#gemrb@irc.freenode.net logs for 14 Jan 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:00:28] <brada> heh
[00:00:56] <brada> i assume you are saying we evaluate everything even if a preceeding portion of the OR is false?
[00:09:35] <lynxlynxlynx> no, that'd be normal, first portion was true
[00:10:31] <lynxlynxlynx> probably required due to side effects
[00:13:33] <brada> thats what i mean
[00:13:49] <brada> my words and thoughts seem to be opposite today :/
[00:13:56] <brada> too much beer last night i think
[00:14:43] <lynxlynxlynx> as long as you don't code at the same time ;)
[00:15:18] <brada> well i dont commit anyway ;)
[00:15:23] <lynxlynxlynx> but yeah, it's now very late here, going to zzz
[00:15:41] <lynxlynxlynx> that works
[00:16:25] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[00:18:48] <-- traveler has left IRC (Ping timeout: 245 seconds)
[01:09:44] <edheldil> hi all, there's a python bug in the current master when running pst: http://paste.debian.net/224404/
[01:09:47] <Seniorita> debian Pastezone
[01:11:39] <edheldil> brada: one possibility is to use a file w/ codepoint->BAM index mapping
[01:12:19] <brada> for what ed?
[01:13:23] <edheldil> <brada> not sure at all how to actually implement it tho
[01:13:23] <edheldil> <brada> since dbString[dbLen] = (string[i]-0x81)*255 + string[i+1] + 128; is quite specific to Chinese ill bet...
[01:13:46] <brada> ah yes. but to do that i would need to know how they are organized for other languages
[01:14:03] <brada> other double byte languages anyway
[01:14:25] <brada> and that is a hell of a big map...
[01:14:31] <brada> vs 1 line of code
[01:15:50] <edheldil> sure :)
[01:17:12] <brada> i dont know anything about the layout of the font bams beyond the extended ascii range
[01:17:28] <brada> that code came from the original chinese patch
[01:19:33] <edheldil> well, the solutions I can see is: hardcoded hack, plugin, external table, external patch
[01:20:58] <brada> thats been my lament; im not particularly happy with any solution
[01:21:37] <brada> i was hoping the other bams for other languages were laid out similar to the chinese ones
[01:22:06] <brada> not sure who would know that
[01:23:16] <brada> i may just leave the bam stuff unfinished for now. at least people will be able to use ttf for the time being
[01:24:37] <edheldil> the korean bg1 and pst does not have bam, afaics
[01:25:00] <brada> really?
[01:25:34] <brada> did they somehow bolt ttf support on or something?
[01:26:20] <edheldil> I don't know of bg1 - there's just dialog.tlk; but pst definitely has an own Torment.exe
[01:26:42] <brada> interesting
[01:27:55] <edheldil> japanese soa has bam
[01:32:17] <edheldil> although I vaguely remember it should have an exe change as well,since it was possible to switch betweenkatakana and hiragana with F3
[01:33:27] <brada> yeah the chinese required an exe patch too
[01:35:06] <edheldil> those are the asian versions I have
[01:35:14] <edheldil> or patches
[02:28:27] <edheldil> good night
[02:31:53] <-- edheldil has left IRC (Read error: Operation timed out)
[03:20:17] <Seniorita> [commit] Brad Allred: Font: this class is a bit easier to work with without "unsigned char" all over the... http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=473d77767287dadb9753502372fa7978d97d646e
[03:20:18] <Seniorita> [commit] Brad Allred: TTF: begin removal of firstchar and last char from fonts.2da http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=d32a4e681d625f0be9381924c8b6b18c5f2c8a75
[03:20:19] <Seniorita> [commit] Brad Allred: BAMFont: remove limits and just make a glyph for every cycle and frame http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=b69f88fee0eb841b8d445fcbd8cc202c311cd3ee
[03:20:20] <Seniorita> [commit] Brad Allred: Fonts: abadon use of first/last char http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=322d2f6512942f752b85253921d794b74f30546d
[03:20:22] <Seniorita> [commit] Brad Allred: Font: remove unused GetCharacterCount method. http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=5df0b840c0ff040dfed19610d21e943121329223
[03:20:23] <Seniorita> [commit] Brad Allred: Merge branch 'multibyte' http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=2708fb242d00fd4c1531dba499e677ae454388b5
[03:21:24] <brada> not sure how we eventually want to handle the huge gaps in the character sprete array, but im content to leave them for now
[03:22:36] <brada> (not really a problem for standard ext ascii mind you)
[03:23:23] <brada> i at least want to prevent the ttf plugin from generating the entire unicode set :/
[03:23:48] <brada> im leaning towards dealing with that in a ttf subclass that will dynamically fetch and cache glyphs as needed
[03:25:46] <brada> in the short term i can probably request a smaller ascii charmap by deafault from freetype
[03:26:15] <brada> but that may not help if the chinese fonts were any indication
[03:31:05] <brada> im prolly worrying too much about ~25K wasted memory give or take :D
[03:31:28] <brada> well thats just for the emptyness in the array (on 64bit)
[03:31:42] <brada> the unused glyphs will make that even larger
[04:07:51] <Seniorita> [wiki] plugins:ttf - [fonts.2da] http://www.gemrb.org/wiki/doku.php?id=plugins:ttf&rev=1358136396&do=diff
[04:26:05] <Seniorita> [wiki] plugins:ttf - [Support for non-ASCII languages] http://www.gemrb.org/wiki/doku.php?id=plugins:ttf&rev=1358137355&do=diff
[04:46:37] <-- brada has left IRC (Quit: brada)
[05:06:11] --> brada has joined #gemrb
[05:10:54] <-- brada has left IRC (Ping timeout: 256 seconds)
[05:26:08] <-- |Cable| has left IRC (Ping timeout: 255 seconds)
[05:27:18] --> brada has joined #gemrb
[05:40:15] --> |Cable| has joined #gemrb
[06:04:15] <-- brada has left IRC (Quit: brada)
[07:03:58] --> miha has joined #gemrb
[07:22:48] --> edheldil has joined #gemrb
[07:22:48] --- ChanServ gives channel operator status to edheldil
[07:32:34] <-- edheldil has left IRC (Ping timeout: 252 seconds)
[08:34:53] --> edheldil has joined #gemrb
[08:34:54] --- ChanServ gives channel operator status to edheldil
[08:36:00] --> WingedHussar has joined #gemrb
[08:46:03] --> lynxlynxlynx has joined #gemrb
[08:46:03] <-- lynxlynxlynx has left IRC (Changing host)
[08:46:03] --> lynxlynxlynx has joined #gemrb
[08:46:03] --- ChanServ gives channel operator status to lynxlynxlynx
[08:50:32] <edheldil> Good moaning!
