ldp__> artem: same here. It doesn't pick up on the changes mapnik includes.
19:08:49 <Ldp__> changed
19:09:00 <Ldp__> but why would it? They're not defined as a dependency in the Makefile
19:09:11 <artem> ok, make sense
19:09:35 <springmeyer> artem: is used to SCons :)
19:09:40 <springmeyer> me too
19:10:02 <Ldp__> artem: compile times for renderd are significantly higher with mapnik trunk, but that's no surprise
19:10:17 <StormTide> artem, i posted a message to the devel mailing list a couple days ago, i'm not sure it got through... regarding python expressions?
19:10:30 <Ldp__> springmeyer: does SCons check every include file, even those outside the project dir?
19:10:46 <StormTide> might be waiting on your moderation
19:10:50 <springmeyer> Ldp__: yes, unless you pass FAST=True
19:10:55 <artem> Ldp__: yep, we should improve compile time
19:11:19 <springmeyer> ya artem - recompiling locks up my machine and takes over 25 minutes
19:11:30 <Ldp__> artem: the memory usage for cc1 has the most impact for my server
19:11:57 <springmeyer> ideally we could do something pretty soon, as it makes development tricky - pesky boost::spirit ;)
19:12:05 <Ldp__> that and the heavy templating is what makes it crawl
19:12:10 <artem> Ldp__: new expression grammar can be re-factored into *cpp files
19:12:33 <artem> and into smaller chunks]
19:13:12 <springmeyer> Ldp__: I think it is __just__ the new boost spirit rather than anything specific to mapnik, but I've said that before...
19:13:29 <springmeyer> artem: let me know if I can help with this...
19:13:31 <Ldp__> springmeyer: yup, filter_factory and all that
19:13:35 <artem> StormTide: did you sign up ?
19:13:38 <springmeyer> yes
19:14:18 <springmeyer> Ldp__: btw FAST=True I think only makes the configure stage faster if you were wondering
19:14:30 <Ldp__> springmeyer: I wasn't, but understood what it did
19:14:33 <StormTide> artem, i did, andit was approved
19:14:34 <springmeyer> k
19:14:45 <StormTide> but im not sure if i sent the mail before i was approved tho
19:14:57 <artem> ok, I see let me check
19:15:01 <StormTide> think it went subscribe, send, approved
19:16:47 <artem> StormTide: There are no pending requests
19:16:56 <artem> hmm.. was it mapnik-devel ?
19:17:46 <artem> only one message from springmeyer this month on mapnik-devel
19:17:48 <StormTide> yes
19:18:01 <artem> why don't you post it again :)
19:18:04 <StormTide> sent to mapnik-devel@lists.berlios.de
19:20:34 <StormTide> anyway, was mostly intended for you anyway...
19:20:59 <StormTide> i'm wondering if its possible to tie the name='[expression]' into python's interpreters
19:21:11 <StormTide> so that we could use a lamba function or similar as an expression
19:21:33 <StormTide> (eg format data in any way python can process it)
19:22:00 <StormTide> name="([attribute]).python(\"'%.3f' % x\");" for example
19:22:06 <StormTide> or something along those lines
19:30:47 <artem> StormTide:  hmm.. interesting
19:31:04 <artem> expression is just a string before it parsed after that ..
19:31:26 <StormTide> ya
19:31:35 <StormTide> figure it could be put in the same as the regex engine really
19:32:06 <artem> right
19:32:42 <artem> c++ lib doesn't know anything about python atm
19:32:56 <StormTide> but since boost is already there
19:33:01 <StormTide> could just preprocessor check for it
19:33:07 <StormTide> and include if available
19:33:21 <StormTide> wouldnt have to change the optional deps, or the build process at all
19:33:33 <StormTide> that verbiage just wouldnt be available if you didnt build with boost
19:33:58 <artem> ok
19:34:22 <artem> StormTide: interesting idea, need to think it through
19:35:41 <artem> StormTide: why don't you create a ticket on trac.mapnik.org - this way we can keep track ..
