Config
Log for #openttdcoop.devzone on 6th June 2011:
Times are UTC Toggle Colours
07:06:03  *** ODM has joined #openttdcoop.devzone
08:45:32  *** andythenorth has joined #openttdcoop.devzone
08:45:46  *** andythenorth has left #openttdcoop.devzone
11:56:20  <planetmaker> http://developer.gimp.org/api/2.0/libgimp/libgimp-gimpdrawable.html#gimp-drawable-get-name <- hm... maybe one could access image layers also by name. That might be much more convenient than the layer number...
11:56:21  <Webster> Title: gimpdrawable (at developer.gimp.org)
11:57:26  <Terkhen> indeed
11:57:34  <Yexo> planetmaker: as alternative: PIL can read psd files and also read the layer names
11:57:58  <planetmaker> can it also read xcf?
11:58:19  <planetmaker> (i.e. gimp)
11:58:44  <planetmaker> I prefer one script for graphics handling, if possible
11:59:03  <planetmaker> PIL actually might be the better choice than gimp...
11:59:07  <planetmaker> we use python anyway
11:59:22  <Yexo> not sure if xcf is supported
12:00:42  <planetmaker> hm... doesn't seem like
12:00:44  <planetmaker> I wonder why
12:01:06  <Terkhen> grep -Ri PSD in the source/docs gives a lot of results, XCF none (except a few hex numbers)
12:01:07  <Yexo> strange indeed
12:01:30  <Terkhen> but gimp can open psd, right? we could move to psd from xcf
12:01:40  <planetmaker> gimp can do that, yes
12:02:09  <planetmaker> Terkhen, I guess the source file is whatever the artist uses
12:02:25  <planetmaker> And I'd not bet on photoshop being capable to import xcf files ;-)
12:02:29  <Terkhen> I'd prefer to not use psd, but not using something as heavy as gimp for the layered image -> png conversion would be great
12:02:52  <planetmaker> yes and yes :-)
12:03:01  <Yexo> xfctools seems to be perfect: http://henning.makholm.net/software
12:03:02  <Webster> Title: Henning Makholm's software (at henning.makholm.net)
12:03:05  <Yexo> never tried it myself though
12:03:37  <Terkhen> The tools are designed to allow efficient use of layered XCF files as sources in a build system that use 'make' and similar tools to manage automatic processing of the graphics. <--- that's us :D
12:03:51  <planetmaker> quite. sounds interesting
12:09:56  <Terkhen> compilation fails on mingw, though
12:10:36  <Terkhen> and it does not seem to have support for compiling on MSVC
12:12:01  <Terkhen> let's see if I can hack it to compile on windows, otherwise it is not a good option :/
12:12:04  <planetmaker> compiles smoothly on suse
12:12:25  <planetmaker> well, it does not handle psd as my current script does
12:12:35  <Terkhen> hm, true
12:12:42  <planetmaker> and is not supported anymore
12:12:48  <planetmaker> Hm...
12:12:52  <Terkhen> not a great option then
12:13:04  <Terkhen> we need something light that supports psd, xcf and layers by name
12:13:18  <Yexo> if this doesn't work I doubt there is anything better
12:13:23  <planetmaker> that'd be ideal. I wonder whether one can add xcf support to PIL
12:13:34  <Yexo> I was thinking of that too
12:14:09  <Terkhen> PIL isn't updated very frequently either
12:15:31  <planetmaker> hm... http://www.google.de/url?sa=t&source=web&cd=5&ved=0CDkQFjAE&url=http%3A%2F%2Fblenderartists.org%2Fforum%2Fshowthread.php%3F114107-import-Gimp-image-(xcf)&rct=j&q=python%20xcf%20read%20support%20gimp&ei=tsTsTajVOYrMtAa1mKDnCg&usg=AFQjCNFin8n0ss5AdoW-_gggd7iJui8x9A&sig2=UAyb53bW-OkEqtcHIi956A&cad=rja
12:15:32  <Webster> Title: Hinweis weiterleiten (at www.google.de)
12:15:37  <planetmaker> ups
12:16:02  <planetmaker> nvm that
12:16:20  <planetmaker> it falls back to xcftools :-P
12:16:54  <Terkhen> :(
12:17:52  <Terkhen> PIL identifies and reads PSD files written by Adobe Photoshop 2.5 and 3.0. <--- isn't that ancient?
