#gemrb@irc.freenode.net logs for 11 Feb 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:46:31] <-- Maighstir has left IRC (Quit: .)
[00:54:41] <-- Drakkar has left IRC (Ping timeout: 255 seconds)
[01:00:28] --> Drakkar has joined #gemrb
[01:24:48] --> rocket_hamster has joined #gemrb
[02:42:08] <-- chiv has left IRC (Ping timeout: 245 seconds)
[03:01:30] <-- rocket_hamster has left IRC (Read error: Connection reset by peer)
[07:59:19] --> edheldil has joined #gemrb
[07:59:20] --- ChanServ gives channel operator status to edheldil
[08:03:54] <-- edheldil has left IRC (Ping timeout: 276 seconds)
[08:07:48] <-- |Cable| has left IRC (Ping timeout: 276 seconds)
[08:20:22] --> |Cable| has joined #gemrb
[08:34:43] --- ermo^ is now known as ermo
[08:43:06] --> fizzle has joined #gemrb
[08:48:44] --> WingedHussar has joined #gemrb
[09:03:05] --> lynxlynxlynx has joined #gemrb
[09:03:05] --- ChanServ gives channel operator status to lynxlynxlynx
[09:06:50] --- ermo is now known as ermo^
[10:06:47] <-- fizzle has left #gemrb
[10:55:49] --- DrMcCoy is now known as DocMcCoy
[10:56:04] --- DocMcCoy is now known as DrMcCoy
[11:39:02] <-- |Cable| has left IRC (Ping timeout: 252 seconds)
[11:51:36] --> |Cable| has joined #gemrb
[12:22:56] <-- |Cable| has left IRC (Ping timeout: 256 seconds)
[12:31:50] --> fizzle has joined #gemrb
[12:35:41] --> |Cable| has joined #gemrb
[12:49:34] --> rocket_hamster has joined #gemrb
[12:51:37] <lynxlynxlynx> fizzle: none of those bg2 areas are ambush areas
[12:52:20] <lynxlynxlynx> http://paste.debian.net/233333/
[12:52:22] <Seniorita> debian Pastezone
[12:53:41] <fizzle> hm. now what?
[12:54:09] <lynxlynxlynx> now it's time to confront dltcep and ieshell
[12:54:21] <lynxlynxlynx> dltcep shows no extra flags for these areas
[12:54:32] <lynxlynxlynx> let me check an actual ambush area
[12:56:37] <lynxlynxlynx> yeah, dltcep seems to be right
[12:57:37] <lynxlynxlynx> what ed ran also looks right, so maybe his data is different
[13:25:56] --- ermo^ is now known as ermo
[13:39:29] <edheldil_> yes?
[13:59:13] <-- traveler has left IRC (Ping timeout: 245 seconds)
[15:27:32] <lynxlynxlynx> are you sure you ran the AF_SAVE check on bg2?
[15:27:55] <lynxlynxlynx> not pst or something else
[15:39:17] <edheldil_> I haven't found any area with this flag set in pst ... What's the problem?
[15:45:17] <edheldil_> is there perhaps a terminological problem? e.g. AR2205 has a flag 0x1 set, which should be can_save, but NI displays it as a flag can't_save=0
[15:55:10] <fuzzie> 0x1 is AREA_AREAFLAGS_NOSAVE, which makes no sense for the bg2 areas listed
[15:56:48] --> traveler has joined #gemrb
[15:58:41] <fuzzie> iwd2's areaflag.ids is consistent with that too, although interestingly it has 0x2 as NOREST and 0x4 as LOCKBATTLEMUSIC
[15:58:58] <fuzzie> so I think we're just confused as to the inconsistency
[16:05:45] <-- WingedHussar has left IRC (Quit: WingedHussar)
[16:07:32] <edheldil_> well, what I listed are those areas with 0x1 set, regardless of meaning
[16:10:11] <fuzzie> thanks for assisting in our confusion, I guess :-)
[16:11:36] <edheldil_> hmmm, rather making it worse .... It looks to me like iesh is opening something else then AR2205, so I will have to check
[16:17:08] --> edheldil has joined #gemrb
[16:17:09] --- ChanServ gives channel operator status to edheldil
[16:17:32] <traveler> sorry edheldil for bugging you, but did your iwd/2 got installed with all character sounds?
