Config
Log for #openttd.dev on 9th October 2012:
Times are UTC Toggle Colours
02:34:39  *** Terkhen has quit IRC
02:34:39  *** DorpsGek has quit IRC
02:35:23  *** Terkhen has joined #openttd.dev
02:35:23  *** DorpsGek has joined #openttd.dev
02:35:23  *** kilo.oftc.net sets mode: +vov Terkhen DorpsGek DorpsGek
09:14:03  *** Webster has joined #openttd.dev
09:14:04  *** ChanServ sets mode: +v Webster
10:55:21  *** Guest1119 is now known as planetmaker
11:32:07  *** Supercheese has quit IRC
11:32:38  *** Supercheese has joined #openttd.dev
13:36:00  <Belugas> hello
14:34:55  <Belugas> is it me or there is no HowTo compile openttd for android from windows?  Tried to look all around and nly see references to linux
14:36:49  *** DorpsGek sets mode: +v planetmaker
14:37:19  * planetmaker knows of no such howto. But I only ever looked for osx and linux
14:42:06  <Terkhen> I did not know that howtos for compiling OpenTTD for android existed
14:43:15  <Terkhen> btw, the guy who did the android port posted at the forums recently WRT his source code, in case you missed it
14:43:35  <Belugas> pelya?
14:43:43  <Belugas> he shown me this page: https://github.com/pelya/commandergenius/tree/sdl_android/project/jni/application/openttd
14:44:10  <Belugas> the readme.txt is not good for me
14:44:12  <Belugas> sadly
14:48:40  <Terkhen> I see
15:44:28  <Belugas> i hope i'm not the only one who tries to compile it in win...
17:19:25  *** Alberth has joined #openttd.dev
17:19:25  *** ChanServ sets mode: +v Alberth
17:45:09  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r24578 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
17:48:56  *** FLHerne has joined #openttd.dev
17:58:32  *** Zuu has joined #openttd.dev
17:58:32  *** ChanServ sets mode: +v Zuu
18:03:21  *** fonsinchen has joined #openttd.dev
18:04:58  *** frosch123 has joined #openttd.dev
18:04:58  *** ChanServ sets mode: +v frosch123
19:03:05  <Belugas> downloading android SDK and NDK
19:05:27  *** Yexo has left #openttd.dev
19:05:27  <planetmaker> http://bugs.openttd.org/task/5322 makes sense to me. Anyone mind?
19:05:29  *** Yexo has joined #openttd.dev
19:05:30  *** ChanServ sets mode: +v Yexo
19:05:34  <planetmaker> http://bugs.openttd.org/task/5322 makes sense to me. Anyone mind?
19:05:43  <Yexo> Belugas: I've been trying to compile it for android (on linux), but so far no luck
19:06:13  <Yexo> there are two ways to compile it: using the ndk framework (with it's makefiles and everything) or using a custom (openttd's) makefile
19:06:23  <Yexo> the second method might not work at all, but it's definitely the way I'd prefer
19:06:43  <Yexo> planetmaker: as I said yesterday: fine :)
19:06:58  <planetmaker> hm, sorry, missed that completely then, Yexo :-)
19:08:03  <Yexo> other channel :)
19:08:05  <Belugas> yexo, method two means gcc or something?
19:08:28  <Yexo> Belugas: a cross-targeted gcc which is part of the android ndk
19:08:52  <Belugas> haa...  looks like a learning curve for the whale
19:09:48  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r24579 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
19:13:33  <planetmaker> Terkhen you can now translate two strings ;-)
19:15:32  <Yexo> ./configure --host=arm-linux-androideabi --endian=LE --without-libpng --without-liblzma --without-freetype --without-liblzo2 --without-zlib --without-libicu CFLAGS="--sysroot $NDK/platforms/android-9/arch-arm" <- Belugas that was my last try
19:15:47  <Yexo> still nowhere near successful
19:19:06  <planetmaker> http://bugs.openttd.org/task/5308/getfile/8640/FS#5308.patch <-- this patch also makes sense to me
19:21:32  <Yexo> see other channel, someone said something about a memory leak
19:21:39  <Yexo> "first compare, than strdup"
19:22:00  <frosch123> also coding style
19:22:05  <frosch123> blanks and such
19:22:50  <planetmaker> yes, coding style needs improvement, sure
19:23:01  <Yexo> the idea of the path/ticket is good for sure
19:25:12  <planetmaker> and yes, I didn't look at .tgp :-) thanks for hint
19:25:41  <frosch123> yeah, we still mess up the channels :)
19:26:31  <planetmaker> did you do anything about the patch or do you plan to comment and wait for improvement?
