#gemrb@irc.freenode.net logs for 13 Mar 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[00:00:46] <gembot> build #301 of nmake-msvc++6 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B6/builds/301 blamelist: Brad Allred <bradallred@me.com>
[02:18:36] <-- kida has left IRC (Ping timeout: 250 seconds)
[03:26:49] <-- GoGi has left IRC (Excess Flood)
[03:28:06] --> GoGi has joined #gemrb
[05:52:00] <-- brada has left IRC (Quit: brada)
[08:02:47] --> lynxlynxlynx has joined #gemrb
[08:02:47] <-- lynxlynxlynx has left IRC (Changing host)
[08:02:47] --> lynxlynxlynx has joined #gemrb
[08:02:47] --- ChanServ gives channel operator status to lynxlynxlynx
[08:10:26] --> WingedHussar has joined #gemrb
[08:33:48] <-- WingedHussar has left IRC (Quit: WingedHussar)
[08:34:29] --> WingedHussar has joined #gemrb
[09:16:33] <-- Coriander has left IRC (Quit: Nettalk6 - www.ntalk.de)
[09:17:30] --> Coriander has joined #gemrb
[09:31:56] <-- DrMcCoy has left IRC (Disconnected by services)
[09:32:06] --> DrMcCoy has joined #gemrb
[10:42:51] <-- lynxlynxlynx has left IRC (Ping timeout: 260 seconds)
[10:58:22] --> lynxlynxlynx has joined #gemrb
[10:58:22] <-- lynxlynxlynx has left IRC (Changing host)
[10:58:22] --> lynxlynxlynx has joined #gemrb
[10:58:23] --- ChanServ gives channel operator status to lynxlynxlynx
[13:08:02] --> kida has joined #gemrb
[13:53:26] --> brada has joined #gemrb
[13:53:57] <brada> psch: could you show me the code you are using to get the paths for script/override/etc?
[14:00:37] <psch> the path i write to the config?
[14:00:55] <psch> that's Context.getExternalFilesDir(null) in the java activity
[14:01:10] <psch> http://nopaste.info/00606e1a2e.html line 30 & 31
[14:01:13] <Seniorita> Nopaste - powered by project-mindstorm IT Services
[14:01:42] <brada> no i mean during core initialization
[14:02:09] <psch> im writing SDL_AndroidGetExternalFilesDir() into $HOME before core initialization
[14:02:18] <psch> which is where it gets the path for the .cfg from
[14:02:25] <psch> the cfg has paths for the folders
[14:02:28] <brada> thats right
[14:02:36] <brada> i remember you saying that now
[14:03:46] <psch> btw, i ordered an OTG cable, so i should be able to test mouse events vs touch events soonish
[14:04:02] <psch> at least if mice work with OTG cables
[14:04:18] <psch> which i assume they should
[14:04:54] <brada> we probably need to apply some transform to mouse coordinates as well
[14:37:35] <lynxlynxlynx> interesting, we got saving trouble reports from android too
[14:51:16] <brada> very odd
[14:51:23] <brada> you have an assert there even
[14:54:34] <lynxlynxlynx> so the lookup succeeds, just the values get trashed somewhere
[14:54:46] <lynxlynxlynx> can you try if it still works for you on mac?
[14:54:57] <lynxlynxlynx> i've tried before your changes and it was fine
[14:58:37] <brada> sure
[15:01:01] <brada> works fine for me
[15:02:04] --> i30817 has joined #gemrb
[15:02:18] <brada> lynx: is it only android users?
[15:09:27] <gembot> build #321 of osx-xcode-binary is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/321 blamelist: Brad Allred <bradallred@me.com>
[15:10:57] <gembot> build #321 of osx-cmake is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/osx-cmake/builds/321 blamelist: Brad Allred <bradallred@me.com>
[15:11:23] <brada> sourceforge suddenly down?
[15:12:29] <brada> out of space?!
