#gemrb@irc.freenode.net logs for 13 Mar 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:50:19] --> MikeChelen has joined #GemRb
[00:51:04] <MikeChelen> can anyone help with this bug? https://sourceforge.net/tracker/?func=detail&aid=3200851&group_id=10122&atid=110122
[00:51:43] --> tomprince has joined #GemRb
[01:03:29] <-- edheldil_ has left IRC (Ping timeout: 255 seconds)
[01:54:30] <-- BaldimerBrandybo has left IRC (Read error: Connection reset by peer)
[01:55:56] --> Drakkar has joined #GemRb
[02:22:36] <-- tomprince has left IRC (Quit: WeeChat 0.3.3)
[02:34:10] <-- Maighstir has left IRC (Quit: ~ Trillian Astra - www.trillian.im ~)
[02:40:23] <-- MikeChelen has left IRC (Ping timeout: 252 seconds)
[06:14:48] --> tomprince has joined #GemRb
[06:58:33] --> pupnik_ has joined #GemRb
[06:59:43] <-- pupnik has left IRC (Read error: Operation timed out)
[06:59:58] <tomprince> https://gist.github.com/867942
[07:00:13] <tomprince> I don't know if those fixes are right, or if they are actually bugs.
[07:38:36] --- pupnik_ is now known as pupaway
[07:50:21] --> edheldil_ has joined #GemRb
[07:57:04] <-- tomprince has left IRC (Quit: WeeChat 0.3.3)
[08:00:38] <-- edheldil_ has left IRC (Ping timeout: 255 seconds)
[09:10:30] --> lynxlynxlynx has joined #GemRb
[09:10:30] --- ChanServ gives channel operator status to lynxlynxlynx
[10:47:41] <fuzzie> re tomprince patches: the resource stuff at least looks wrong at a glance
[10:48:04] <fuzzie> i mean, the patching of them
[10:48:39] <fuzzie> i'd have to look at source for the rest, so later
[11:28:29] <Aki_> Hi hi
[11:50:50] <lynxlynxlynx> blasted demo
[11:51:08] <lynxlynxlynx> i was all happy when i found a few files that it doesn't have, but are in soa
[11:51:27] <lynxlynxlynx> but it turned out the chitin is the same, so we don't detect they're missing
[12:08:49] <Aki_> What is that chitin.key in fact ?
[12:13:26] <lynxlynxlynx> it's a map of all the resources
[12:13:55] <lynxlynxlynx> a lot of files are archived in bif files, so this is a clear optimisation
[12:15:32] <Aki_> Ok so it say musib1.mp3 is between ofset X and Y of file z" ?
[12:16:15] <Aki_> So to extrat a BIF we need the .key file ?
[12:23:04] --> Maighstir has joined #GemRb
[12:27:57] --> SiENcE has joined #GemRb
[12:43:08] <lynxlynxlynx> i don't think so, the bif file format is very simple (check iesdp), header with the offsets and then the data
[12:43:27] <lynxlynxlynx> maybe without the filenames though
[12:47:43] <CIA-27> GemRB: 03avenger_teambg * rb23a026114e3 10gemrb/gemrb/core/Inventory.cpp: drop gold stored as stat
[12:47:48] <CIA-27> GemRB: 03avenger_teambg * r2d1541aa4f14 10gemrb/gemrb/plugins/CREImporter/CREImporter.cpp: duplicated items are now tolerated by GemRB's creature importer (fixes some missing items)
[12:47:55] <Aki_> ok lynxlynxlynx , understood
[12:55:54] <lynxlynxlynx> nice
[12:59:24] <Maighstir> The key file mostly tells which bif a specific file sits it, doesn't it?
[12:59:53] <Maighstir> Not where in the bif.