19:37:16 <StormTide> i dont have a login...
19:37:35 <artem> you can create one, it's open to everyone
19:37:42 <StormTide> ah ok
19:37:43 <StormTide> sec
19:37:51 <artem> the only reason for login is to keep spammers away:)
19:38:49 <artem> springmeyer: thinking about spam - we're still getting a few on mapnik.org
19:39:33 <springmeyer> artem: ya, very few get through to being public but I need to also disable them getting saved (been lazy)
19:39:49 <springmeyer> the ones that get through as public I can handle deleting
19:40:18 <artem> no probs it's a big step forward from just publishing spam anyway :)
19:40:18 <nikq> Mapnik Trac: Ticket #510 (Python Interpreter as attribute expression.) created | http://trac.mapnik.org/ticket/510
19:40:25 <springmeyer> :)
19:40:30 <StormTide> there, created as 510
19:40:32 <artem> StormTide: excellent, thanks!
19:40:45 <springmeyer> ya, can't believe how lazy I am, I'll fix that up asap artem
19:41:09 * artem think he is even lazier 
19:41:12 <springmeyer> artem: we could also just disable comments in general?
19:41:14 <springmeyer> :)
19:41:51 <artem> springmeyer: perhaps optional comments?
19:42:02 <springmeyer> per post you mean?
19:42:19 <artem> yes, a checkbix - allow-comments ?
19:42:21 <springmeyer> I think that functionality is already in place (little radio button in admin)
19:42:23 <artem> checkbox
19:42:32 <artem> ha
19:42:36 <springmeyer> yep, I think we have it, but I'll have to check
19:42:52 <springmeyer> I think you have to hit the 'advanced' tab the the js opens it up
19:44:24 <artem> don't see advanced tab on http://mapnik.org/admin/news/entry/add/
19:45:25 * springmeyer looks
19:45:56 * springmeyer ah, making shit up again
19:46:07 <springmeyer> its in the flatpages I realize
19:46:22 <springmeyer> cool. I will look into adding to our "news" items...
19:50:23 <artem> awesome :D
19:50:47 * artem away for a bit 
19:55:39 <springmeyer> catch you later artem
20:15:34 <StormTide> springmeyer, is there any way to do a conditional statement in mapnik... eg if [someattribute] = whatever then write _____ to the map
20:16:22 <Ldp__> <Filter>[someattribute] = 'whatever'</Filter><TextSymbolizer name="literal string">
20:16:31 <StormTide> aha
20:16:33 <StormTide> excellent
20:16:34 <StormTide> thanks
20:16:41 <Ldp__> only in trunk!
20:16:59 <StormTide> works for me, im running the latest commit ;)
20:17:18 <StormTide> sounds like i tried to do this at the exact right point in this project ;)
20:17:45 <Ldp__> StormTide: you might want to read this (again): http://www.mapnik.org/news/2009/dec/08/future_mapnik2/
20:17:49 <springmeyer> StormTide: yep, right time, right place
20:18:08 <springmeyer> Ldp__: do you accept payment for your services?
20:18:09 <StormTide> Ldp__, thx ;)
20:18:16 <Ldp__> springmeyer: only in cookies
20:18:33 <springmeyer> hmm, duly noted
20:18:46 * StormTide can pay only in php books ;)
20:19:11 <Ldp__> springmeyer: should I? :)
20:19:29 <springmeyer> until your iPhone app hits number1 :)
20:19:53 <springmeyer> Ldp__: I think you should consider more than cookies, yes :)
20:20:04 <StormTide> springmeyer, do iphone apps make money these days?
20:20:17 * springmeyer chuckles
20:20:18 <StormTide> im doing this more for myself than to make money haha
20:22:49 <StormTide> springmeyer, where do you work anyway?