[09:05:49] <lynxlynxlynx> oj
[09:06:17] <lynxlynxlynx> brad broke the build, how ironic
[09:17:20] --> thySEus1 has joined #gemrb
[09:20:45] --> chiv has joined #gemrb
[09:24:18] --> chiv_ has joined #gemrb
[09:25:03] <-- chiv has left IRC (Ping timeout: 245 seconds)
[09:25:39] <chiv_> heh, i was about the checkout the chinese support, but it just came out with errors
[09:29:37] <fuzzie> that is a silly commit :-p
[09:31:11] <fuzzie> lynxlynxlynx: you want me to commit a fix or is it already handled?
[09:32:46] <lynxlynxlynx> go ahead, i'm working working
[09:33:30] <lynxlynxlynx> i'll need to pick your brains later for some scripting stuff
[09:34:06] <Seniorita> [commit] Alyssa Milburn: Font: partially revert 473d7776 to fix the build http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=6a8535548ee2e2332e85462d9a56ef99fa335bf2
[09:34:14] <lynxlynxlynx> gembot: unmute
[09:34:15] <gembot> You hadn't told me to be quiet, but it's the thought that counts, right?
[09:34:30] <lynxlynxlynx> hmm, why didn't you complain then?
[09:35:12] <chiv_> was it a warning instead of an error?
[09:35:21] <fuzzie> only for one build
[09:35:40] <fuzzie> oh, no, for all builds
[09:36:14] <fuzzie> we force warnings to be errors by default, but apparently not on buildbot for some reason
[09:36:54] <lynxlynxlynx> plenty of them failed, are you sure?
[09:37:07] <fuzzie> they failed because brada broke the BAM fonts
[09:37:14] <fuzzie> which is different to failed compiles I guess?
[09:37:19] <lynxlynxlynx> ah indeed, the minimal test
[09:38:59] <chiv_> im really just curious to see what happens when I prod it with the chinese input method...
[09:39:08] <chiv_> so i can be patient...
[09:39:24] <fuzzie> (specifically numeric fonts are broken)
[09:39:44] <fuzzie> presumably there's a useless cycle in the in-game ones which stops it from crashing
[09:39:58] <lynxlynxlynx> i can't even get ttf ones to work
[09:40:07] <fuzzie> well, I've never tried ttf ones
[09:40:23] <lynxlynxlynx> crashes in getfont
[09:40:29] <fuzzie> in the ttf one?
[09:40:33] <lynxlynxlynx> but argh, i should go work
[09:40:35] <fuzzie> bg2 works fine for me, with trunk.
[09:40:40] <lynxlynxlynx> yeah, gamedemo
[09:40:48] <lynxlynxlynx> with the extra cols removed
[09:40:50] <fuzzie> we can fix it later by means of despairing at brad.
[09:40:56] <lynxlynxlynx> yep
[09:42:18] <Seniorita> [commit] Jaka Kranjc: demo: removed now obsolete fonts.2da columns http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=80263ae62540bccf90f9c09773546076bf8ba757
[09:42:41] <edheldil> chiv_: have you read my night comment? You have broken PS:T
[09:43:28] <chiv_> i missed that one... what is broken?
[09:44:03] <edheldil> edheldil> hi all, there's a python bug in the current master when running pst: http://paste.debian.net/224404/
[09:44:05] <Seniorita> debian Pastezone
[09:44:23] <fuzzie> edheldil: and that isn't from an install, right?
[09:45:08] --> edheldil_ has joined #gemrb
[09:45:22] <lynxlynxlynx> and is it really the first error?