[16:17:52] <traveler> i mysteriously have Sounds/ both in iwd2 and iwd populated by only female ones
[16:18:05] <edheldil> sorry for the confusion. Iesh now came up with AR0041
[16:18:06] <edheldil> AR0042
[16:18:06] <edheldil> AR0043
[16:18:06] <edheldil> AR0044
[16:18:06] <edheldil> AR0045
[16:18:06] <edheldil> AR0046
[16:18:08] <edheldil> AR2601
[16:18:10] <edheldil> in bg2
[16:21:30] <traveler> and there wasn't even install options, so probably all installs were 'full by default'. yeah, it's pretty minor but still weird. got me wondering what _else_ can be botched in those installs.
[16:24:46] <edheldil_> I have male sounds, but I installed iwd2 completely by hand
[16:26:05] <lynxlynxlynx> edheldil_: those are the ambush areas, yes
[16:26:13] <lynxlynxlynx> just the last needs a check, sec
[16:27:08] <lynxlynxlynx> yeah, that's one too (my bet is on drizzt)
[16:28:26] <edheldil_> sorry for the confusion :(
[16:34:13] <-- traveler has left IRC (Ping timeout: 245 seconds)
[17:19:33] <-- rocket_hamster has left IRC (Quit: bye!)
[17:23:59] --> chiv has joined #gemrb
[17:26:09] <-- edheldil has left IRC (Ping timeout: 276 seconds)
[17:43:52] <fizzle> hah
[17:44:13] <fizzle> do I see any green lights there?
[17:56:12] <lynxlynxlynx> yep
[18:07:15] * fizzle starts stretching his fingers
[18:16:00] --> traveler has joined #gemrb
[18:23:32] <lynxlynxlynx> :)
[18:36:56] <Seniorita> [wiki] todo http://www.gemrb.org/wiki/doku.php?id=todo&rev=1360607500&do=diff
[18:48:29] --> Yoshimo has joined #gemrb
[19:00:06] <-- Yoshimo has left IRC (Read error: Connection reset by peer)
[19:04:59] <fizzle> brada: [OpenAL/ERROR]: Unable to queue buffer: 0xa004
[19:05:08] <fizzle> nothing else but this a couple of times
[19:05:19] --> Yoshimo has joined #gemrb
[19:05:24] <lynxlynxlynx> he needs the context, whatever it was
[19:05:50] <fizzle> I don't have any context
[19:06:04] <fizzle> other than no more background music after this
[19:07:13] <-- Yoshimo has left IRC (Client Quit)
[19:13:46] <lynxlynxlynx> what are the previous lines?
[19:14:30] <fizzle> [ResourceManager]: Searching for 'amb_n21a'...
[19:14:32] <fizzle> [ResourceManager]: Found 'amb_n21a.wav' in 'chitin.key'.
[19:14:35] <fizzle> for example
[19:20:50] <fizzle> same for paper.wav, WAL_04a.wav, amb_m32b...
[19:32:29] --> brada has joined #gemrb
[19:36:27] <traveler> edheldil: right, plain unshield will do. i was expecting unified installer like my bg version, so just install onto pendrive with windows laptop.
[19:36:47] <traveler> *installed
[19:38:52] --> edheldil has joined #gemrb
[19:38:52] --- ChanServ gives channel operator status to edheldil
[19:41:14] <brada> fizzle: looks like your openal context is lost
[19:41:29] <brada> im not sure what we should be doing
[19:41:44] <brada> we can i guess catch this error and try to recreate the context
[19:42:06] <brada> but AFIK you shouldnt just loose the contxt for no reason
[19:42:10] <fizzle> shouldn't that also affect other sounds?