20:23:03 <springmeyer> cookie factory
20:23:35 * StormTide half wonders if thats a new web 2.0 startup
20:23:55 <springmeyer> mmmm, good idea :)
20:24:03 <Ldp__> springmeyer: I do accept hardware as well :)
20:24:14 <StormTide> cookiefactry.com or something haha
20:24:17 <springmeyer> first time I visited NYC I went to the "knitting factory"
20:24:26 <springmeyer> twas the hippest jazz club
20:24:51 <springmeyer> Ldp__: there you go.
20:25:48 <springmeyer> StormTide: I work for myself out of Seattle. http://dbsgeo.com
20:26:03 <springmeyer> I do a lot of django with a bit of mapnik on the side
20:26:08 <springmeyer> I need more cookies in my life
20:26:14 <StormTide> cool
20:26:26 <StormTide> im on vancouver island
20:26:30 <StormTide> just a stones throw away ;)
20:27:03 <springmeyer> yep yep !
20:27:18 * springmeyer it totally distracted dreaming about your backyard
20:27:20 <springmeyer> http://www.clearwind.ca/djangoski/
20:27:56 <springmeyer> StormTide: you in victoria?
20:28:04 <StormTide> springmeyer, close enough... duncan
20:28:12 <springmeyer> ah, nice
20:28:27 <StormTide> my boat lives in victoria tho ;)
20:28:43 <StormTide> (hence the map of oak bay ;])
20:28:50 <springmeyer> nice
20:29:11 <springmeyer> ya, I gotta hit you up for ideas when I get around to visiting salt spring
20:29:21 <springmeyer> want to do some bike touring there sometime
20:29:35 <StormTide> its absolutely beautiful there
20:30:00 <StormTide> get over there a couple times a year
20:30:06 <springmeyer> cool
20:30:35 <StormTide> sailed around the southern half this year, amazing coast line... if you could get a bike up there, you'd have views like no other
20:30:57 <springmeyer> ya, I bet, thats great
20:38:33 *** ajturner (~ajturner@209.155.228.129) has joined #mapnik
21:55:49 <StormTide> is it faster for mapnik to use one shapefile with multiple layers referenced by <layer> tags... or a series of small shape files with one layer each individually represented by <layer>
22:10:59 <springmeyer> depends on whether these layers have different extents
22:11:47 <StormTide> ok, so within an extent they're better as one?
22:12:08 <StormTide> that was what i was thinking anyway, instead of splitting this .000 into a bunch of shpfiles for each layer, create one
22:12:11 <StormTide> and reference it
22:12:25 <StormTide> by layer...  but since some are areas, some points and others linestring...
22:12:34 <StormTide> wasnt sure if if was better for them to all be separate
22:14:08 <springmeyer> well, its hard to say, it also depends on how your style filters are working
22:14:32 <springmeyer> there is a penalty to requiring mapnik to loop through and check the bounds of many layers
22:14:34 <StormTide> polygons, linestrings and textsymbolizers....
22:14:38 <springmeyer> just to end up not rendering them
22:14:59 <springmeyer> but there is an even greater penalty to pulling data from a shapefile
22:15:13 <springmeyer> just to have it discarded by the filters in the Style>Rules
22:16:13 <StormTide> well, as it goes right now its either one shapefile per .000 file (per chart extent).. or lots of shpfiles per .000
22:16:18 <StormTide> and theres about 1000 .000 files
22:17:26 <StormTide> in theory i could probably join all those shpfiles into one... but i suspect thats a really bad idea
22:17:35 <springmeyer> hmm, ya, requiring bounds checks on 1000 layers for ever map render could be bad
22:17:58 <springmeyer> so I would tend toward joining....
22:18:17 <springmeyer> and indexing
22:18:19 <StormTide> joining them all into one massive shpfile? or into one shape file per regional extent?