[09:45:22] <chiv_> thats odd, i didnt even touch messagewindow.py
[09:46:04] <chiv_> can you tell me when it happens? ill find out whats wrong
[09:47:08] <edheldil> Ah, sorry, deleteing *.pyc files fixed it
[09:47:17] <fuzzie> bug fixed \o/
[09:47:17] <chiv_> yay
[09:47:29] <lynxlynxlynx> how did you get them at all? we disabled their generation ages ago
[09:47:54] <chiv_> good show everyone, beers and cheeses all round
[09:48:01] <fuzzie> the disabling code only works on python 2.6+
[09:48:12] <fuzzie> so possibly edheldil has old python?
[09:48:29] <edheldil> 2.6.5
[09:49:20] <edheldil> hm, they had to be there for some time, because they dd not appear again, but I thought I had to delete them already
[09:49:29] <edheldil> apparently not
[09:50:27] <fuzzie> oh, sorry, to be clear about what I said a few minutes ago: bg2 works fine for me in trunk, except, numeric fonts are broken :)
[09:50:36] <edheldil> I apologize ... But I still think that str(Drinks['Price']) is ugly :-D
[09:51:52] <fuzzie> also the icon fonts are all broken
[09:52:39] <fuzzie> and capitals ;p
[09:53:23] <chiv_> well, for me its a choice between that or 50 if/else's...
[09:54:28] <chiv_> unless you just mean you don't like the dict's ?
[09:55:32] <fuzzie> um
[09:55:33] <fuzzie> so
[09:56:01] <fuzzie> if I remove an unmemorised spell then it removes the memorised variant instead
[09:56:06] <fuzzie> that is quite inconvenient
[09:58:09] <fuzzie> i mean, unmemorised as in depleted, memorised as in undepleted
[09:58:29] <chiv_> i would like to say something else about that area... is there any way that it would be possible to not change the max-count when you get level drained/swap a ring of wizardry etc, and just expend the spells instead?
[09:59:34] <chiv_> i really hated that part of bg2....
[09:59:42] <chiv_> stupid vampires...
[09:59:51] <wjp> or at least remember the spell selection for when the lost levels get restored...
[10:00:12] <chiv_> well, that is what i meant to say, just phrased better :)
[10:01:27] <fuzzie> ah
[10:01:45] <fuzzie> lynxlynxlynx: so you changed UnmemorizeSpell to take a spellresref, not a spell, and so you delete the wrong indexes :)
[10:02:09] <fuzzie> oh, no, that can't be it
[10:02:14] <fuzzie> phft
[10:02:19] <fuzzie> ah, I see
[10:02:21] <Gekz> type safety says what?
[10:02:22] <Gekz> :P
[10:02:31] <fuzzie> so you changed the UnmemorizeSpell call to call a different function.
[10:02:42] <fuzzie> one which takes a spellresref and not a spell.
[10:02:45] <fuzzie> and so you delete the wrong indexes.
[10:03:09] <fuzzie> I'm not sure how to fix that, I'm not quite sure what lynx was trying to do here.
[10:03:53] <lynxlynxlynx> i fixed the problem you're now having for iwd2
[10:04:13] <fuzzie> yes, I can see this is the commit where you broke it
[10:04:26] <fuzzie> I guess iwd2 doesn't have individual slots?
[10:04:34] <lynxlynxlynx> you had let's say 1/2 of cure light wounds - clicking on it would get you to 0/1 instead 1/1
[10:04:36] <fuzzie> bg2 has individual slots, so you can't do it by resref.
[10:05:07] <fuzzie> Since if I have both depleted and undepleted spells in two slots, I can click on either slot.
[10:05:07] <lynxlynxlynx> ok, just ifdef the call in gemrb_unmemorizespell, we still have all the needed data
[10:05:20] <lynxlynxlynx> ah
[10:05:24] <fuzzie> you think it's okay to key it just on the onlydepleted flag?
[10:05:45] <lynxlynxlynx> sure
[10:06:59] <fuzzie> thanks
[10:08:35] <fuzzie> um, how do I update docs?
[10:10:47] <fuzzie> if I run add_missing_gui_docs.sh then it adds a whole bunch of things but doesn't update the one I changed
[10:11:33] <lynxlynxlynx> its name is true
[10:11:35] <fuzzie> have to update manually?
[10:11:37] <lynxlynxlynx> yep
[10:13:37] <fuzzie> hopefully that isn't terrible
[10:13:58] <fuzzie> I really have to get some study done now though
[10:14:13] <Seniorita> [commit] Alyssa Milburn: GUIScript: Fix non-iwd2 case of UnmemorizeSpell. http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=4dd5c7c0b8bd114837aa7a1be93d1c3a6c64859f
[10:34:26] <-- thySEus1 has left IRC (Ping timeout: 252 seconds)
[11:27:16] <Seniorita> [commit] Jaka Kranjc: sorcerer is spelled sorcerer http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=1b98e2b7e70a8ce8aab2d82ef2ae9f1b270215c7
[11:30:47] <chiv_> hmm, pcmanfm... the file manager that could have been written by someone who stole my brain and read its contents
[12:02:53] <lynxlynxlynx> i need to write another limerick, maybe it'll bring mattinm back :D
[12:03:41] <lynxlynxlynx> all this boring levelup gui stuff
[12:04:17] <chiv_> how well do you know 3e rules?
[12:06:02] <wjp> what do you want to know?