12:18:07  <planetmaker> yes
12:18:11  <planetmaker> current is 5.5
12:18:25  <planetmaker> maybe higher... CS is 5.5
12:18:25  <Yexo> I was able to read a file george uploaded yesterday
12:18:39  <planetmaker> but... ^ all of andy's files and georges can be read
12:19:16  <planetmaker> at least with gimp ;-)
12:21:42  <Terkhen> it seems that we are stuck with gimp :P
12:23:03  <planetmaker> it's darn heavy. But has the advantage to be both, scriptable and cross-platform
12:26:49  <Terkhen> it increases compilation time a lot
12:27:03  <Terkhen> I think that something more light wouldn't take that much time
12:27:33  <planetmaker> yes, agreed
12:27:49  <planetmaker> though image processing is never a light-weight task
12:28:28  <Yexo> the data format seems pretty well documented (by the author of xcftools), if that's not out of date it isn't too hard to write a PIL plugin to read xcf files
12:28:55  <planetmaker> his documentation was merged to the gimp docs. It's part of their documentation - and meanwhile a bit updated
12:29:02  <planetmaker> (he writes it on his page somewhere)
12:29:15  <Yexo> see http://svn.gnome.org/viewvc/gimp/trunk/devel-docs/xcf.txt?view=markup
12:29:16  <Webster> Title: [gimp] Annotate of /trunk/devel-docs/xcf.txt (at svn.gnome.org)
12:29:28  <Yexo> the only update was s/The GIMP/GIMP/ in 2007
12:29:34  <Yexo> which is 4 years ago now
12:29:34  <planetmaker> :-D ok
12:32:57  <Terkhen> is there any newgrf that uses persistent storage for airports?
12:33:05  <Yexo> no
12:34:09  <Ammler> Yexo: maybe they switched to git?
12:34:38  <Terkhen> ok, I guess I'll test industries first then :P
12:35:44  <Yexo> Ammler: ok, there was a single fix in 2009 in git :)
12:35:46  <Yexo> still not really updated
12:36:22  <planetmaker> it's not like openttd's savegame format need frequent updates ;-)
12:36:51  <planetmaker> or nfo specs ;-)
12:37:02  <Yexo> but they have both been updated in the last 4 years
12:37:14  <Yexo> it's not just the format, it also describes the data it contains
12:37:32  <Yexo> so to compare: every time a new setting is added to openttd, or a new property or callback to the nfo specs
12:37:40  <Ammler> well, that just means, the format spec is pretty stable, doesn't?
12:37:47  <planetmaker> hm, I rather meant the container format
12:37:51  <planetmaker> v7
12:37:59  <Yexo> that doesn't change much indeed :)
12:38:05  <planetmaker> additional stuff is also stored in new xcf formats
12:38:13  <planetmaker> but the format itself doesn't need changing
12:38:26  <planetmaker> nor did openttd need that in the last 4 years afaik
12:38:35  <Yexo> only this text document seems to be the documentation for both the container and the contents
12:38:47  <Yexo> and neither part has been updated for 4 years
12:38:52  <planetmaker> hm....