[19:42:16] <brada> so something else may be the proper fix
[19:43:07] --> chiv_ has joined #gemrb
[19:43:10] <brada> the music and sounds have diffrent sources
[19:43:23] <-- traveler has left IRC (Ping timeout: 245 seconds)
[19:43:48] <-- chiv has left IRC (Ping timeout: 245 seconds)
[19:44:42] <brada> being able to replicate that would help me
[19:45:16] <brada> if i could replicate it while running with a debugger i could maybe see what is happening
[19:46:07] <fizzle> I haven't managed to reproduce it deliberately
[19:46:17] <fizzle> and it happens a lot less often since your last changes
[19:47:32] --> traveler has joined #gemrb
[19:48:17] <brada> sure thats understandable
[19:48:55] <brada> before rather benign errors that werent caugth could cause us to think other bad things happend the next time we checked the al error state
[19:49:20] <traveler> edheldil: well, do'h... gemrb doesn't see extracted male sounds, because they have polish glyphs in names which got copied as '?' which is not very unix friendly way
[19:49:42] <brada> so rechecking the documentation the exact reasons for AL_INVALID_OPERATION after calling alSourceQueueBuffers
[19:49:46] <brada> There is no current context, an attempt was
[19:49:47] <brada> made to add a new buffer which is not the
[19:49:48] <brada> same format as the buffers already in the
[19:49:49] <brada> queue, or the source already has a static buffer
[19:49:50] <brada> attached
[19:49:55] <brada> so probably the context is fine
[19:50:01] <brada> maybe the other reasons
[19:54:15] <brada> i think i have an idea
[19:54:37] <brada> no time right now tho
[19:55:07] <-- edheldil has left IRC (Ping timeout: 260 seconds)
[19:55:57] --- ermo is now known as ermo^
[21:14:35] <brada> i think the problem might be somewhere we queue a null buffer wich resets the source to AL_UNDETERMINED instead of AL_STREAMING
[21:17:19] --> edheldil has joined #gemrb
[21:17:19] --- ChanServ gives channel operator status to edheldil
[22:16:05] <brada> fizzle/traveler: i pushed some more openal fixes / error handling
[22:42:41] <fizzle> brada: finally got one in the debugger
[22:42:50] <fizzle> what do you need to know?
[22:43:12] <fizzle> with your latest changes
[22:43:16] <brada> bt i guess would be a nice start
[22:43:29] <brada> and i pushed a couple more changes did you get those?
[22:43:35] <fizzle> yes
[22:43:43] <brada> what happened to seniorita?
[22:44:06] <fizzle> GemRB::OpenALAudioDriver::QueueAmbient (this=0x7de970, stream=1, sound=0x1a86cc0 "amb_e06a")
[22:44:26] <brada> so ambient audio is your problem?
[22:44:28] <brada> not music?
[22:44:47] <fizzle> not sure if the two are related
[22:44:53] <fizzle> or separate issues
[22:45:28] <fizzle> missing background music is more obvious
[22:46:06] <brada> so what is the new error?
[22:46:13] <fizzle> same as before
[22:46:29] <fizzle> Unable to queue buffer
[22:46:41] <brada> AL_INVALID_OPERATION?
[22:47:06] <fizzle> yes
[22:47:16] <brada> so pretty much if other audio continues to work then the context is still valid
[22:47:30] <brada> and we now check that the source is not static
[22:47:49] <brada> so that leaves the only cause as something is already buffered in a diffrent format i guess
[22:48:13] <brada> not sure how to cope with that
[22:49:36] <brada> btw is this openal 1.0 or 1.1?
[22:49:42] <brada> logged at init
[22:50:14] <brada> and are there any other openal errors prior to that one?
[22:50:33] <brada> or even just openal messages of any kind
[22:52:08] <fizzle> what message?
[22:52:17] <fizzle> I don't see anything in the log
[22:53:27] <brada> well there should at least be a message saying the version :p
[22:53:50] <fizzle> not under component "OpenAL"
[22:54:25] <brada> should have [OpenAL]: Initializing OpenAL driver:
[22:54:35] <brada> followed by version / vendor
[22:55:04] <fizzle> I don't see that in the log or in the code
[22:55:30] <brada> openalaudiodriver.cpp line 178
[22:55:57] <lynxlynxlynx> added minutes ago
[22:56:49] <brada> those changes i asked you if you got and you said yes :p
[22:57:03] <fizzle> how could I know you added even more :P
[22:57:09] <lynxlynxlynx> fizzle: not sure if it's new since your changes, but polymorphing from a non-human to other non-human fails to update the avatar ingame (eg. jaheira's bear to wolf form)
[22:57:38] <lynxlynxlynx> inventory paperdoll is not affected
[22:58:34] <fizzle> brada: so I'll let the debugger go then...
[22:58:46] <fizzle> lynx: only case I tried is human to wolf
[22:59:09] <fizzle> (are there others in bg1?)