[13:17:07] <lynxlynxlynx> yeah
[13:18:12] <lynxlynxlynx> http://iesdp.gibberlings3.net/file_formats/ie_formats/key_v1.htm
[13:18:36] <lynxlynxlynx> 4 dwords that then match in the bif header
[14:26:58] <CIA-27> GemRB: 03lynxlupodian * r5564edf2a57d 10gemrb/gemrb/GUIScripts/ (GUICommon.py bg2/Autodetect.py): added bg2 demo autodetection
[14:45:57] <-- SiENcE has left IRC (Quit: bye)
[14:50:43] <CIA-27> GemRB: 03lynxlupodian * ra2f3c2452e6a 10gemrb/gemrb/GUIScripts/bg2/Start2.py:
[14:50:43] <CIA-27> GemRB: bg2: don't allow entering the tutorial from the demo, since the data is
[14:50:43] <CIA-27> GemRB: missing (crash); multiplayer and movies for consistency
[15:23:17] <CIA-27> GemRB: 03lynxlupodian * r9253b90749e5 10gemrb/gemrb/GUIScripts/bg2/Start2.py: bg2: correct a few ui glitches for the demo
[15:23:23] <CIA-27> GemRB: 03lynxlupodian * r757d114d3d7c 10gemrb/gemrb/GUIScripts/bg2/Start2.py: bg2: removed a double negation
[15:33:36] <lynxlynxlynx> another two items off the list :)
[15:43:13] <lynxlynxlynx> web also updated
[15:55:38] <fuzzie> ugh, stupid troll bug
[17:46:06] --> MikeChelen has joined #GemRb
[17:48:21] --- pupaway is now known as pupnik
[18:32:50] --> tomprince has joined #GemRb
[18:42:56] <-- tomprince has left IRC (Ping timeout: 246 seconds)
[18:54:48] --> tomprince has joined #GemRb
[19:52:10] <-- tomprince has left IRC (Quit: WeeChat 0.3.3)
[20:54:06] --> Avenger has joined #GemRb
[20:54:06] --- ChanServ gives channel operator status to Avenger
[20:55:15] <Avenger> the sound3 resref in .pro files is actually a vvc resource. So the field name in IESDP is closer to the truth. 0x00020 8 (resref) Travelling BAM
[20:55:34] <Avenger> it is in fact, a vvc, but bams work as a substitute, of course
[20:55:58] <Avenger> i don't really see why there are two travel animations, though
[20:56:54] <fuzzie> did you look at wherever the notes are on the forum?
[20:57:11] <Avenger> no
[20:57:30] <Avenger> it also has a different speed, heh
[20:57:46] <Avenger> are there notes?
[20:57:53] <fuzzie> on PRO files
[20:58:15] <Avenger> i can never find anything on the forums
[20:58:25] <fuzzie> um
[20:58:46] <Avenger> i think we talked about this and i insisted it is a wav resource :)
[20:59:41] <Avenger> this was something i remember testing, but i must have been mixing it up with the second sound resource
[20:59:53] <Avenger> so, gemrb's code is a mess
[20:59:54] <fuzzie> http://forums.gibberlings3.net/index.php?showtopic=20117&st=0&p=168902&#entry168902
[21:00:04] <fuzzie> look at top post
[21:00:21] <Avenger> 0x20 WAVC file
[21:00:23] <Avenger> ah yes
[21:01:32] <CIA-27> GemRB: 03lynxlupodian * re54e06db18f5 10gemrb/gemrb/core/GameScript/GSUtils.cpp: AttackCore: stop the attack if the two parties can't see each other
[21:01:34] <fuzzie> ascension64's assorted notes are helpful but a bit scattered randomnly around the forum
[21:01:40] <fuzzie> lynxlynxlynx: argh, wha?
[21:02:04] <lynxlynxlynx> wha argh?
[21:02:06] <fuzzie> that doesn't sound good :P
[21:02:24] <fuzzie> it is not good
[21:02:29] <lynxlynxlynx> explain, i have one test case
[21:02:43] <fuzzie> well, obvious test case is chases between areas
[21:03:00] <fuzzie> what is your test case?
[21:03:13] <lynxlynxlynx> shooting through blocked cells
[21:03:15] <fuzzie> that is definitely not good though
[21:03:24] <lynxlynxlynx> doors, walls, infinite range
[21:03:46] <fuzzie> enemies will happily follow you all the way through an area for an attack
[21:03:53] <lynxlynxlynx> i can move it down then
[21:04:08] <lynxlynxlynx> after these area chase checks
[21:04:15] <fuzzie> well i'm confused as to what has happened with this code
[21:05:15] <fuzzie> i mean, the CanSee belongs in the check around MoveNearerTo
[21:06:18] <lynxlynxlynx> nothing recently, most of it is 2009
[21:06:25] <fuzzie> oh, it got temporarily commented out and then Avenger removed the comment
[21:06:40] <fuzzie> along with the TODO above it :P
[21:06:47] <Avenger> heh
[21:07:08] <fuzzie> so yes, put it there, i guess..