12:39:33  <planetmaker> probably the actual content is documented as in our case: in the source ;-)
12:39:42  <Yexo> yes
13:29:43  <Rubidium> planetmaker: the container format is still the same as for nfo v1
14:52:08  *** Lakie has joined #openttdcoop.devzone
16:36:49  *** ODM has quit IRC
16:40:40  *** frosch123 has joined #openttdcoop.devzone
17:10:20  <Brot6> nml: update from r1371 to r1374 done - http://bundles.openttdcoop.org/nml/nightlies/r1374
17:17:25  <Brot6> Following repos didn't need a nightlies update: 2cctrainset (r750), 32bpp-extra (r40), ai-admiralai (r75), ai-aroai (r39), ailib-common (r21), ailib-direction (r17), ailib-list (r32), ailib-string (r29), ailib-tile (r16), airportsplus (r93), basecosts (r25), belarusiantowns (r8), bros (r52), chips (r141), comic-houses (r71), firs (r1996), fish (r653), frenchtowns (r6), german-townnames (r34), grfcodec (r830), grfpack (r279), heqs
17:17:25  <Brot6> (r605), indonesiantowns (r41), manindu (r7), metrotrackset (r56), narvs (r37), newgrf_makefile (ERROR r293), nml (r1374), nutracks (r201), ogfx-industries (r119), ogfx-landscape (r69), ogfx-rv (r107), ogfx-trains (r241), ogfx-trees (r49), opengfx (r673), openmsx (r97), opensfx (r97), smts (r19), snowlinemod (r49), spanishtowns (r10), sub-landscape (ERROR r66), sub-opengfx (ERROR r666), swedishrails (r202), swisstowns (r22),
17:17:27  <Brot6> transrapidtrackset (r15), ttdviewer (r34), ttrs (r36), worldairlinersset (r672)
17:20:21  <Brot6> newgrf_makefile: compile of r293 still failed (#2656) - http://bundles.openttdcoop.org/newgrf_makefile/nightlies/ERROR/r293
17:23:29  <Brot6> sub-landscape: compile of r66 still failed (#2616) - http://bundles.openttdcoop.org/sub-landscape/nightlies/ERROR/r66
17:24:15  <Brot6> sub-opengfx: compile of r666 still failed (#2586) - http://bundles.openttdcoop.org/sub-opengfx/nightlies/ERROR/r666
17:25:05  <Brot6> Following repos rebuilds successful without any difference to earlier nightlies builds: airportsplus (Diffsize: 12), belarusiantowns (Diffsize: 30), frenchtowns, german-townnames, indonesiantowns (1 errors) (Diffsize: 1), manindu (Diffsize: 1), narvs, ogfx-industries, ogfx-landscape (Diffsize: 21), ogfx-rv, ogfx-trains (Diffsize: 813), spanishtowns (Diffsize: 1), swedishrails (Diffsize: 385), swisstowns
18:22:37  *** andythenorth has joined #openttdcoop.devzone
18:37:14  <Ammler> andythenorth: vm working for your tests?
18:37:24  <andythenorth> will try shortly
18:37:25  <andythenorth> thanks
18:37:28  <andythenorth> :)
19:35:52  *** ODM has joined #openttdcoop.devzone
19:38:18  <andythenorth> Ammler: I need to setup my key somehow?
19:38:27  <andythenorth> I am lacking knowledge about ssh
19:38:38  <andythenorth> I add it to a hosts file or some such?
19:39:34  <Ammler> I have already added your key
19:39:55  <andythenorth> Permission denied (publickey).
19:40:09  <andythenorth> I need to specify the port when I ssh?
19:40:13  <Ammler> yes
19:40:20  <Ammler> 12122
19:40:43  <andythenorth> hmm
19:41:22  <Ammler> ssh -p 12122 root@andy.openttdcoop.org
19:41:36  <andythenorth> I'm in
19:42:07  <andythenorth> Ammler: there's some package manager I need to use to install python 2.6?
19:42:22  <Ammler> python is already installed
19:42:55  <Ammler> zypper in <package>
19:45:26  * andythenorth disdains scp in favour of an sftp client
19:45:28  <andythenorth> scp sucks
19:45:43  <Ammler> it supports also sftp
19:46:34  <Ammler> just use sftp://andy.openttdcoop.org:12122 as address
19:48:01  <andythenorth> I have to figure how to teach my gui client that
19:48:08  <andythenorth> it wants passwords and such
19:48:21  <Ammler> username root
19:48:33  <andythenorth> no password needed - I'm in
19:48:37  <andythenorth> it found my key I guess
19:48:41  <Ammler> yep
19:49:13  <andythenorth> where should I put stuff I want to serve publicly?
19:49:15  <andythenorth> var?
19:49:16  <andythenorth> usr?
19:49:24  <Ammler> I would create a user
19:49:41  <Ammler> shall I setup uwsgi for you?
19:49:49  <andythenorth> not needed yet
19:49:52  <andythenorth> but soon
19:49:59  <Ammler> what you mean with publicy then?