22:18:33 <StormTide> indexing resulted in a crash if i recall correctly
22:18:48 <springmeyer> that was a different problem
22:18:52 <springmeyer> I'd start with regional groups first
22:19:06 <StormTide> k
22:19:09 <Ldp__> I have a tiled shapefile dataset that I ended up joining into 1 big shapefile. It renders pretty fast, but it's mostly landuse (large polygons) and buildings
22:19:16 <springmeyer> StormTide: that crash resulted because an .index file was built for the shapefile
22:19:23 <springmeyer> but it was the 'ogr' plugin reading it
22:19:42 <springmeyer> which also can use .index files, but it needs to create them itself (with a different tool)
22:20:15 <springmeyer> so consider a bug of the ogr plugin, which I only had you using because the edgecase of shape plugin not correctly reading pointz shapefiles
22:20:18 <springmeyer> which is now fixed
22:20:25 <StormTide> ah
22:20:36 <StormTide> excellent
22:20:40 <StormTide> *tries to index*
22:21:04 <springmeyer> in simple words: don't use the 'ogr' plugin unless I tell you too again :)
22:21:23 <springmeyer> just don't index the point shapefiles!
22:21:31 <springmeyer> only lines and polygon ones
22:21:53 <Ldp__> springmeyer: don't let shapeindex grok point shapefiles then!
22:22:10 <artem> springmeyer: why not ?
22:22:43 <springmeyer> artem: bugs!
22:22:50 <StormTide> springmeyer, how would i, if joining, only index some
22:22:53 <artem> ok
22:22:58 <StormTide> or should i have different shp files for different types
22:23:07 <StormTide> eg regional area groups, regional point groups...
22:23:35 <springmeyer> artem: #504
22:23:35 <nikq> Ticket #504: Support for Pointzm broken in shapefile driver, http://trac.mapnik.org/ticket/504
22:24:01 <artem> StormTide: sounds like you might be better of using sqlite.input - you can have multiple tables in one file
22:24:06 <springmeyer> but even with r1605, only a few points are returned from an indexed pointz
22:24:07 <nikq> http://trac.mapnik.org/changeset/1605, at , by dane: improve shapeindex support for pointz
22:24:42 <artem> springmeyer: ok,thanks
22:24:44 <springmeyer> or of course postgis too StormTide
22:24:48 <StormTide> springmeyer, with the index on the point file, it seems to be working with my mapnik ver
22:25:03 * springmeyer can't recall why you didn't want to use a database - was there a reason?
22:25:46 <springmeyer> StormTide: can you paste the output of $ ogrinfo -so -al <your point.shp>
22:25:48 <StormTide> not really... other than it didnt seem to offer any advantage, and my final output format is for route-me and is just a tiles database
22:26:41 <springmeyer> the advantage is speed of rendering even with all data in one table, or a few large tables
22:27:04 <springmeyer> `OptimizeRenderingWithPostGIS
22:27:04 <nikq> http://trac.mapnik.org/wiki/OptimizeRenderingWithPostGIS
22:27:13 <springmeyer> relevant ^^
22:28:00 <StormTide> http://www.phpriot.com/2667
22:34:14 *** artem has quit (Quit: artem)
22:35:47 <springmeyer> hmm StormTide - that should not work right with an index based on my testing
22:36:01 <springmeyer> only about %% of the points render
22:36:04 <springmeyer> er 5%
22:36:26 <StormTide> thats odd
22:37:12 <StormTide> http://dev.stormtide.ca/test.htm seems to be working for me?