19:26:59  <frosch123> i am always unsure about such cases :)
19:27:14  <frosch123> i could fix the two lines, and commit
19:27:19  <Yexo> why not do that?
19:27:28  <frosch123> but then the author might not notice/learn anything
19:27:41  <Yexo> post whatever you fixed in the fs issue, I don't see a problem
19:30:59  <planetmaker> it teaches more if he finds out himself about the leak. But... I'm with Yexo there, I guess
19:35:03  <Belugas> thanks Yexo.  preciously saved :)
19:39:25  *** fonsinchen has quit IRC
20:03:20  <planetmaker> Alberth, oh, you our comment correctness police ;-) : how should multi-line doxygen be aligned? with the @? one letter further? with the first of the description of that entry? We use all of it
20:04:35  <Alberth> personally, I prefer to align the descriptions, as it makes them easier to find then
20:05:02  <Alberth> but many devs don't follow that idea :)
20:05:48  <Terkhen> planetmaker: thanks, I'll translate them now
20:06:10  <Terkhen> wow, those were complicated
20:06:11  <Alberth> it is not that bad, I usually read the generated html anyway
20:06:18  <planetmaker> yup, told you, Terkhen ;-)
20:06:37  <Alberth> it took a whole 22 seconds? :)
20:06:50  <Terkhen> I'm watching TV too :P
20:07:00  <Alberth> with multi-tasking :p
20:08:06  <planetmaker> http://hg.openttd.org/openttd/trunk.hg/file/6e7d952dc44e/src/script/api/script_engine.hpp#l112 <-- so like this, Alberth ?
20:09:49  <Alberth> oh, I am thinking about the wrong part of the texts :(
20:10:00  <planetmaker> :-)
20:10:44  <Alberth> I always align everything under each other with the introducory text
20:11:12  <Alberth> I think the @ is clear enough to find the first non-explanation line
20:11:48  <Alberth> I am not sure whther the @ lines also support multi-line
20:12:10  <planetmaker> http://paste.openttdcoop.org/show/1768/
20:12:20  <planetmaker> they seem to do. at least we use it
20:12:26  <Alberth> you should try to build the docs, and check (or find an existing case, and check the already built docs)
20:12:28  <planetmaker> or it's all broken
20:12:49  <planetmaker> let's just check api docs as they are :-)
20:14:20  <Alberth> the docs can be quite broken, I have had a round of adding % characters already to prevent doxygen from making it a link in the @file entries
20:14:21  <planetmaker> http://noai.openttd.org/api/trunk/classAIEngine.html#c98e6ab7a63bdb52d226f47ca29e7b8c works
20:14:32  <planetmaker> the note is two-line note in source
20:14:37  <Alberth> k
20:15:47  <Alberth> I like your change, it makes very clear the @return has more lines of text
20:16:01  <planetmaker> ok :-) Then I'll go for that style
20:16:49  * Alberth gives a spare space-bar to planetmaker
20:17:20  <planetmaker> :)
20:17:28  <Alberth> although writing a script may be better here
20:18:04  <planetmaker> The Spacebar has room# 518 ;-)
20:18:28  <planetmaker> I didn't intend to change all. But I was looking at FS5318
20:18:43  <planetmaker> And then I can do it right at least in those cases
20:19:32  <planetmaker> indeed getting that uniform might be a script task
20:19:51  <Alberth> that's also my strategy, just fix the things you run into
20:20:38  <Alberth> Hmm, given that room# 518 has 'About 54,800,000 results', it must be something special, but I am missing it completely :p
20:21:39  <planetmaker> :-) It's the community room's name in my institute
20:23:01  <Alberth> it sounds unlikely to be the sole cause of 54 million hits
20:23:27  <planetmaker> :-)
20:23:30  <frosch123> planetmaker: so, did you do something with fs#5308?
20:23:53  <planetmaker> nope. I thought you had fixed it (locally)?
