#gemrb@irc.freenode.net logs for 3 Mar 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:08:25] <-- Maighstir has left IRC (Quit: ~ Trillian Astra - www.trillian.im ~)
[00:19:13] --- pupni-af-k is now known as pupnik
[03:37:23] <-- pupnik has left IRC (Quit: leaving)
[05:55:36] <xrogaan> hey
[05:56:02] <xrogaan> this page needs an update
[05:56:03] <xrogaan> http://gemrb.sourceforge.net/wiki/doku.php?id=getting_started
[05:56:17] <xrogaan> specially the part talking about autogen
[07:05:16] <-- tomprince has left IRC (Ping timeout: 246 seconds)
[07:19:36] --> Bo_Thomsen has joined #GemRb
[07:32:12] --> tomprince has joined #GemRb
[08:07:08] <-- Bo_Thomsen has left IRC (Quit: Leaving.)
[08:30:51] --> adominguez has joined #GemRb
[08:32:10] --> edheldil_ has joined #GemRb
[08:33:31] --> lynxlynxlynx has joined #GemRb
[08:33:31] <-- lynxlynxlynx has left IRC (Changing host)
[08:33:31] --> lynxlynxlynx has joined #GemRb
[08:33:31] --- ChanServ gives channel operator status to lynxlynxlynx
[09:21:38] --> lubos has joined #GemRb
[09:39:36] <edheldil> wjp: yesterday I have noticed that for a couple of minutes the wiki was unresponsive
[09:56:05] <-- |Cable| has left IRC (Remote host closed the connection)
[12:13:53] --> lynxlynxlynx_ has joined #GemRb
[12:14:35] --> lynxlynx has joined #GemRb
[12:14:35] --- ChanServ gives channel operator status to lynxlynx
[12:15:12] --> Maighstir has joined #GemRb
[12:15:57] <-- lynxlynxlynx has left IRC (Ping timeout: 240 seconds)
[12:19:39] --- lynxlynxlynx_ is now known as lynxlynxlynx
[12:28:11] <-- Maighstir has left IRC (Quit: ~ Trillian Astra - www.trillian.im ~)
[12:31:49] --> Maighstir has joined #GemRb
[13:12:10] <-- DrMcCoy has left IRC (Quit: Reconnecting)
[13:12:27] --> DrMcCoy has joined #GemRb
[16:01:19] --> _dev_urandom has joined #GemRb
[16:02:43] <-- devurandom has left IRC (Ping timeout: 260 seconds)
[16:49:20] --> lynxlynxvargrrr has joined #GemRb
[16:49:20] <-- lynxlynx has left IRC (Read error: Connection reset by peer)
[16:53:17] <-- lubos has left IRC (Quit: Leaving.)
[17:21:28] --> Bo_Thomsen has joined #GemRb
[17:22:13] <-- edheldil_ has left IRC (Ping timeout: 272 seconds)
[17:36:49] --> lynxlynx has joined #GemRb
[17:36:49] <-- lynxlynx has left IRC (Changing host)
[17:36:49] --> lynxlynx has joined #GemRb
[17:36:49] --- ChanServ gives channel operator status to lynxlynx
[17:37:23] <-- lynxlynxvargrrr has left IRC (Write error: Connection reset by peer)
[17:44:39] --> DocMcCoy has joined #GemRb
[17:45:19] <-- DrMcCoy has left IRC (Quit: Reconnecting)
[17:45:35] --- DocMcCoy is now known as DrMcCoy
[17:47:26] <-- DrMcCoy has left IRC (Changing host)
[17:47:26] --> DrMcCoy has joined #GemRb
[17:49:42] <-- Bo_Thomsen has left IRC (*.net *.split)
[17:51:06] --> Bo_Thomsen has joined #GemRb
[17:57:42] <-- Bo_Thomsen has left IRC (*.net *.split)
[17:58:22] --> Bo_Thomsen has joined #GemRb
[18:00:05] <-- lynxlynx has left IRC (Ping timeout: 276 seconds)
[18:01:54] --> PixelScum has joined #GemRb