[12:06:58] <chiv_> nothing really, just wondered how much of a headache it was to write them ...
[12:07:19] <chiv_> I think I used to know alot, but that is all gone now
[12:17:46] <chiv_> I think the irony is, they streamlined it so much that you barely had to read the book
[13:03:20] --> kida has joined #gemrb
[13:10:00] --> traveler__ has joined #gemrb
[13:19:39] <Seniorita> [wiki] todo - [Shared] http://www.gemrb.org/wiki/doku.php?id=todo&rev=1358169475&do=diff
[13:20:37] <lynxlynxlynx> traveler__: you mean you went there for the first time and it was empty?
[13:20:40] <traveler__> guys, bad news
[13:20:49] <lynxlynxlynx> otherwise it is the same item as the one before it
[13:20:53] <traveler__> free action is the one of blessing (white cross)
[13:20:59] <traveler__> *icon
[13:21:06] <traveler__> no counters on stacked items
[13:21:19] <lynxlynxlynx> yeah, brad broke a lot this morning
[13:21:22] <traveler__> fresh git
[13:21:41] <lynxlynxlynx> come later for a yelling session >>
[13:21:48] <traveler__> speaking of wiki, i went there several times (repeated ambushes) each time epmty
[13:21:50] <traveler__> haha
[13:21:58] <fuzzie> he broke fonts
[13:22:13] <fuzzie> so the number and icon fonts are pretty crazy, as you've discovered
[13:22:16] <lynxlynxlynx> like i said, only the first time matters or it is the same as we already have on the todo
[13:22:35] <traveler__> first time was empty too
[13:22:44] <traveler__> no corpses etc
[13:22:55] <traveler__> blank area, i've made dump to be sure
[13:23:11] <fuzzie> probably it's the same as what we already have on the todo anyway
[13:23:59] <wjp> I'm guessing removing this firstChar is breaking everything?
[13:25:09] <fuzzie> wjp: well, the crash on the demo data is presumably when cycles==0
[13:25:55] <fuzzie> so I'm not sure if it's more complicated than the obvious, so haven't looked into it
[13:27:19] <traveler__> you mean that if repeated ambushes each time enemies should be revived? and if blank filled? it's strange, since usually if there was empty ambush, corpses would be there, meaning i've cleared it earlier, with this area it was clean. in bg1 corpses in ambushes stay a long way, i don't know if they disappear at all.
[13:27:38] <fuzzie> traveler__: ambush areas should always be deleted after you leave them.
[13:28:00] <fuzzie> so you get a new copy every time.
[13:28:19] <traveler__> well, they are not for sure.
[13:28:32] <traveler__> by bg1 i've mean gemrb/bg1 there
[13:28:40] <fuzzie> yes, gemrb is broken there
[13:28:57] <fuzzie> wjp: it confuses me because brad added the whole firstchar=='0' logic in the first place
[13:29:51] <traveler__> what i'm trying to say, is that this area looks like it had never enemies, because they would left corpses at least (data bug?)
[13:30:10] <fuzzie> traveler__: It's presumably a failing spawn.
[13:30:14] <fuzzie> Unless someone fixed bg1 spawns.
[13:31:52] <fuzzie> wjp: ah, nm, there used to be as SetFirstChar method. so yes it's presumably firstChar.
[13:31:53] <lynxlynxlynx> fuzzie: in bg2 the same ambush only happens once, you can't repeat them
[13:32:15] <fuzzie> lynxlynxlynx: well I have no idea about bg2 :)
[13:32:46] <fuzzie> but the theory is that the ambush areas don't get saved
[13:32:46] <traveler__> mhm, with clean never visited before save this area should have huge spiders
[13:33:08] <fuzzie> so it's re-loaded anew from the ARE the next time you visit it
[13:33:13] <traveler__> i've tried with save that has none areas visited and mta to 6001
[13:33:25] <fuzzie> monsters being present or not is a different issue, but I'm usually happy to blame spawns in bg1
[13:33:34] <lynxlynxlynx> possible, but then the ambush chance would have to be zeroed in bg2
[13:33:49] <fuzzie> yes, I think I wrote the code for that in bg2
[13:33:51] <fuzzie> let me look
[13:33:51] <lynxlynxlynx> possibly for multiple areas
[13:35:50] <wjp> fuzzie: the case cyclecount <= 1 seems to have completely different behaviour now
[13:36:12] <wjp> fuzzie: including presumably this crash for == 0 you mentioned
[13:36:35] <fuzzie> yes, you can see the ==0 crash easily, since it tries using glyphs[0] without actually having created any glyphs
[13:36:46] * wjp nods
[13:37:06] <fuzzie> but I figured I'd leave it and see what brad says
[13:37:29] <fuzzie> we can always fix it if brad has problems or is too frustrated or whtever :)
[13:38:14] <fuzzie> lynxlynxlynx: hm I guess I didn't write any relevant code for that. let me think
[13:39:03] <lynxlynxlynx> if you have time, i'd rather divert you to iwd2 - i have two scripting problems there
[13:40:59] <fuzzie> well, I am avoiding studying, so sure
[13:41:40] <chiv_> this is studying.. its just not studying what you are meant to be studying :)
[13:44:11] <lynxlynxlynx> http://paste.debian.net/224559/
[13:44:13] <Seniorita> debian Pastezone
[13:44:27] <fuzzie> as long as it's not change ai script
[13:45:05] <lynxlynxlynx> nope
[13:45:16] <fuzzie> oh great, iwd2 script behaviour
[13:45:50] <lynxlynxlynx> the other one is likely simpler
[13:46:03] <lynxlynxlynx> at least it's all about gemrb
[13:46:47] <lynxlynxlynx> this cutscene was slow in the original too, just don't remember if so much (but definitely on the annoying side)
[13:47:18] <fuzzie> the cutscene thing makes me wonder if you want GF_CUTSCENE_AREASCRIPTS but it's also possible that new areas always get a chance for new scripts..