[21:07:29] <lynxlynxlynx> i can see it was an error
[21:08:31] <fuzzie> yes, and it only matters for attack actions which would try following you if you go out of sight, which is not all of them
[21:08:44] <fuzzie> so another silly thing we missed?
[21:10:28] <lynxlynxlynx> do you have Avenger's commit hash? nothing strikes out from the commit log
[21:10:56] <fuzzie> 557fa26d, a long time ago
[21:11:11] <fuzzie> i'm fairly sure i mentioned it on irc at the time and promptly forgot about it
[21:11:52] <fuzzie> since at the time we were busy struggling with the dream scene etc
[21:12:52] <fuzzie> and half the work was out-of-tree
[21:13:33] <fuzzie> now you should assure me that i'm not crazy and the commit does do it :)
[21:13:52] <lynxlynxlynx> it does
[21:15:17] <fuzzie> but i think there should work perfectly
[21:15:24] <fuzzie> if that is your bug
[21:17:07] <Avenger> fuzzie, what will happen if i don't release travel_handle?, only Stop it?
[21:17:16] <lynxlynxlynx> i see why it was a todo though, GoNearAndRetry is not implemented
[21:17:26] <Avenger> i want this: if it is looping, then it stops, but if it isn't looping then it plays till the end
[21:17:52] <Avenger> would .Stop turn off the looping flag?
[21:17:59] <Avenger> or also stop the sound
[21:18:01] <fuzzie> no
[21:18:08] <fuzzie> call StopLooping() :)
[21:18:12] <Avenger> ok
[21:18:33] <fuzzie> or is that not what you want?
[21:18:40] <Avenger> do i still have to call release?
[21:18:54] <Avenger> i want the explosion sound of a projectile to play fully
[21:19:17] <fuzzie> well, at the moment there is no way to see whether something is looping or not
[21:19:29] <Avenger> i will call stoplooping
[21:19:32] <Avenger> that's enough
[21:19:39] <fuzzie> but if you just want to turn off the looping sound, stoplooping is good
[21:19:59] <fuzzie> and if it is not enough, it is easy to add something else
[21:20:13] <Avenger> i just don't call .release, and hope it won't leak
[21:20:26] <fuzzie> it won't leak
[21:20:53] <fuzzie> but if you don't call .release(), then travel_handle remains set
[21:21:02] <fuzzie> so some 'if (travel_handle)' check still returns true
[21:21:12] <Avenger> it is in the destructor :)
[21:21:32] <fuzzie> that is the only difference
[21:21:51] <Avenger> it won't be touched after that so that's ok
[21:22:20] <lynxlynxlynx> http://sprunge.us/aJWA?diff <-- better?
[21:22:44] <fuzzie> the TODO is v.outdated
[21:23:02] <lynxlynxlynx> sounds reasonable though
[21:23:07] <fuzzie> the important bit was "when sender can't see target" and you fixed it
[21:23:14] <fuzzie> i think i deleted GoNearAndRetry a while ago
[21:23:34] <lynxlynxlynx> the declarations remain, so i thought we were lacking it
[21:23:41] <fuzzie> oops
[21:24:01] <fuzzie> ok well that is stupid on my part :)
[21:25:10] <CIA-27> GemRB: 03lynxlupodian * rc628073c6483 10gemrb/gemrb/core/GameScript/GSUtils.cpp: AttackCore: better impossible attack prevention
[21:25:18] <fuzzie> thankyou for noticing
[21:25:50] <lynxlynxlynx> want me to remove those too?
[21:29:05] <lynxlynxlynx> heh, it was the same thing, you commented it out, Avenger deleted it
[21:29:25] <fuzzie> ah, hehe
[21:29:41] <Avenger> i don't like pieces of dead code around :P
[21:29:45] <Avenger> i thought it was not needed
[21:30:16] <fuzzie> well this is wqhy there was a TODO comment above :-) but it did look like a very old one
[21:30:52] <fuzzie> i might have accidentally removed it also
[21:30:53] <lynxlynxlynx> there's a loop with the remains of GoNearAndRetry though
[21:30:56] <fuzzie> was just trying to work out where it had gone!