[18:04:44] <-- Drakkar has left IRC (Ping timeout: 240 seconds)
[18:07:10] <-- lynxlynxlynx has left IRC (Read error: Connection reset by peer)
[18:07:12] --> lynxlynxlynx_ has joined #GemRb
[18:11:33] <-- adominguez has left IRC (Quit: Saliendo)
[18:14:51] --> Maighstir_ has joined #GemRb
[18:15:38] <-- Maighstir has left IRC (Read error: Connection reset by peer)
[18:22:43] <-- Bo_Thomsen has left IRC (Ping timeout: 260 seconds)
[18:22:55] <-- _dev_urandom has left IRC (Read error: Connection reset by peer)
[18:23:10] --> devurandom has joined #GemRb
[18:23:19] --> Bo_Thomsen has joined #GemRb
[18:32:53] <-- devurandom has left IRC (Read error: Connection reset by peer)
[18:33:05] --> devurandom has joined #GemRb
[18:39:16] --> |Cable| has joined #GemRb
[18:42:56] --> _dev_urandom has joined #GemRb
[18:43:08] <-- devurandom has left IRC (Ping timeout: 260 seconds)
[18:50:08] <-- lynxlynxlynx_ has left IRC (Ping timeout: 276 seconds)
[18:50:20] --> lynxlynxlynx has joined #GemRb
[18:50:20] <-- lynxlynxlynx has left IRC (Changing host)
[18:50:20] --> lynxlynxlynx has joined #GemRb
[18:50:20] --- ChanServ gives channel operator status to lynxlynxlynx
[19:06:56] <CIA-36> GemRB: 03avenger_teambg * r90236ab29692 10gemrb/gemrb/core/DialogHandler.cpp: consolidated GameControl calls, and fixed an unlikely crash
[19:08:00] <CIA-36> GemRB: 03avenger_teambg * r47e48c040bbd 10gemrb/gemrb/core/Map.cpp: delete the original searchmap after copy (fixes a leak)
[19:11:16] <CIA-36> GemRB: 03avenger_teambg * r3ad47580b407 10gemrb/gemrb/plugins/GUIScript/GUIScript.cpp: when targeting self with an item it is not a point target (similar to spells)
[19:13:44] <CIA-36> GemRB: 03avenger_teambg * r0e6ab5bd4b99 10gemrb/gemrb/GUIScripts/pst/GUICommonWindows.py: missing EmptyControls caused unnecessary lines on console
[20:49:02] <-- _dev_urandom has left IRC (Remote host closed the connection)
[20:49:18] --> devurandom has joined #GemRb
[22:02:58] <-- Bo_Thomsen has left IRC (Quit: Leaving.)
[22:20:36] --> edheldil_ has joined #GemRb
[22:44:18] --> Avenger has joined #GemRb
[22:44:18] --- ChanServ gives channel operator status to Avenger
[22:44:26] <Avenger> hi
[22:45:12] <Avenger> fuzzie you were right about destruct order. I see this, after core was already destructed:
[22:45:15] <Avenger> CachedFileStream::~CachedFileStream() line 111 + 6 bytes
[22:45:17] <Avenger> CachedFileStream::`scalar deleting destructor'(unsigned int 1) + 40 bytes
[22:45:19] <Avenger> IDSImporter::~IDSImporter() line 39 + 35 bytes
[22:45:20] <Avenger> IDSImporter::`scalar deleting destructor'(unsigned int 1) + 37 bytes
[22:45:22] <Avenger> Held<Plugin>::release() line 33 + 77 bytes
[22:45:23] <Avenger> Holder<SymbolMgr>::~Holder<SymbolMgr>() line 64
[22:45:55] <fuzzie> is that the bug?
[22:46:00] <Avenger> should be
[22:46:04] <Avenger> this happened on quit
[22:46:05] <fuzzie> i mean, does it die there?
[22:46:13] <Avenger> caused an exception in msvc
[22:46:27] <fuzzie> 'after core was actually destructed' you mean, after main() returned?