19:49:59  <andythenorth> I prefer one step at a time :)
19:50:10  <andythenorth> ah
19:50:16  <andythenorth> I see the problem
19:50:28  <andythenorth> we need a proxy / web front end?
19:50:40  <Ammler> I would recommend nginx
19:50:52  <Ammler> I have no clue how to run pyhton with apache
19:51:07  <andythenorth> default pyramid can serve
19:51:14  <Ammler> ah ok
19:51:15  <andythenorth> but will I be able to get to its port?
19:51:19  <andythenorth> http://andy.openttdcoop.org:some_port
19:51:19  <Ammler> that might be fine for test
19:51:33  <Ammler> which port does it use per default?
19:51:53  <Ammler> currently 12180 will be forwarded
19:51:57  <Ammler> to 80
19:52:14  <andythenorth> 6543 is default
19:52:17  <andythenorth> it's trivial to change
19:52:23  <Ammler> so I forward that
19:53:33  <Ammler> I created a user andy
19:53:51  <Ammler> I would switch to that user for the pyramid
19:53:56  <Ammler> su - andy
19:55:32  <andythenorth> where should I sftp my stuff to first?  /usr/local/??
19:55:46  <andythenorth> my gui client sftp is root
19:56:14  <Ammler> ok, port 6543 goes to your vm
19:56:35  <Ammler> andythenorth: again, I would use user andy
19:56:35  <andythenorth> cool
19:56:45  <andythenorth> so I need to add a key for that user?
19:56:49  <andythenorth> or how?
19:57:47  <Ammler> copied key
19:58:00  <andythenorth> I'm in thanks
19:58:03  <Ammler> so you use same address, just user andy instead root
19:58:19  <Ammler> then copy your stuff to that userdir
19:58:47  <andythenorth> so the path is /home/andy
19:58:48  <andythenorth> ok
19:58:49  <Ammler> as the port is >1024, you can run pyramid as user
20:01:01  <andythenorth> lets see if the buildout works :)
20:01:23  <Ammler> just tell me, if there is any python module missing
20:02:05  * andythenorth watches buildout :P
20:03:55  <andythenorth> Ammler: if you're interested, pyramid uses a paster script to serve for development purposes.
20:04:04  <andythenorth> I'm not sure quite how it works, but it's using wsgi
20:04:04  <andythenorth> http://en.wikipedia.org/wiki/Python_Paste
20:04:05  <Webster> Title: Python Paste - Wikipedia, the free encyclopedia (at en.wikipedia.org)
20:04:44  <Ammler> andythenorth: yep, that is ok, wsgi is used in many python apps
20:04:49  <Ammler> e.g. also for hgweb
20:04:52  <andythenorth> yes
20:04:54  <andythenorth> it's cunning
20:05:12  <andythenorth> you can use wsgi to mix e.g. django, pyramid and zope in one app
20:05:12  <Ammler> that is why we use nginx/uwsgi
20:05:21  <Ammler> yep
20:05:34  <andythenorth> we could actually migrate bananas one method at a time if inclined
20:05:39  <andythenorth> although I'm not so inclined :P
20:06:20  <andythenorth> buildout finished o
20:06:21  <andythenorth> k
20:06:33  <andythenorth> can't run tests, the test runner is not configured correctly by me
20:07:49  <Ammler> anything I can help?
20:08:15  <andythenorth> not with tests, that is something I will fix later
20:08:21  <andythenorth> but meanwhile
20:08:23  <andythenorth> the buildout failed
20:08:34  <andythenorth> it needs to compile some c libraries
20:08:38  <andythenorth> and misses gcc
20:08:57  <andythenorth> can we get gcc?
20:09:45  <Ammler> of course
20:10:27  <Ammler> just run zypper in gcc"
20:10:30  <Ammler> done
20:10:53  <andythenorth> hmm
20:10:59  <andythenorth> No update candidate for 'gcc-4.5-19.1.i586'. The highest available version is already installed.
20:11:20  <Ammler> yep, "done" means i have done it :-P
20:11:25  <andythenorth> ok :)
20:11:40  * andythenorth runs buildout again
20:14:43  <andythenorth> meh
20:14:55  <Ammler> :-)
20:15:18  <andythenorth> markupsafe/_speedups.c:12:20: fatal error: Python.h: No such file or directory
20:15:18  <andythenorth> compilation terminated.