22:37:22 <StormTide> i didnt do anything with the xml file to make it use the index tho
22:37:31 <StormTide> i just created the .index file with shapeindex test3
22:38:28 *** artem (~artem@i-83-67-73-6.freedom2surf.net) has joined #mapnik
22:39:57 <StormTide> springmeyer, i guess i'll try the postgis route, but that seemed harder to me cuz i have to get the .000's into there, and then it'll be hard to regenerate partial updates
22:40:22 <nikq> Mapnik Trac: Ticket #504 (Support for Pointzm broken in shapefile driver) updated | http://trac.mapnik.org/ticket/504#comment:5
22:41:23 <nikq> Mapnik Trac: Ticket #504 (Support for Pointzm broken in shapefile driver) updated | http://trac.mapnik.org/ticket/504#comment:6
22:43:26 <nikq> Mapnik Trac: before_after_index_on_pointz.png attached to Ticket #504 | http://trac.mapnik.org/attachment/ticket/504/before_after_index_on_pointz.png
22:45:08 <springmeyer> oh crap
22:45:38 <springmeyer> Qnik is using the ogr driver
22:45:43 * springmeyer shoots himself in foot
22:46:40 <StormTide> ;)
22:46:57 <StormTide> if i try to load the indexed shp file in qnic i get /usr/bin/qgis.bin: symbol lookup error: /usr/local/lib/mapnik/input/ogr.input: undefined symbol: _ZN5boost9iostreams18mapped_file_source4initEv
22:47:01 <StormTide> as a crash
22:47:12 <StormTide> but that'd make sense if qnic was using ogr
22:47:54 * springmeyer had switched back to Mapnik 0.7.0, forgot
22:48:12 <springmeyer> if running 0.7.0 Qnik will use the ogr driver, as it should
22:48:27 <StormTide> im running trunk, and its still using ogr
22:48:45 <StormTide> tho i havent updated qnik from source lately
22:48:48 <springmeyer> hmm, you running latest Qnik?
22:49:06 <StormTide> its relatively recent
22:49:12 <StormTide> how do i check the ver
22:49:36 <springmeyer> cool, works
22:49:41 <springmeyer> sorry about the false alarm
22:50:11 <springmeyer> latest Mapnik + shapfile pointz + index works fine
22:50:21 <StormTide> k cool
22:50:27 <StormTide> so i'll just ignore this qnik crash then
22:50:39 <springmeyer> latest Qnik will use shp plugin if you are running Mapnik trunk
22:50:42 <springmeyer> ya
22:50:42 <StormTide> cuz i dont think it uses ogr when i load the xml separately
22:51:00 <springmeyer> right, Qnik does nothing fancy when loading from external xml
22:51:13 <springmeyer> but when creating a map on the fly from a QGIS layer it does
22:51:22 <StormTide> still, there does appear to be a crash from test3.shp when loaded into qgis then using create canvas
22:51:24 <springmeyer> try to sidestep mapnik bugs in the shapefile driver
22:51:39 <StormTide> but that wont get in my way
22:51:40 <springmeyer> ya, because you need to update Qnik
22:51:44 <StormTide> ah ok
22:52:58 <nikq> Mapnik Trac: Ticket #504 (Support for Pointzm broken in shapefile driver) updated | http://trac.mapnik.org/ticket/504#comment:7
22:53:12 *** artem has quit (Quit: artem)
22:53:59 <StormTide> there, works flawless
22:54:15 <StormTide> so no worry about merging the points into the regional file then, and i can still index.... score
22:54:29 <springmeyer> yep
22:54:43 <StormTide> will get this working, and if the shpfile stuff is too slow i'll try to setup postgis to handle it
22:54:46 <springmeyer> post r1605
22:54:46 <nikq> http://trac.mapnik.org/changeset/1605, at , by dane: improve shapeindex support for pointz
22:55:06 <StormTide> since its ultimately to get to generate_tiles.py, probably doesnt matter
23:22:45 <StormTide> ERROR 6: Can't create fields of type IntegerList on shapefile layers.
23:22:46 <StormTide>  <-- it seems to work, but it spits out those errors...  any idea what that means?
23:27:00 <springmeyer> #gdal <-- ask there
23:49:14 <StormTide> apparently they say you cant have shp files with multiple layers ~_0
23:49:19 <StormTide> so stuff that idea i guess
23:54:27 <springmeyer> well thats pretty much the case for all geodata formats as far as attributes go
23:54:52 <StormTide> i guess i just thought it was like the .000 with all the layers in one file
23:54:55 <StormTide> like a zipfile
23:55:08 <springmeyer> no, shapefiles don't have layers
23:55:22 <springmeyer> sorry, was referring to different limitation
23:57:19 *** jfreeman (~jfreeman@mail.agileware.net) has joined #mapnik