[13:48:06] <lynxlynxlynx> good point, i'll try with that
[13:48:20] <fuzzie> (the OnCreation block is almost certainly a red herring, it's present in all area scripts)
[13:48:26] <lynxlynxlynx> ok, next is the Nothing trigger
[13:48:49] <lynxlynxlynx> it's at 255 so i had to increase MAX_OBJECTS to not skip it
[13:48:58] <lynxlynxlynx> but that was not enough
[13:49:14] <lynxlynxlynx> we do have an implementation for it, but it is not reached
[13:49:14] <fuzzie> what's the script line?
[13:49:36] <lynxlynxlynx> ItemIsIdentified("10Vellum", Nothing)
[13:50:03] <lynxlynxlynx> we return NULL near that hack for attack([0]), since no filters applied (all the ids fields are 0)
[13:50:23] <fuzzie> it's in a script or dialog?
[13:50:29] <lynxlynxlynx> so no targets, so no item
[13:50:31] <lynxlynxlynx> dialog
[13:50:33] <fuzzie> ah
[13:50:34] <fuzzie> :)
[13:51:06] <lynxlynxlynx> scripts mostly use SetSpellTarget(Nothing)
[13:51:34] <fuzzie> my first guess would be that we're simply parsing it wrong
[13:51:44] <fuzzie> is that likely, if all the ids fields are 0?
[13:53:10] <lynxlynxlynx> parse where, it has no parameters?
[13:53:26] <fuzzie> dialogs are all text
[13:53:26] <lynxlynxlynx> and yep, the gameflag did it
[13:53:31] <fuzzie> we parse them from the text
[13:53:56] <fuzzie> in any case, Nothing is an object filter, so MAX_OBJECTS shouldn't be involved..
[13:54:07] <lynxlynxlynx> sure, in ItemIsIdentified, dump of parameters did show the first filter to be 255
[13:54:30] <lynxlynxlynx> we ignore it on load, since 255 > 128
[13:54:50] <fuzzie> that is a silly bug
[13:55:08] <fuzzie> we're using MAX_OBJECTS there where we shouldn't
[13:55:37] <fuzzie> oh, no
[13:55:44] <fuzzie> other way around, ok
[13:56:05] <fuzzie> let me work out what I'm talking about :)
[13:56:57] <lynxlynxlynx> :)
[14:03:05] <fuzzie> right, so it's Nothing rather than something like Nothing([0]), so it has no ids entry, so it ends up at DoObjectFiltering..
[14:03:15] <fuzzie> so you should get a call
[14:04:27] <fuzzie> you changed MAX_OBJECTS to 256?
[14:05:09] <lynxlynxlynx> locally
[14:05:24] <fuzzie> and you no longer get a complaint at startup about Nothing?
[14:05:51] <lynxlynxlynx> yes as in no, also not later when parsinng
[14:06:56] <fuzzie> I was expecting it to be squeezed into a char somewhere (where 255 <-> -1) but I don't see any problems.
[14:12:50] <fuzzie> so: the EvaluateObject returning NULL doesn't matter, this case is handled by the DoObjectFiltering later in GetAllObjects.
[14:12:54] <fuzzie> do you have a save?
[14:13:14] <lynxlynxlynx> sure
[14:17:18] <lynxlynxlynx> it wasn't handled later though, i can give you the exact path if you think that's easier
[14:17:26] <fuzzie> sure
[14:17:43] <lynxlynxlynx> after that null return everything just backtracked and no object was found
[14:19:09] <fuzzie> which is weird, because that should only happen if objectName is valid
[14:19:45] <fuzzie> since that only happens if Matching.cpp:235 matches, i.e. if EvaluateObject returns NULL *and* objectName[0] is set
[14:20:41] <lynxlynxlynx> tgts->Count() is zero, so it looks like the trigger itself is buggy
[14:21:16] <fuzzie> you mean, ItemIsIdentified?
[14:21:27] <lynxlynxlynx> i guess i was looking at this before the max_object change, since then it bugged out at Unknown object filter:
[14:21:28] <lynxlynxlynx> no
[14:21:35] <lynxlynxlynx> doobjectfiltering -> Nothing
[14:21:41] <fuzzie> I mean, I would expect the Nothing filter to get called, and then I'd expect ItemIsIdentified to fail immediately due to lack of actor.
[14:22:10] <lynxlynxlynx> parameters->Clear();
[14:22:10] <lynxlynxlynx> return parameters;
[14:22:11] <lynxlynxlynx> haha
[14:22:28] <lynxlynxlynx> so yeah
[14:23:07] <lynxlynxlynx> is it supposed to return the area-ful of actors?
