00:00:00 <cmarqu> Because it doesn't print *that* setting when starting up
00:00:09 <springmeyer> copy your openlayer.html and hg update will grab the revised config example
00:00:12 <cmarqu> I just saw watch_mapfile = False
00:01:57 <springmeyer> hmm, no it should be there
00:02:43 <cmarqu> Sorry, it is indeed. Just far away.
00:02:52 *** aub has quit ()
00:03:12 <springmeyer> ya, the sorting of the items gets lost since it is a dictionary
00:03:19 <springmeyer> not sure how to fix that...
00:03:54 <cmarqu> You could print them alphabetically and name the config items to that they have the same root for the same thing
00:04:06 <cmarqu> Which you have already
00:04:15 <springmeyer> hmm, yes
00:08:15 <cmarqu> Mapfile **changed**, reloading...
00:08:15 <cmarqu> Segmentation fault
00:08:44 <springmeyer> shoot. and that's with the latest code?
00:09:06 <cmarqu> I would claim so, but let me look again.
00:10:09 <springmeyer> do hg pull/update and then $ sudo python setup.py install
00:10:15 <cmarqu> pull/up didn't bring anything new, so I installed again.
00:10:54 <springmeyer> okay
00:11:07 <springmeyer> and hg st # does not reveal any locally changed files?
00:12:20 <cmarqu> Well, there is a diff in tilelite.py, but it's just
00:12:26 <cmarqu> -    doctest.testmod()
00:12:26 <cmarqu> \ No newline at end of file
00:12:26 <cmarqu> +    doctest.testmod()
00:12:47 <cmarqu> My emacs might have added that newline.
00:13:02 <springmeyer> well not a problem
00:13:07 <cmarqu> Other than that, only utils/openlayers.html is different
00:13:21 <springmeyer> you might manually clear out your python site packages and reinstall
00:13:26 <springmeyer> just to be safe
00:13:29 <cmarqu> And now I got
00:13:30 <cmarqu> Listening on port 8000...
00:13:30 <cmarqu> Mapfile **changed**, reloading...
00:13:30 <cmarqu> Failed to reload mapfile, will try again in 4 seconds
00:13:30 <cmarqu> Segmentation fault
00:13:34 <cmarqu> Ok.
00:13:46 <springmeyer> okay, nope that is the latest code
00:14:08 <cmarqu> The the clearing the packages won't help?
00:14:22 <springmeyer> (that 'will try again in 4 seconds'... means you've got the changeset that I thought would make the mapfile reloading threadsafe
00:14:35 <springmeyer> so, I've got to go back to that
00:14:39 <cmarqu> I see.
00:15:07 <cmarqu> Take it easy, I'll have to go to bed soon anyway and won't be around for the rest of the weekend
00:15:12 <springmeyer> I could not prompt that on mac osx, so perhaps something different on your linux version
00:15:24 <springmeyer> cmarqu: ya, I'm off in a bit to the pub
00:15:48 <springmeyer> cmarqu: can you tell me your python version, mapnik svn revision, boost version, and linux version?
00:16:13 <cmarqu> Python 2.5.2
00:17:00 <cmarqu> Mapnik is r973
00:17:04 <nikq> http://trac.mapnik.org/changeset/973, at , by dane: scons: move all configure options to 'scons configure' stage, using pickling to maintain configure options across builds
00:17:36 <cmarqu> Linux is 2.6.27-12-generic, x86_64
00:18:05 <springmeyer> okay, 64 bit - that may be related
00:18:12 <springmeyer> and boost?
00:19:00 <cmarqu> How do I find that out? Could it be 1.34.1?
00:19:36 <springmeyer> python scons/scons.py configure # in the mapnik source directory would do it
00:20:10 <cmarqu> Found boost lib version... 1_34_1
00:20:25 <springmeyer> cool, thanks
00:20:38 <cmarqu> Heck, I'll just update mapnik to latest. But that won'd finish compiling soon.
00:21:04 <cmarqu> Or do you prefer debugging without that?
00:21:09 <springmeyer> no, all that sounds good
00:21:34 <springmeyer> surely a tilelite problem, just helpful to know your differences for when I get to debugging further
00:22:21 <cmarqu> "Could not find required header or shared library for icudata" - do I want that?