[21:31:16] <lynxlynxlynx> it used movetopoint and added 1 to distance
[21:31:25] <lynxlynxlynx> movetopoint still removes it to account for that
[21:31:55] <fuzzie> yes, i guess that whole codeblock is dead
[21:34:35] <lynxlynxlynx> nothing from us uses int0 anymore
[21:35:36] <lynxlynxlynx> huh, no move* actions in iesdp
[21:38:14] <lynxlynxlynx> ah, again the page bug
[21:38:31] <lynxlynxlynx> how can such a simple design be so buggy
[21:42:58] <lynxlynxlynx> yep, dead also in the dialog and script dumps
[21:46:33] <lynxlynxlynx> http://sprunge.us/gOfC?diff ack?
[21:47:10] <fuzzie> yes
[21:47:31] <fuzzie> (0 is default param for distance, so could just remove that)
[21:49:36] <lynxlynxlynx> now for the other half of the bug :)
[21:49:47] <CIA-27> GemRB: 03lynxlupodian * r4f3918583921 10gemrb/gemrb/core/GameScript/ (Actions.cpp GSUtils.h): finish the GoNearAndRetry removal
[21:55:50] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.15/20110303024726])
[22:05:22] <lynxlynxlynx> for spells looks like an arbitrary constant is to blame
[22:11:18] <fuzzie> spellcasting without visibility?
[22:13:38] <lynxlynxlynx> that too, the first noticeable thing is the extreme range
[22:13:58] <lynxlynxlynx> GetSpellDistance multiplies by 15 in the end
[22:14:07] <lynxlynxlynx> no info in the parent commit why
[22:14:15] <lynxlynxlynx> 5 appears more reasonable from testing
[22:15:04] <lynxlynxlynx> the visibility part is just more missing cansee/isvisible calls
[22:27:14] <CIA-27> GemRB: 03lynxlupodian * raf70c373e368 10gemrb/gemrb/core/GameScript/Actions.cpp: take visibility into account when spellcasting (in the normal cases)
[22:27:21] <CIA-27> GemRB: 03lynxlupodian * r03b1022bf363 10gemrb/gemrb/core/GameScript/GSUtils.cpp: GetSpellDistance: use a more reasonable undocumented scaling factor
[22:28:01] <fuzzie> well 15 sounds like it's trying to approximate search square size, which is how the original engine usually measured these things
[22:33:09] <lynxlynxlynx> the math before that doesn't like it, being all point based
[22:40:03] <fuzzie> well it's strange because 5 doesn't match anything much at all that i can think of
[22:40:42] <fuzzie> "if (dist != 0xfffffff) {" <- that is missing an f, deliberate?
[22:44:51] <lynxlynxlynx> sharp eye
[22:45:04] <lynxlynxlynx> return 0xffffffff;
[22:45:25] <lynxlynxlynx> kerning sucks for this
[22:45:46] <fuzzie> i keep finding those in various bits of code
[22:46:57] <lynxlynxlynx> seems to be the only case not in guiscripts
[22:48:38] <fuzzie> oh yes, i see them in guiscripts
[22:49:00] <CIA-27> GemRB: 03lynxlupodian * r89bc76e0e977 10gemrb/gemrb/core/GameScript/Actions.cpp: GameScript::Spell: fuzzie fixed ffffffffff from faulty fffffff
[22:49:13] <fuzzie> and the other case, where 0xfffffff is passed to DisplayString, one too many f i guess?
[22:49:38] <lynxlynxlynx> probably, i doubt it supports alpha
[22:49:56] <fuzzie> well i noticed the core doesn't pass alpha anywhere :)
[22:50:53] <lynxlynxlynx> something for another day, good night
[22:50:59] <fuzzie> night!
[22:51:32] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:34:59] <CIA-27> GemRB: 03avenger_teambg * r6bdbaac40f61 10dltcep/ (7 files): dltcep update (projectiles)
[23:40:41] <CIA-27> GemRB: 03avenger_teambg * rb6b9d9e45124 10gemrb/gemrb/ (3 files in 2 dirs): cleared projectile wav/vvc resource fields
[23:44:29] --> Avenger has joined #GemRb
[23:44:29] --- ChanServ gives channel operator status to Avenger
[23:44:51] <Avenger> i see fuzzie you found a bad 0xfffffff constant. rgrep -w 0xfffffff shows a few more :)
[23:45:16] <Avenger> only in guiscript
[23:45:52] <-- Avenger has left IRC (Client Quit)