[14:23:12] <fuzzie> no
[14:23:17] <fuzzie> it's meant to return nothing :)
[14:23:24] <lynxlynxlynx> wierd use
[14:23:30] <fuzzie> yes, it makes utterly no sense
[14:23:40] <fuzzie> the MAX_OBJECTS change sounds like a good one though
[14:25:03] <fuzzie> all the uses in the scripts seem legitimate
[14:25:04] <lynxlynxlynx> yeh
[14:26:43] <lynxlynxlynx> all the dialog uses seem to be with this item function
[14:27:32] <fuzzie> is it possible that it's simply a no-op?
[14:28:20] <lynxlynxlynx> mhm
[14:28:34] <lynxlynxlynx> i supposed if there was something else, it would only check that object's inventory
[14:30:35] <lynxlynxlynx> yep
[14:30:37] <lynxlynxlynx> ItemIsIdentified("00SwdL07",Protagonist)
[14:31:19] <lynxlynxlynx> and Player1 etc
[14:31:44] <lynxlynxlynx> and that's it
[14:32:08] --> Cuvieronius has joined #gemrb
[14:32:22] <lynxlynxlynx> i should just hardcode this nothing exception in the action, right?
[14:33:36] <fuzzie> do you need to?
[14:34:00] <fuzzie> sorry, I don't know what your original problem was, can't the trigger just return false?
[14:36:36] <lynxlynxlynx> the item is identified, so no, it needs to return true now┼ż
[14:37:22] <-- fireglow has left IRC (Remote host closed the connection)
[14:37:22] <fuzzie> and it's the inventory of Sender?
[14:37:31] <lynxlynxlynx> nope
[14:37:39] <lynxlynxlynx> that's the odd bit
[14:37:53] <lynxlynxlynx> in this case any party member should do
[14:38:27] <fuzzie> hm
[14:38:32] <fuzzie> I wouldn't dare to try and work it out.
[14:39:00] <lynxlynxlynx> i would just replace it with a party filter before fetching the object (iff filter is 255)
[14:39:20] <lynxlynxlynx> good first approx
[14:41:07] <lynxlynxlynx> 19 LastTalkedToBy would be simpler
[14:43:07] --> fireglow has joined #gemrb
[14:44:24] <lynxlynxlynx> worked like a charm
[14:45:04] <lynxlynxlynx> using [PC] would be more player friendly though
[14:45:29] <lynxlynxlynx> doesn't matter, it only checks one inventory
[14:50:28] <-- chiv_ has left IRC (Ping timeout: 245 seconds)
[14:50:37] <traveler__> *it is very offtopic, but if someone has problem with hanging pmp which is supported by rockbox.org, i recommend it over original firmware, as it turns out it was the culprit in my case (sansa clip+)
[14:50:43] <lynxlynxlynx> http://ebm.si/arhiv/delme1.png :D
[14:52:19] <traveler__> still in targos :P ?
[14:52:22] <fuzzie> o.O
[14:53:06] <-- miha has left IRC (Quit: Lost terminal)
[15:03:58] <lynxlynxlynx> ok, next bug
[15:04:05] <lynxlynxlynx> unrelated :)
[15:09:45] <lynxlynxlynx> a locked door with difficulty of 100
[15:10:47] <traveler__> are you sure that all doors were meant to be unlockable?
[15:11:53] <lynxlynxlynx> they weren't
[15:12:16] <lynxlynxlynx> this one in particular looks like a script or the dialog should take care of it, but there's no mention
[15:12:31] <traveler__> *shut up already*
[15:12:34] <traveler__> *s
[15:12:50] <lynxlynxlynx> "You can't reach the shield on this rack while this moon-eyed soldier is standing in front of it. " fail string
[15:13:40] --> miha has joined #gemrb
[15:14:06] <lynxlynxlynx> the dude was sent away, so it's odd
[15:16:21] <fuzzie> iwd2's scripting is really crazily different in places
[15:16:33] <fuzzie> but I have no idea about how much it's different otherwise
[15:17:53] <lynxlynxlynx> doesn't look scripting related at all
[15:18:29] <lynxlynxlynx> the "door" was trapped though and i could remove that, but of course that didn't change anything
[15:19:12] <lynxlynxlynx> oh, maybe it is
[15:19:39] <lynxlynxlynx> SetDoorFlag(Myself,HIDDEN,TRUE) <-- maybe this isn't about our hidden doors
[15:21:07] <Seniorita> [commit] Jaka Kranjc: iwd2: run area scripts during cutscenes, fixes Swift Thomas from Targos http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=2bb0a9963eb560311bf4e31f52df4494cf3034f2
[15:21:09] <Seniorita> [commit] Jaka Kranjc: GemRB_CheckVar: print the value in the DEBUG context, so it stands out http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=ff609e922efc38b1ce074852094475fbc25be75f
[15:21:10] <Seniorita> [commit] Jaka Kranjc: bump MAX_OBJECTS so iwd2's Nothing is reckognized (255) http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=8778d10541f501009e17ef85dc28d57d816f67ae
[15:21:11] <Seniorita> [commit] Jaka Kranjc: ItemIsIdentified: handle the Nothing filter specially or lots of iwd2 uses fail http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=f0d491ad3fb4ae294bc8ae754f0331598566e4ca
[15:21:26] <lynxlynxlynx> looks like it
[15:22:30] <lynxlynxlynx> we don't handle 0x2000 at all yet
[15:29:59] <lynxlynxlynx> four more flags of which one looks like a duplicate