00:22:38 <springmeyer> :)
00:22:43 <springmeyer> ya, required
00:22:59 <cmarqu> I see. Is is a new dependency?
00:23:06 <springmeyer> nope
00:23:16 <cmarqu> Also, these optional dependencies were not found: - boost system - ociei
00:23:23 <springmeyer> but its paths were not previously being checked properly
00:23:24 <cmarqu> Interesting.
00:23:55 <springmeyer> so if your mapnik build worked it was just because icudata happend to be linkable on the same path as another library
00:24:06 <springmeyer> (which is common on linux of course)
00:24:33 <springmeyer> yup, boost_system is not available until boost version > 1.35 and hence optional
00:24:56 <cmarqu> Ok.
00:24:57 <springmeyer> ociei is oracle lib, so no mortal will have that one :)
00:25:25 * cmarqu goes to install it just because
00:25:44 <springmeyer> you are not a mortal?
00:26:10 <cmarqu> I am still, but soon...
00:31:27 <cmarqu> Apparently, icudata should be in libicu-dev which I have installed.
00:33:36 <springmeyer> huh, so scons still can't find it?
00:34:04 <springmeyer> can you http://dpaste.com your 'config.log'?
00:34:10 <cmarqu> Hmm, I now deleted .sconf_temp/ and .sconsign.dblite, and it works.
00:34:22 <springmeyer> ya, that was my hunch :)
00:36:43 <cmarqu> Ok, bedtime. I'll return some day to ask how to delete stuff outside a bounding box in PostGIS, but not tonight.
00:37:00 <springmeyer> ST_Difference
00:37:03 <springmeyer> good night!
00:37:13 <cmarqu> Have fund in the pub. Night.
00:37:17 <cmarqu> -d
01:10:13 *** springmeyer has quit ()
05:42:14 *** springmeyer (n=dane@c-24-19-50-92.hsd1.wa.comcast.net) has joined #mapnik
12:55:50 *** aub (n=aubrey@cpe-72-227-134-148.nyc.res.rr.com) has joined #mapnik
16:41:37 <CIA-6> mapnik-utils: migurski * r590 /trunk/serverside/cascadenik/cascadenik/__init__.py: oops
17:00:38 <CIA-6> mapnik-utils: dane.springmeyer * r591 /trunk/serverside/cascadenik/cascadenik/compile.py: remove other reference to unused output import
17:10:43 *** scruggs__ (n=chris@75-121-84-235.dyn.centurytel.net) has joined #mapnik
17:23:07 *** scruggs has quit (Read error: 110 (Connection timed out))
17:23:26 *** scruggs (n=chris@75-121-84-235.dyn.centurytel.net) has joined #mapnik
17:23:36 *** scruggs_ has quit (Read error: 110 (Connection timed out))
18:30:13 <springmeyer> ping cmarqu
18:30:44 <springmeyer> cmarqu: I just tried to prompt a segfault on a couple machines, including 64 bit ubuntu 8.10
18:31:26 <springmeyer> no luck, even while running a python loop that updated the saved time on the mapfile each second while blasting TileLite with map requests
18:32:03 <springmeyer> the locking I added seems to work solidly to avoid allowing a render request to hit mapnik while the map is reloading
18:32:16 <springmeyer> (which is seemingly not working for you)
18:32:55 <springmeyer> I'm thinking that perhaps the raster layers you are using are either taking a lot longer to load, or for some other reason are messing up the thread lock
18:33:24 <springmeyer> so, cmarqu: could you wrap some of them up for me to download/try with your hike_n_bike styles?
19:38:31 *** aub has quit ()
20:56:37 *** audifahrer (n=andreas@p57AF675F.dip.t-dialin.net) has joined #mapnik
21:53:21 *** cmarqu_ (n=chatzill@p54B608C4.dip0.t-ipconnect.de) has joined #mapnik
22:49:39 <audifahrer> could one help me with a plugin development question?
23:29:19 <audifahrer> bye
23:29:37 *** audifahrer has quit (Remote closed the connection)
23:38:20 *** _closure has quit (Remote closed the connection)
23:38:23 *** _closure (n=E78C8EE4@li48-120.members.linode.com) has joined #mapnik