[22:46:42] <Avenger> i see core fields filled with feeefeee
[22:47:47] <fuzzie> oh, right
[22:47:53] <fuzzie> it's going to be the GSUtils tables
[22:48:07] <fuzzie> which presumably aren't being deleted
[22:48:09] <Avenger> yep, some table is not explicitly freed before core is dead
[22:48:20] <fuzzie> global variables are bad :(
[22:48:37] <Avenger> so their automated garbage collection tries to run, but core is already liberated
[22:48:56] <fuzzie> so ok, let me look at how this worked before, if they didn't leak
[22:49:20] <Avenger> before the coming of the Holders?
[22:49:41] <Avenger> i'm almost sure we ignorantly leaked them :)
[22:50:08] <fuzzie> maybe so
[22:50:12] <fuzzie> ok, just clear them in the destructor?
[22:50:30] <fuzzie> they are already accessible from the Interface destructor
[22:50:57] <fuzzie> just 'triggersTable.release();'
[22:51:04] <fuzzie> and same for actionsTable, overrideActionsTable, objectsTable
[22:51:41] <tomprince> Or register a cleanup?
[22:51:45] <Avenger> does GameScript have some global release?
[22:51:58] <Avenger> register?
[22:52:01] <fuzzie> tomprince: in core?
[22:52:07] <tomprince> Why not?
[22:52:18] <Avenger> i thought all tables are registered anyway
[22:52:44] <fuzzie> global variables with destructors are a nightmare anyway
[22:53:33] <fuzzie> i guess RegisterCleanup would be fine too
[22:54:01] <fuzzie> just taking the lazy way, here
[22:54:19] <lynxlynxlynx> now for something completely different: how do you mages feel about applying gemrb for gsoc?
[22:54:56] <lynxlynxlynx> we still have time and i'm willing to take over the administrative part (already did that once too)
[22:55:12] <lynxlynxlynx> i don't think i'm capable enough to mentor though
[22:55:16] <fuzzie> i think you would be the only person with time for either :/
[22:56:03] <Avenger> all ids tables are already in symbols list
[22:56:25] <Avenger> and i thought Interface would destroy them
[22:56:31] <fuzzie> Avenger: if you keep a smart reference to the symbol around, you can't delete the symbol, to avoid crashes
[22:56:33] <Avenger> but apparently no
[22:56:39] <Avenger> ahh
[22:56:46] <Avenger> that's so 'smart' :P
[22:56:48] <fuzzie> these symbols shoudl not be global really, they should be in Interface, then it would be magic
[22:57:02] <Avenger> but they are in interface
[22:57:08] <Avenger> i mean, in symbols
[22:57:15] <Avenger> which is a nice list
[22:57:17] <fuzzie> they are global variables in GSUtils.cpp, though
[22:57:31] <lynxlynxlynx> fuzzie: not interested in participating as a student?
[22:57:43] <fuzzie> lynxlynxlynx: if the timetable is like previous years, it covers two exam periods for me
[22:57:48] <Avenger> well, the holders are a secondary development
[22:58:00] <lynxlynxlynx> oh
[22:58:29] <lynxlynxlynx> Avenger: would you be willing to mentor?
[22:58:31] <Avenger> initially there was this symbols list, we should still use them somehow
[22:58:32] <fuzzie> (mid May until mid June, then first half of August)
[22:58:43] <Avenger> mentor? i don't think i can do that
[22:59:00] <Avenger> i never been one, and i'm generally a bad supervisor :)
[22:59:07] <lynxlynxlynx> why not? you don't need pedagogic knowledge
[22:59:13] <fuzzie> Avenger: everything in the symbols list gets zapped at the right time, unless you can't delete it because you keep a reference to it
[22:59:26] <lynxlynxlynx> i can help you with the supervision bit
[22:59:28] <fuzzie> just do .release() on the stupid globals :-)
[22:59:43] <Avenger> ok fuzzie, we need some release call into GameScript
[22:59:46] <fuzzie> you might ask tomprince re: mentoring?