[15:40:59] --> CamDawg has joined #gemrb
[15:41:48] <Seniorita> [commit] Jaka Kranjc: iwd2: fixed ar1200 shield rack not being reachable after sending Tabard away http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=56dc6ec0587f9dd666d3acf96ac9a5d2cf1a1812
[15:47:51] <-- miha has left IRC (Ping timeout: 276 seconds)
[15:51:22] --> thySEus1 has joined #gemrb
[15:54:18] --> miha has joined #gemrb
[15:57:41] <-- thySEus1 has left IRC (Remote host closed the connection)
[16:05:05] <-- edheldil_ has left IRC (Read error: Operation timed out)
[16:09:44] <-- WingedHussar has left IRC (Quit: WingedHussar)
[16:19:12] --> brada has joined #gemrb
[16:19:52] <brada> i neglected to update fonts 2da for demo
[16:20:17] <brada> i wonder why i didnt get a warning for that type change
[16:20:38] <brada> 9yes i did check)
[16:21:16] <fuzzie> all the non-plaintext fonts being broken is our only concern now, though :)
[16:21:35] <brada> ok i didnt notice that
[16:21:44] <fuzzie> numeric, icons and capitals
[16:22:42] <fuzzie> plus the demo crashes, but it's just because the bam has no cycles and you don't check that
[16:23:06] <brada> yeah im unfortunately ignorant on bam font layout
[16:23:21] <brada> with since when does the demo use a bam font?
[16:23:48] <fuzzie> not the demo, sorry. the test.
[16:23:51] <brada> ah
[16:24:09] <brada> i didnt think it was possible to have a bam without a cycle...
[16:24:16] <brada> how does that make sense?
[16:24:19] <fuzzie> Afaik the theory is that (a) for cycles <=1 you should use frames, not cycles and (b) you need the firstchar thing for numeric fonts.
[16:24:47] <fuzzie> But I thought I'd better not touch it before you were around.
[16:25:07] <brada> yup i knew about how to fix numeric fonts
[16:25:16] <brada> i actually thought about that as i was going to bed :p
[16:25:21] <fuzzie> are capitals likely to be the same problem? and icons?
[16:25:56] <brada> im unsure there. i can think of why caps would be diffrent from standard text
[16:26:00] <brada> cant
[16:28:16] <fuzzie> afaict it's just cosmetic for the games anyway (just bad glyphs), so no hurry
[16:29:06] <brada> ok
[16:29:30] <brada> but now decicions need to be made so that i can merge the actual multibyte stuff
[16:29:41] <brada> that was just a primer
[16:33:30] <Seniorita> [commit] Jaka Kranjc: Map::AnyEnemyNearPoint: ignore actors with avatar removal http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=d3c5865ab1ca7ad647d82fc011e1be40d98b4739
[16:35:59] <brada> ok i see what you mean about broken caps and states
[16:36:06] <brada> jsut offset by 1 it seems
[16:36:40] <fuzzie> the caps didn't work for me at all
[16:36:50] <brada> really?
[16:36:54] <brada> in what game?
[16:36:56] <fuzzie> bg2
[16:37:04] <brada> thats what im using...
[16:37:07] <fuzzie> hm
[16:37:13] <fuzzie> ok, now they all appear when I restart it
[16:37:14] <fuzzie> that is worrying
[16:37:25] <traveler__> re fuzzie "numeric, icons and capitals": i might be seeing things but i think that problems with weapon animation (lack of it) resurfaced (no weapon for hobgoblins, gnolls etc) i will recheck now
[16:37:26] <brada> i would have notice if i werent in chinese lol
[16:37:55] <fuzzie> brada: ok never mind me then.
[16:37:59] <fuzzie> i don't know what the heck I had earlier.
[16:38:08] <brada> ok
[16:38:13] <brada> well still need to be fixed
[16:38:15] <fuzzie> sorry for the noise.
[16:38:20] <brada> dont be
[16:38:45] <brada> i need more noise from you regarding what i should do to finish the multibyte stuff
[16:39:52] <fuzzie> what's mystrlen?
[16:40:00] <traveler__> +1 for seeing things, gnolls still have their weapons
[16:40:20] <fuzzie> (it's in core/Font.cpp but I don't see any users..)
[16:40:44] <fuzzie> hm, dates back to times unknown maybe
[16:41:30] <brada> ha
[16:41:32] <fuzzie> and, well, I don't know what to do about the non-GBK case
[16:41:36] <brada> i was wondering that yesterday too
[16:41:37] <fuzzie> lots of options
[16:41:47] <brada> yeah but none sounded fun :(
[16:44:36] <fuzzie> so I quite like the idea of just coding the GBK case for now (i.e. only modifying the currChar code)
[16:44:42] <fuzzie> but I have no time to try it myself today
[16:44:47] <lynxlynxlynx> brada: also better to lookup fonts.2da values by column names
[16:44:59] <lynxlynxlynx> that way changes like this don't break existing files like now
[16:45:06] <brada> yes i coded that before i knew about that then forgot
[16:45:17] <lynxlynxlynx> any user out of our control is screwed
[16:45:28] <brada> yeah i agree
[16:45:49] <lynxlynxlynx> and why did you push this in the first place, if you knew it was not finished?