[15:23:41] <lynxlynxlynx> also windows
[15:24:31] <lynxlynxlynx> i'll go delete some old builds
[15:25:15] <lynxlynxlynx> woah, 2,4G each
[15:25:21] <lynxlynxlynx> better do it via ssh
[15:25:48] <brada> it was my buildbot that was out of space
[15:25:55] <brada> but yeah probably not a bad idea
[15:48:51] <gembot> build #328 of mingw32 is complete: Exception [6exception upload] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/328 blamelist: Brad Allred <bradallred@me.com>
[15:51:35] <lynxlynxlynx> cleanup complete
[15:51:53] <lynxlynxlynx> was deleting while this was uploading, so the failure is expected
[16:04:41] <-- WingedHussar has left IRC (Remote host closed the connection)
[16:09:46] <gembot> build #302 of nmake-msvc++6 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B6/builds/302
[16:27:33] --> Avenger has joined #gemrb
[16:27:33] --- ChanServ gives channel operator status to Avenger
[16:32:15] <-- Avenger has left IRC (Client Quit)
[16:33:28] <-- i30817 has left IRC (Remote host closed the connection)
[17:04:51] <brada> does this look sane to everyone?
[17:04:52] <brada> http://paste.debian.net/241500/
[17:04:57] <Seniorita> debian Pastezone
[17:08:07] <brada> psch: will that work for your android build?
[17:08:36] <psch> i don't know
[17:08:47] <psch> i don't yet understand it tbh
[17:12:22] <brada> hmm. i should probably move the DATADIR cases into that function
[17:16:07] <brada> there
[17:16:08] <brada> http://paste.debian.net/241504/
[17:16:11] <Seniorita> debian Pastezone
[17:20:49] <brada> psch: maybe you should use a special env variable instead of $HOME
[17:20:56] <brada> like $GEM_DATA
[17:21:01] <brada> or similar
[17:21:13] <psch> yeah, i can use whatever, meaningful of course preferred
[17:21:17] <brada> yes
[17:21:21] <brada> jsut let me know
[17:21:30] <psch> additionally, not using $HOME also let's us not have to look in .GemRB, right?
[17:21:42] <brada> yeah that patch fixes that
[17:21:45] <psch> which i think is a bit not-so-pretty, seeing as the whole subdir is already application specific
[17:21:53] <brada> yes
[17:22:07] <psch> ill have to adjust the activity then
[17:22:12] <brada> sure
[17:22:15] <psch> i'm fine with GEM_DATA
[17:22:22] <brada> ok
[17:22:25] <brada> thats what ill use
[17:27:22] <brada> updated with GEM_DATA env var
[17:27:23] <brada> http://paste.debian.net/241505/
[17:27:27] <Seniorita> debian Pastezone
[17:29:06] <brada> psch: go ahead and apply that and let me know what happens
[17:29:14] <psch> alright
[17:34:30] <psch> missing semicolon on line 208 in the diff, line 478 in VFS.cpp
[17:34:38] <psch> 477
[17:34:50] <brada> thanks :D
[17:35:06] <psch> heh, i did need the compiler to tell me ;)
[17:44:13] <psch> loads fine without explicit paths in the cfg
[17:45:11] <psch> which is good since it means i can remove the config-file-editing from the activity
[17:45:47] <psch> which is actually at least part of why you rewrote LoadConfig, so i'd say it was successful
[17:47:46] <psch> im still a bit sad i can't easily add $GEM_DATA to the env on the java side, but well
[17:48:15] <psch> reflecting into a method and adding code doesn't seem sane, so it stay in GemRB.cpp for now :/
[17:56:01] <psch> yeah, in any case, it works
[17:57:13] <brada> cool
[17:57:52] <brada> config editing capabilities would still be handy
[17:57:55] <brada> at least for now
[17:58:15] <brada> still working on a nice patch to get configuration from other sources
[17:58:24] <brada> such as gui ;)
[17:59:07] <psch> of course
[17:59:19] <psch> but what i did in the activity was just a simple replace if i found a placeholder
[17:59:38] <psch> seeing as override etc now look in the right path anyway, that's not needed anymore
[18:01:31] <psch> im thinking though, the whole GUI thing should rather be done inside gemrb like is written on gsoc_ideas, shouldn't it?