20:24:06  <frosch123> well, then i am doing that now :p
20:24:12  <planetmaker> :D
20:24:39  <Alberth> planetmaker: I am somewhat pondering about making a devzone project for cl.py (the improved language file analyzer), but it sounds a bit overkill for a single file
20:25:46  <frosch123> oh, now that i look at that diff more closely
20:25:50  <planetmaker> well, if you like you can :-) 'd consider that better be part of NML or of the newgrf framework
20:25:56  <frosch123> it has a weird usage of iterator
20:28:19  <Alberth> planetmaker: an other direction is to extend to a newgrf translation service, but that is a bit premature :)
20:28:45  <planetmaker> both can be done. Do you feel comfortable to replace http://dev.openttdcoop.org/projects/make-nml-common/repository/changes/check_language ?
20:28:57  <planetmaker> then I just give you commit rights and it's yours to toy with
20:29:06  <planetmaker> I don't mind that bash file being gone
20:29:41  <planetmaker> I think it's terkhen's "hacky language checker". But my memory might be faulty
20:30:48  <planetmaker> oh, rather... nml_langcheck.py maybe. That's already from you, I recon
20:30:52  <frosch123> http://paste.openttdcoop.org/show/1769/ <- that's what remains
20:31:33  <planetmaker> hm :-) Definitely an improvement, frosch123
20:32:10  <planetmaker> http://dev.openttdcoop.org/projects/make-nml-common/repository/entry/nml_langcheck.py <-- Alberth
20:32:48  <Yexo> contains |= strcmp(*iter, ip) == 0; and in the for && !contains
20:33:12  <Yexo> but fine like this, easier to read :)
20:33:43  <Alberth> planetmaker: yep, that looks like mine :p
20:33:44  <Alberth> I had a talk with Ammler a few days ago, and he liked to have an additional 'index.html' file generated with an overview of the state of all languages + a link to the 'real' file
20:34:38  <planetmaker> fine with me
20:35:02  <planetmaker> maybe by means of extra parameter (not) generated. Not sure whether it's needed, though
20:36:09  <Alberth> I added a '--index' this morning, it just doesn't do anything yet :p
20:36:32  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r24580 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
20:37:21  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r24581 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
20:38:46  <frosch123> planetmaker: are you sure about that?
20:38:55  <frosch123> iirc it was different for ships and aircraft
20:39:34  <frosch123> while aiengine and aiorder functions would match each other for the vehicle types
20:40:26  <Zuu> Yep, from my readup of the implementation, it differs depending on transport mode.
20:40:41  <frosch123> i think we already denied such a documentation before
20:40:50  <frosch123> michi_cc had a strong opinion
20:40:52  <planetmaker> hmpf, drat
20:42:23  <frosch123> yeah, ships are manhattan
20:42:50  <Zuu> The old docs encourage AIs/GSs to ues the OrderDistance API for those cases and not mix with distances obtained through DistanceManhattan/DistanceSquared etc.
20:42:54  <frosch123> so, the comment should be like "the unit returned by GetOrderDistance/GetMaximumOrderDistance match each other, but the unit itself is unspecified"
20:43:51  <planetmaker> you're right :-(
20:43:55  <frosch123> iirc michi_cc reserved to change the units at some point
20:44:14  <planetmaker> still, the API should say what kind of distance it is, not?
20:44:27  <frosch123> no, the idea was to explcitly not specify it
20:44:41  <frosch123> it should only only be used to compare the results of those 3 functions
20:44:42  <frosch123> nothing else
20:44:52  <frosch123> noone should rely on the units
20:45:07  <planetmaker> oh, I see.
20:45:14  <Alberth> good night
20:45:18  <planetmaker> night, Alberth
20:45:31  *** Alberth has left #openttd.dev
20:46:39  <Zuu> For practical cases it can help when on a strategic level make computations and measurements of ideal transport distances etc. to know the limits without having a real case, but that can be worked around by an AI/GS by picking two tiles at some square/manhattan distance at the map and obtain the order distance to feed into this strategic model.
20:47:06  <planetmaker> Hm... so simply revert?
20:48:48  <Zuu> a revert and/or a note about the fact that the unit of the return value is unspecified and possible even a subject for change in the future.
20:52:10  <planetmaker> http://paste.openttdcoop.org/show/1770/ like this?