[22:59:54] <lynxlynxlynx> it's more about design decisions and code sanity
[23:00:01] <Avenger> lynx: i can give spiritual support :)
[23:00:02] <fuzzie> Avenger: they're in the header file, you should be able to do it from anywhere
[23:00:17] <Avenger> code sanity --> muhahaha
[23:00:23] <lynxlynxlynx> that doesn't cut it
[23:00:31] <lynxlynxlynx> new people would be lost here
[23:00:58] <Avenger> not more than we are
[23:01:19] <lynxlynxlynx> optimist :P
[23:01:34] <fuzzie> i think you'd have to start by deciding on possible projects
[23:01:57] <lynxlynxlynx> sure, but first i need to know if it makes sense at all
[23:02:10] <fuzzie> i think that might be a better way to work that out, i mean
[23:02:29] <Avenger> is this some google summer of code mentoring?
[23:02:33] <tomprince> I'd be somewhat available for the first part of the summer. I have a month were I am essentially incommunicado.
[23:02:52] <fuzzie> e.g. an opengl renderer might be fairly independent of the rest of the code
[23:02:54] <tomprince> I do know the infrastructure stuff, but I am not really up to speed on the IE stuff. :)
[23:03:09] <fuzzie> and i think that kind of infrastructure stuff is probably the best bet for projects
[23:03:16] <Avenger> yeah fuzzie is right, or native directx support for windows gurus :)
[23:03:25] <Avenger> or sdl sound for windows?
[23:03:47] <lynxlynxlynx> yes, ie-light stuff is easier to approach
[23:04:02] <Avenger> a bigger bit: multiplayer support, HAHA
[23:04:13] <lynxlynxlynx> uh no
[23:04:17] <Avenger> :D
[23:04:32] <fuzzie> i think any serious non-infrastructure work would need so much IE knowledge and RE to make it ridiculous
[23:04:32] <lynxlynxlynx> it has to be reasonably managable
[23:05:00] <Avenger> well, then some non ie specific plugin as fuzzie said
[23:05:10] <lynxlynxlynx> a gl 2 / gl es renderer sounds like a great idea
[23:05:30] <lynxlynxlynx> the embedded people would be happy too
[23:05:36] <Avenger> since we already have an example sound or gfx plugin, writing another should be easy if the guy knows the target
[23:05:55] <fuzzie> the graphics plugin would be a nightmare i think
[23:05:56] <lynxlynxlynx> well, sdlvideo needs to be trimmed first
[23:06:17] <fuzzie> but maybe not an impossible nightmare to get something working in 3 months
[23:06:21] <Avenger> part of the project would be to refactor the common parts into video :)
[23:06:52] <lynxlynxlynx> or just copy all the bloat initially
[23:06:55] <fuzzie> yes, once there was a second plugin and someone could see what the common parts are :-)
[23:07:04] <Avenger> hehe
[23:07:25] <fuzzie> it is pretty tricky and we sabotaged it a bit already with things like the MOS importer
[23:08:04] <Avenger> i have to go. someone please think of the symbol table destructors
[23:08:08] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014])
[23:08:36] <lynxlynxlynx> what about windowing? this is pretty light on ie too
[23:08:55] <lynxlynxlynx> do we need a new design or just fix all the corner cases?
[23:08:57] <fuzzie> well, you'd have to have all the games
[23:09:27] <fuzzie> there's some mumbling by me, and i think discussion by tomprince, in the logs
[23:09:47] <fuzzie> but it depends what you want
[23:10:13] <fuzzie> tomprince was pondering having a layer on top of the CHU files and removing a lot of the complexity of the python, and hopefully dealing with a lot of things on the way
[23:10:45] <lynxlynxlynx> wasn't this mostly about the initial setup?
[23:11:04] <fuzzie> i think our bugs mostly are about the initial setup/removal
[23:11:38] <lynxlynxlynx> mostly removal
[23:11:38] <tomprince> Any reason cmake was changed to not allow in tree builds?
[23:11:55] <CIA-36> GemRB: 03avenger_teambg * r00387dcba431 10gemrb/gemrb/override/pst/ (pacify.pro pacify.spl): pacify projectile
[23:12:03] <lynxlynxlynx> someone screwed up, so i made it mandatory
[23:12:06] <fuzzie> tomprince: people kept doing it, and then having a CMakeCache.txt in their root breaking all their out-of-tree builds
[23:12:29] <lynxlynxlynx> we can add an override flag for people who know what they want
[23:12:33] <tomprince> Do people regularly have multiple out-of-tree-builds?