[18:01:59] <psch> no idea how practical it is, but the idea is already there
[18:02:16] <psch> compared to writing a configurator in java exclusively for android, it seems smarter
[18:04:14] <brada> the problem with that is gemrb requires guiscripts and such to have a gui like that
[18:04:32] <brada> which is fine
[18:04:48] <brada> but i dont see it happening anytime soon
[18:05:37] <brada> and its not like throwing together a config gui takes more than an afternoon :p
[18:06:16] <brada> and most of the patch series im working on is stuff that would need to be done either way
[18:06:21] <brada> or should be done anywya
[18:06:57] <brada> and i dont see a CHU gui being able to easily open a file browser to locate the game dir etc
[18:07:12] <brada> it would take quite some work i imagine
[18:07:49] <psch> i guess firing a file browser intent to chose gamepath is something i could add
[18:09:20] <brada> also people seem to mis the pelya interface so reimplementing the nice features from that would be something not doable in CHU
[18:10:03] <brada> beback in a bit. if nobody has objected by then ill push that commit
[18:10:05] <-- brada has left IRC (Quit: brada)
[18:29:39] <lynxlynxlynx> well, if the commit message is true, then it is false :)
[18:33:13] <lynxlynxlynx> the code looks better
[18:37:37] --> brada has joined #gemrb
[18:40:01] <-- kida has left IRC (Ping timeout: 256 seconds)
[18:48:47] <brada> psch: what else needed to be fixed for android?
[18:50:24] <psch> what do you mean fixed?
[18:50:30] <brada> i mean patched i guess
[18:50:33] <psch> http://nopaste.info/0b9fdf4df0.html
[18:50:34] <Seniorita> Nopaste - powered by project-mindstorm IT Services
[18:50:42] <psch> is the only change i have from master, excluding the android subdir
[18:51:03] <brada> SDL/SDL.h vs SDL.h is really more an environment thing
[18:51:41] <psch> $ find SDL/ -iname 'SDL.h'
[18:51:42] <psch> SDL/include/SDL.h
[18:51:43] <brada> your search path should point to a directory called SDL that contains the headers
[18:51:57] <psch> that's how it comes from upstream
[18:51:59] <brada> eyah thats from the source dir
[18:52:07] <brada> that wouldnt be the same if you had installed libraries
[18:52:28] <brada> when you make install sdl it pust them in an SDL directory
[18:52:30] <psch> i have sdl installed
[18:52:43] <psch> but i can guarantee that's sdl hg
[18:53:21] <psch> i'm not saying you're wrong, mind, i'm just saying what i needed to change
[18:53:23] <brada> you should probably symlink it
[18:53:33] <brada> if thats the way its going to be built
[18:54:04] <psch> *i can't guarantee
[18:54:29] <psch> i don't really understand what you think i should do to be honest
[18:54:40] <brada> wasnt there an android specific cpp filethat calls main?
[18:54:53] <brada> if so that setenv should probably go there
[18:54:56] <psch> yeah, SDL_android_main.cpp
[18:55:30] <brada> you dont need to check HAVE_SETENV if you know its there
[18:55:56] <psch> i didn't know it'd be there
[18:56:01] <brada> btw what doesnt have setenv?
[18:56:10] <brada> ?
[18:56:14] <-- Drakkar has left IRC (Ping timeout: 245 seconds)
[18:56:21] <brada> isnt setevn part of libc?
[18:56:28] <psch> i don't know
[18:56:32] <psch> probably?