[22:59:14] <lynxlynxlynx> i don't know why polymorphing back is fine
[22:59:48] <lynxlynxlynx> either jaheira is weaker in bg1 or we have a bug, but in my save she had no polymorphing innates
[23:00:06] <fizzle> no innate
[23:00:11] <fizzle> cloak of wolf
[23:00:19] <lynxlynxlynx> not sure if the polymorph self spell is available (4th level or so)
[23:00:28] <fizzle> haven't seen it
[23:00:51] <fizzle> I guess polymorphing back works because something is constantly resetting the anim id
[23:01:07] <fizzle> which is why morphing away didn't work without the pcf
[23:01:22] <lynxlynxlynx> i don't see much change now
[23:01:34] <lynxlynxlynx> the big morphs still have the problem
[23:02:10] <lynxlynxlynx> but yeah, some of it is being reset all the time, since we give a lot of console spam
[23:03:34] <fizzle> brada: "Version:(null)" :P
[23:03:52] <lynxlynxlynx> we could use the base stat, but then that brings problems of its own
[23:04:40] <fizzle> brada: got it again, same place
[23:07:16] <lynxlynxlynx> i guess the reversion to human form just removes the original effect
[23:07:20] <brada> version null eh
[23:07:46] <traveler> brada: missed changes because null seniorita output, recompiling now
[23:08:29] <brada> well it looks like im going to have to add some debug output to output the format anytime something is queue
[23:08:35] <brada> later tho
[23:08:41] <brada> going to school now
[23:08:44] <-- brada has left IRC (Quit: brada)
[23:09:53] <fizzle> I'm off as well
[23:10:39] <-- fizzle has left #gemrb
[23:10:52] <traveler> fizzle/lynx: i just tested polymorphing in bg1 with jaheira, and for first time, it was just fine O_o....
[23:11:59] <traveler> bear, cave bear and wolf (innate)
[23:12:05] <traveler> this multi save included in tosc
[23:14:51] <traveler> ah, right
[23:15:02] <traveler> bear -> wolf does not indeed update
[23:15:07] <traveler> but i see other problem
[23:15:14] <traveler> when bear form bumps CON
[23:15:22] <traveler> it does not grant additional HP
[23:15:41] <traveler> and jaheira just goes from 57/57 to 57/64 HP
[23:16:10] <lynxlynxlynx> not much we can do about that
[23:16:35] <lynxlynxlynx> this full hp case yes, others not really
[23:17:03] <traveler> that's too bad
[23:17:13] <lynxlynxlynx> we do try somewhere already, but the con bonus sounds like being done later
[23:17:20] <traveler> i mean, isn't it general problem with bumping CON midplay?
[23:17:35] <lynxlynxlynx> sure
[23:17:59] <lynxlynxlynx> it's the same problem with anything bumping both hp stats too
[23:18:33] <lynxlynxlynx> in a non-absolute way - you get continuous healing
[23:18:57] <lynxlynxlynx> or in the iwd2 case, a permanent bonus
[23:19:03] <traveler> uh oh
[23:19:16] <lynxlynxlynx> we need some extra infrastructure to be able to deal with this
[23:20:02] <lynxlynxlynx> i think an "on removal" effect callback would helpful (for this avatar case and hide in shadows, where we use a separate stat)
[23:20:21] <lynxlynxlynx> for hp, some extra accounting
[23:21:16] <lynxlynxlynx> hp healing probably just requires some careful thinking
[23:22:04] <lynxlynxlynx> with that hook, it would be simple, since the base stat could be used and later reverted
[23:22:59] <traveler> *totally unrelated - it would be cool (in original? in EE?) if somebody would have thought that getting near e.g. bears in bear form shouldn't 'aggro' them as is usual case.*
[23:25:22] <lynxlynxlynx> that's not how it work in rl either
[23:30:16] <traveler> i only mention it, should my iwd2 version have some exotic audio content; i know iwd2 is very much WIP, anyway audio in iwd2 movies plays here at about half speed judging from pitch shift
[23:31:34] <lynxlynxlynx> iwd2 movies work fine here
[23:32:00] <lynxlynxlynx> they use a different format though, so it depends on how well the decoder works on your platform
[23:32:13] <lynxlynxlynx> oh, maybe you're even using the vlc plugin for it
[23:34:24] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:34:25] <traveler> lynx: you have experience in getting near bears while shapeshifted :)?
[23:34:46] <traveler> nah without vlc here
[23:35:47] <traveler> but i have history of playing pitch shifted audio
[23:36:00] <traveler> due to bitperfect playback so maybe it's local thing
[23:39:50] <traveler> nope, it's not it
[23:40:17] <traveler> i think that dubbing team could have in someway messed them
[23:40:31] <traveler> as only dubbed movies had this problem, not logo ones etc
[23:41:27] <traveler> prologue dub and other ones are just fine
[23:46:00] --- ermo^ is now known as ermo