20:15:33  <Rubidium> python-dev (or similar) missing?
20:15:39  <Ammler> python-devel installed :-)
20:15:56  <andythenorth> my local buildout is borked for different reasons
20:16:22  <andythenorth> my laptop ran out of battery yesterday :P
20:18:03  <Rubidium> pff... it can't cope with that?
20:19:09  <planetmaker> @base 16 10 53f
20:19:09  <Webster> planetmaker: 1343
20:24:43  <andythenorth> meh
20:24:53  <andythenorth> my local issues were probably upstream dependencies
20:25:17  <andythenorth> one of the problems with buildout is that if someone checks in broken code to py-pi, buildouts fail...
20:25:22  <andythenorth> the solution is version pins
20:25:34  <andythenorth> but that's beyond my knowledge right now
20:27:48  <andythenorth> bah
20:27:52  <andythenorth> ok my code is broken :P
20:29:11  * andythenorth will fix
20:30:13  <andythenorth> this is now an exercise in debugging pyramid
20:30:46  <Ammler> but the pyramid is working?
20:34:14  <andythenorth> yes
20:34:17  <andythenorth> my app isn't
20:34:32  <andythenorth> I've been using hg purge a lot
20:34:39  <andythenorth> which is bad if you forget to commit your src
20:43:37  <Ammler> andythenorth: now I would use hg there with a update hook
20:44:13  <andythenorth> in this specific case I needed to add some things to hgignore
20:44:23  <andythenorth> specifically the stuff fetched by buildout into the working directory
20:44:36  <andythenorth> e.g. /parts, /eggs, /downloads
21:01:37  *** frosch123 has quit IRC
21:04:32  *** andythenorth is now known as Guest3660
21:04:32  *** andythenorth has joined #openttdcoop.devzone
21:09:54  *** Guest3660 has quit IRC
21:18:53  <andythenorth> Ammler: http://andy.openttdcoop.org:6543/
21:18:59  <andythenorth> it's not impressive yet :P
21:19:11  <andythenorth> I spent most of my time fixing some local screwups, not working with the vm
21:19:45  <andythenorth> but it works
21:20:00  <andythenorth> it's running in foreground, when I exit I think it will stop
21:20:37  <Ammler> add "&" at the end
21:20:40  <Ammler> or use screen
21:20:41  * andythenorth experiments
21:20:57  <andythenorth> it's in development configuration, which is dangerous to run on web
21:21:08  <Ammler> if you are interested on the output, screen
21:21:11  <Ammler> else &
21:21:12  <andythenorth> it has more debugging on, and arbitrary code can be executed
21:21:57  <andythenorth> Ammler: works with &
21:22:01  <andythenorth> dunno how I stop it now :P
21:22:11  <andythenorth> probably there's a paster stop
21:23:53  <andythenorth> hmm
21:23:56  <andythenorth> dunno how to stop it
21:23:58  <andythenorth> but anyways
21:24:05  <andythenorth> there is a --daemon option in paster
21:24:08  <andythenorth> for backgrounding
21:24:13  <andythenorth> it can also restart etc
21:24:25  <andythenorth> but if it was in production, it should use nginx or apache or scuh
21:24:32  <andythenorth> and probably supervisord or similar
21:25:11  * andythenorth -> bed
21:25:14  <andythenorth> Ammler: thanks for that
21:25:41  <Ammler> http://andy.openttdcoop.org/ <-- setup proxy done
21:25:45  <Terkhen> good night andythenorth
21:25:56  <andythenorth> good night
21:25:59  *** andythenorth has quit IRC
21:26:00  <Ammler> andythenorth: how did you start?
21:26:04  <Ammler> oh well :-)
21:26:11  <Ammler> killall <whatever>
21:27:02  <Terkhen> :P
21:34:48  *** ODM has quit IRC
22:19:20  *** OwenS has quit IRC
22:20:03  *** OwenS has joined #openttdcoop.devzone

Powered by YARRSTE version: svn-trunk