[18:56:46] <psch> i haven't touched the file for a few days at least
[18:56:54] <brada> stdlib.h
[18:57:09] <psch> back when i tried it thought it smart-ish to put it inside the HAVE_SETENV and didn't move it
[18:57:24] <psch> but i can see why you would put it into SDL_android_main.cpp
[18:57:28] --> Yoshimo has joined #gemrb
[18:57:34] <psch> the setenv() call that is
[18:57:38] <brada> yes
[18:57:52] <brada> id rather all the android specific code from that file go there if it can
[18:58:10] <psch> it'd get a bit harder to keep that in sync with what happens with sdl2 though
[18:58:14] <psch> if we keep those seperate
[18:58:21] <psch> like with the constants for the pixel formats
[18:58:33] <brada> i doubt that file will ever change
[18:58:38] <brada> i really wouldnt worry
[18:58:48] <brada> and if it did it would be irrelevant to you
[18:59:15] <brada> if you did your own gui interface especially
[19:00:11] <psch> for now there's not much i'd do actually, just let the user input gamepath, if possible with a browser, and write that to gemrb.cfg
[19:00:13] <brada> what im trying to say is that file is just a helper and is in no way required for an sdl app to work :)
[19:00:44] <psch> i'd just stack stuff around what happens - i don't really want to mess with the JNI bindings
[19:01:21] <brada> in fact you may as well copy that file and rename it
[19:01:40] <brada> jsut to show that it is customized
[19:04:53] --> Drakkar has joined #gemrb
[19:05:53] <psch> i don't get it
[19:06:58] <psch> oh
[19:07:22] <psch> this is about "just put stuff we did something with into the repo" in the end, isn't it?
[19:07:34] <brada> pretty much yes
[19:07:40] <psch> i haven't even started pulling my build env apart to be honest
[19:07:50] <psch> i only have the prep_env.sh + prepared files in now
[19:07:58] <psch> because i figured that should be documented somewhere
[19:08:08] <psch> but then i haven't pushed it anywhere, so it's still only on my pc
[19:08:34] <psch> yeah i'll get to that, i'm not sure how it make sense though
[19:08:41] <psch> i mean, the idea itsself is obviously fine
[19:08:54] <psch> but i'm not sure how far i can/should pull the SDL project template apart
[19:09:21] <brada> i believe we are looking for the makefile and co inside an "android" directory at gemrb root
[19:09:38] <brada> that should be able to run without neeeding to move them
[19:09:56] <brada> along with a readme
[19:10:25] <brada> then add whatever customized android code into the same directory
[19:10:57] <brada> so some java activity stuff i assume
[19:11:06] <brada> and a manifest file i guess
[19:11:16] <brada> no idea what else
[19:11:47] <brada> the important thing is that everything needed to build aside from dependancies is in the repo
[19:12:08] <psch> from my experience, env vars inside Android.mk files don't really work well, as in they break the script-supplied vars
[19:12:17] <psch> stuff like $(LOCAL_PATH) e.g.
[19:12:33] <psch> which doesn't make sense to me, but it happens
[19:12:56] <brada> the android directory is in a known location in the repo so you can use ../../whatever
[19:12:57] <psch> idk, the ndk build process seems really fragile to me
[19:13:04] <psch> yeah, but dependencies
[19:13:13] <psch> i remember you saying something about referring to them with env vars
[19:13:24] <brada> thats what im doing with ios build
[19:13:31] <brada> do whatever works i suppose
[19:13:43] <brada> im hardly a good person to ask regarding android development
[19:13:56] <psch> well, what currently works is building the build environment before actually building :/
[19:14:10] <psch> i'll have to put some more time into cleaning this up i guess
[19:14:40] <brada> you can check if some variables are set and if not run the setup script i guess
[19:15:13] <brada> just wrap everything in a shellscript
[19:16:04] <brada> id just like a way to direct the build to an existing set of precompiled libs
[19:16:18] --> vampi-the-frog has joined #gemrb
[19:16:22] <brada> and headers
[19:22:12] <psch> yeah, makes sense
[19:23:35] <vampi-the-frog> I don't get it
[19:23:48] <vampi-the-frog> I find garbled data in the containers section in the ARE files inside SAV files
[19:24:16] <vampi-the-frog> it's basically containers with bogus names (31 chars), with proper world positions, but with 0 items
[19:24:39] <vampi-the-frog> InfExp also displays them with garbled characters, it's the files themselves that contain the weird data
[19:25:08] <lynxlynxlynx> what's the real problem?
[19:25:34] <vampi-the-frog> I'm trying to find ingress' teeth (PST), by writing some code that searches in the save files :)
[19:25:56] <vampi-the-frog> and my fear is that PST garbled the save files so the teeth are lost forever
[19:26:09] <lynxlynxlynx> why not just summon them?