[23:12:37] <fuzzie> no
[23:13:04] <fuzzie> but people find cmake confusing and do an in-tree build first by accident, then re-read the instructions, make a build dir, and get very upset when cmake doesn't work
[23:13:54] <fuzzie> but it is an avoiding-problems change, an override flag would make sense :)
[23:15:30] <edheldil_> fuzzie: no wonder, "cmake ." is an example in some help
[23:17:51] <CIA-36> GemRB: 03tom.prince * rdc1f38ba8e6f 10gemrb/gemrb/core/GameScript/GameScript.cpp:
[23:17:51] <CIA-36> GemRB: GameScript: Explicitly cleanup symbol tables on exit.
[23:17:51] <CIA-36> GemRB: There were getting cleaned up after ~Interface, causing exceptions on MSVC.
[23:17:51] <CIA-36> GemRB: Signed-off-by: Tom Prince <tom.prince@ualberta.net>
[23:17:55] <lynxlynxlynx> must be old, definitely not ours
[23:18:02] <fuzzie> tomprince: thanks!
[23:18:13] <fuzzie> i think possibly also causing the insane behaviour of mingw builds
[23:20:23] <lynxlynxlynx> can't think of any other project ideas
[23:20:41] <lynxlynxlynx> the animation system one sounds too hard
[23:21:25] <tomprince> :q
[23:22:05] <edheldil_> animation system? What do you mean, fuzzie?
[23:24:03] <lynxlynxlynx> avatars
[23:24:51] <lynxlynxlynx> we have a lot of problem with bad animation types, since they're so varying and we're not adaptable
[23:25:10] <fuzzie> i think it would require considerable RE work
[23:25:13] <lynxlynxlynx> so things just disappear on the screen sometimes
[23:25:23] <fuzzie> except for PST, of course
[23:25:50] <edheldil_> of course?
[23:26:03] <fuzzie> pst's animation data is all in a nice ini file, as discussed
[23:26:12] <edheldil_> ah, ok
[23:26:26] <fuzzie> it is all documented in a forum post somewhere from RE, too
[23:26:55] <tomprince> One wacky idea I had was to write a libgit2 backed ResourceSource.
[23:26:59] <fuzzie> i guess bg2's code must be partially known since there's a modding for it
[23:27:10] <fuzzie> modding? mod
[23:27:13] <tomprince> And a corresponding weidu backend.
[23:27:55] <fuzzie> anyone willing to try and implement things in weidu is a brave person :)
[23:27:56] <lynxlynxlynx> i don't think ia adds any new animations
[23:28:31] <fuzzie> there's some bg2 engine mod which allows addition of new anims by means of automated exe modding based on a 2da file and weidu scripting
[23:28:36] <fuzzie> is it ia?
[23:28:59] <lynxlynxlynx> i'm not sure, that description sounds new
[23:29:01] <fuzzie> i know that underneath it is just reusing anim ids, but i think the actual avatar stuff is quite complex
[23:29:08] <lynxlynxlynx> i had infinity animations in mind
[23:29:15] <lynxlynxlynx> maybe it was expanded
[23:29:37] <lynxlynxlynx> maybe i misunderstood :)
[23:29:37] <fuzzie> it does seem likely to be IA
[23:29:41] <tomprince> IA adds a bunch of animations from the other games.
[23:29:58] <fuzzie> "pick from a list of 44 different animation types", "add custom game animations for creatures and characters", etc
[23:29:59] <lynxlynxlynx> iwd mostly, but bg2 supports that anyway
[23:31:04] <fuzzie> i looked at it because i was wondering how they ported pst anims
[23:31:17] <fuzzie> and it turned out that most of their ports were into iwd slots, AAAAA :( but they stopped doing that
[23:32:27] <fuzzie> but yes IA is smart stuff, and since i seem unable to write something more coherent than 'smart stuff' i am off to bed, ninight
[23:33:08] <edheldil_> gn
[23:34:30] <lynxlynxlynx> uh, it's already tommorow
[23:34:57] <lynxlynxlynx> best be going too
[23:52:24] <-- lynxlynxlynx has left IRC (Ping timeout: 264 seconds)