20:53:06  <frosch123> distanc/distance
20:53:32  <planetmaker> 	 * @note   The unit of the order distances is unspecified and should not
20:53:32  <planetmaker>  maybe even
20:53:40  <planetmaker> then it's clearer that order distances != map distances
20:54:09  <Zuu> Yep, that is a good improvement
20:59:30  *** DorpsGek changes topic to "OpenTTD Dev Channel || Latest SVN: r24582 || Logs: http://webster.openttdcoop.org/?channel=openttd.dev || Voice (talk-right) upon request via #openttd; make sure you are registered to NickServ before asking"
21:02:24  <Belugas> mmh... that patch has some big problems.. the way its made, it assumes no integration in trunk
21:02:48  <planetmaker> it doesn't assume that. yes
21:03:12  <Belugas> but sometimes, it checks for ifdef ANDROID, sometimes not
21:03:20  <Belugas> not consistent at all!
21:03:32  <planetmaker> is it needed in more cases?
21:03:51  <Belugas> dunno yet, i'm just reading the patch
21:04:04  <Belugas> like removing stdafx.h
21:04:11  <Belugas> for everyone
21:04:13  <planetmaker> my impression was that they are in the right places. But ^^ is pointless
21:04:25  <planetmaker> shouldn't matter anywhere, should it?
21:04:32  <planetmaker> did you read my comment on the patch?
21:04:42  <Belugas> on forums?
21:04:45  <planetmaker> yes
21:04:54  <Belugas> mmh.. can't remember if i did
21:04:55  <Belugas> i'
21:05:01  <Belugas> ll checck that tonight
21:05:01  <Yexo> http://paste.openttdcoop.org/show/1771/ <- I can get the sources to compile with this, next step is linking
21:05:09  <Yexo> which probably means building SDL first
21:05:30  <planetmaker> http://www.tt-forums.net/viewtopic.php?p=1048932#p1048932 @ Belugas
21:06:46  <Belugas> thanks to you both :)
21:07:06  <planetmaker> Yexo, is there no android-sdl available?
21:07:21  <planetmaker> as I understand the android guy he has a custom sdl there, too...
21:07:34  <Belugas> looks like that to me as well
21:09:21  <frosch123> night
21:09:25  *** frosch123 has quit IRC
21:11:00  <Belugas> haaa... ok... i read it
21:16:57  <Yexo> planetmaker: the android guy has a custom sdl in his repo but afaik no sources for it
21:17:11  <Terkhen> planetmaker: I hope that my hacky script is a standard for anything :P
21:17:14  <Yexo> I think he hacks around it by using system-installed headers for SDL with his own library
21:17:26  <Yexo> which is not a good solution
21:20:48  <Belugas> home he goes, night
21:21:58  <planetmaker> Yexo, doesn't he ave the same kind of "source" as for openttd?
21:23:09  <Yexo> night Belugas
21:24:16  <planetmaker> yes... I find there .c and .h files. though no patch(es)
21:24:23  <Yexo> planetmaker: I can cross-compile sdl without problems, now it's just a matter of fixing a correct sdl-config
21:24:44  <planetmaker> :-)
21:25:12  <planetmaker> you guys should document how to get to a working android dev environment :-)
21:25:22  <Yexo> I'll do that as soon as I have one
21:25:33  <planetmaker> :-)
21:25:50  <Yexo> still in the progress of creating exactly that
21:27:06  <Terkhen> good night :)
22:14:34  <michi_cc> planetmaker: The idea of not having units for those functions was for two reasons. Firstly, we might one day have it for rail/RVs as well (with whatever units) and secondly especially the current crude distance for aircraft might get improved.
22:23:35  *** FLHerne has left #openttd.dev
22:31:59  <Yexo> planetmaker: seems like the complete source of SDL is in that android repo
22:32:49  <Yexo> I've managed to link everything together, so I now have libOpenTTD.so
22:33:04  *** FLHerne has joined #openttd.dev
22:33:07  <Yexo> tomorrow trying to get the java code to work to get a working application
22:33:22  * Yexo is happy with the progress :)
22:33:24  <Yexo> good night all
22:48:20  <planetmaker> good night here, too
23:48:18  *** Zuu has quit IRC

Powered by YARRSTE version: svn-trunk