[19:26:15] <vampi-the-frog> PST doesn't have a lua console
[19:26:23] <lynxlynxlynx> gemrb does
[19:26:26] <vampi-the-frog> hrm
[19:26:31] <vampi-the-frog> PS:T is extremely buggy in gemrb
[19:27:02] <lynxlynxlynx> anyway, if they're not part of some random treasure or on an actor (slight chance of moving between areas), their location should be fixed
[19:27:06] <lynxlynxlynx> and mentioned in guides
[19:27:13] <vampi-the-frog> no no no
[19:27:15] <vampi-the-frog> I got them
[19:27:22] <vampi-the-frog> but then the character carrying them died
[19:27:34] <vampi-the-frog> and lost them in a pile somewhere
[19:27:44] <lynxlynxlynx> piles expire
[19:27:59] <vampi-the-frog> still, they should still be stored, with an expired timestamp, right?
[19:28:10] <lynxlynxlynx> not sure
[19:28:37] <vampi-the-frog> now that I think about it though, why would the character lose a weapon if they're wearing it?
[19:28:39] <lynxlynxlynx> waste of space to keep destroyed piles around
[19:28:46] <-- Cable_ has left IRC (Ping timeout: 245 seconds)
[19:28:55] <lynxlynxlynx> game critical items are immune
[19:29:02] <vampi-the-frog> yeah I noticed that...
[19:29:09] <lynxlynxlynx> maybe magic items too, but i think that's it
[19:29:22] <vampi-the-frog> hrm
[19:30:33] <brada> who cares if PST is buggy in gemrb
[19:30:33] <vampi-the-frog> I think I'll start a new game with the original engine, and try to generate one of those bogus containers
[19:30:40] <brada> use gemrb to create the item and save
[19:30:41] <vampi-the-frog> brada: people playing it?
[19:30:44] <brada> then load in PST
[19:30:50] <vampi-the-frog> yes I was thinking the same thing
[19:31:08] <vampi-the-frog> and I'll resort to that if I can't figure it out by grepping the files
[19:31:30] <vampi-the-frog> I'm writing my own code to load chitin.key, TLK files, SAV files, ARE files and so on
[19:31:31] <brada> sounds a lot more painful/difficult then jsut doing it in gemrb :D
[19:31:39] <vampi-the-frog> it's more fun for me
[19:31:58] <fuzzie> you learn fairly quickly in the IE games that you have to pick up all the items from a dead char before leaving
[19:32:22] <fuzzie> very annoying when you run out of inv space for things you consider vital :P
[19:33:20] <brada> i learned to just reload when a char dies :D
[19:35:10] <lynxlynxlynx> i can't stand it too, everyone must be alive
[19:35:34] <lynxlynxlynx> endgame being the only exception
[19:42:20] --> Cable_ has joined #gemrb
[19:48:03] <-- vampi-the-frog has left IRC (Remote host closed the connection)
[20:46:10] <-- Yoshimo has left IRC (Quit: Yoshimo)
[21:18:33] --> thomcom has joined #gemrb
[21:24:43] <brada> http://paste.debian.net/241559/
[21:24:46] <Seniorita> debian Pastezone
[21:24:51] <brada> fuzzie: any comments?
[21:24:55] <brada> lynx?
[21:25:20] <brada> it can wait till after release since it doesnt do much until i have time to implement gui interfaces
[21:25:35] <brada> and id hate to break something so close to release :p
[21:25:43] <brada> if i havent already ;)
[21:29:54] <brada> i jsut now realized im calling it INIConfig when it is not an INI file...
[21:31:40] <lynxlynxlynx> ATTEMPT_INIT could easily be a function
[21:32:42] <lynxlynxlynx> i expected more code to die, but the idea looks ok
[21:36:53] <-- thomcom has left IRC (Quit: Page closed)
[21:50:47] <-- brada has left IRC (Quit: brada)
[22:33:30] --> brada has joined #gemrb
[22:36:46] --- ermo^ is now known as ermo
[23:23:24] --- ermo is now known as ermo^
[23:37:55] --> i30817 has joined #gemrb
[23:46:30] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:59:08] --> i30817_ has joined #gemrb