[16:46:07] <lynxlynxlynx> introducing-regressions-not-finished
[16:46:29] <brada> i didnt notice the problem with the caps and states being offset
[16:47:03] <lynxlynxlynx> and it built for you?
[16:47:07] <brada> yes
[16:47:10] <brada> no warning at all
[16:47:26] <brada> the min test has never worked for me
[16:47:28] <lynxlynxlynx> please do more testing, there's a pattern already :(
[16:48:06] <brada> how? like i said min test has never worked for me
[16:48:11] <fuzzie> if you could work out what's broken with min test, would appreciate it
[16:48:29] <brada> ok
[16:49:17] <lynxlynxlynx> well test with games, minimal is only for the init
[16:49:17] <brada> what warning would i need to turn on to get the warning that broke things?
[16:49:25] <brada> i did test with the games
[16:49:40] <lynxlynxlynx> i said more :)
[16:49:44] <brada> unfortunately most testing was done in chinese :p
[16:49:46] <fuzzie> brada: -Wformat.
[16:49:48] <brada> yeah i will
[16:58:00] <brada> ok any other w flags i should ensure are enabled?
[17:01:14] <lynxlynxlynx> this was unavoidable, it is just on by default in gcc
[17:01:55] <brada> i dont have gcc
[17:05:02] <brada> btw min test being broken is just due to the paths in the cfg iirc
[17:06:42] <traveler__> http://img5.imageshack.us/img5/7630/201301141803361440x900s.png ar1900 very subtle, but travel area /border looks out of place
[17:06:55] <traveler__> few pixels wide long strip
[17:07:41] <traveler__> this is very border of map
[17:15:17] <-- duckpunch has left IRC (Ping timeout: 246 seconds)
[17:28:34] --> duckpunch has joined #gemrb
[17:28:45] <-- Seniorita has left IRC (Remote host closed the connection)
[17:29:24] --> Seniorita has joined #gemrb
[17:37:39] --> edheldil_ has joined #gemrb
[17:38:29] --> chiv has joined #gemrb
[17:38:47] <chiv> ni men hao
[17:39:30] <chiv> would anyone like to point me in the direction of where things like clickable infopoints are handled so i can try and get the broken torment quests working, or is somebody already on it?
[17:39:51] <lynxlynxlynx> infopoint.cpp among others
[17:40:32] <chiv> ok, ill have a look
[18:10:20] <fuzzie> how're they broken?
[18:11:45] <chiv> you just cant get close enough to any of them
[18:12:32] <chiv> or perhaps you can, but they jump to the 'too far away to use' conclusion for some reason
[18:13:04] <chiv> ar1800 is a good example
[18:13:24] <-- brada has left IRC (Quit: brada)
[18:19:21] <chiv> a disproportionate amount of the game is uncompletable because of various trigger points not working, eg glaive and the decanter of water for ignus, exposing soego to escape imprisonment of the silent king, no sensory stones work, you cant get the stygian things to give to merriman in order to work towards finding ravel, you cant do any of the fortress of regret...etc... so it might be worth focusing on :)
[18:20:19] <chiv> oh yeah, you cant get into the tenements either...
[18:21:10] <chiv> they all say 'you are too far away to use this' or somesuch
[18:21:15] <-- |Cable| has left IRC (Ping timeout: 276 seconds)
[18:23:17] <chiv> oh yeah, it would stop you from leaving baator as well... if I had managed to load that area and test...
[18:23:48] <chiv> its like one thing that breaks the whole game, everything else even if it is a bit rough, is still mostly playable
[18:24:43] <lynxlynxlynx> iwd2 prologue successfully completed
[18:25:04] <lynxlynxlynx> apart from the inventory bugs and no levelup, there were no annoyances
[18:25:08] <chiv> *standing ovation*
[18:26:41] <lynxlynxlynx> starting chapter 1 immediately exposes a bug
[18:27:09] <lynxlynxlynx> textscreen loads and a dialog starts, not sure which first, and the gui gets messed up
[18:29:58] <lynxlynxlynx> looks like it was the dialog first, since textscreen does force a pause
[18:34:05] --> |Cable| has joined #gemrb
[18:37:21] --> Yoshimo has joined #gemrb
[18:43:34] --> brada has joined #gemrb
[18:59:23] <-- brada has left IRC (Read error: Connection reset by peer)
[19:09:31] --> brada has joined #gemrb
[19:28:22] <edheldil_> chiv: I have baator in save games at my site
[19:29:35] <chiv> hmm, have you tried it lately?
[19:29:54] <chiv> mta("AR1000") just crashes instantly for me
[19:30:38] <-- Yoshimo has left IRC (Read error: Connection reset by peer)
[19:31:00] <chiv> but if it worked at one point, i could bisect...
[19:34:11] <lynxlynxlynx> nah, not much sense in that
[19:34:20] <lynxlynxlynx> a backtrace is what would be useful
[19:35:43] <chiv> thats a point, I should do that
[19:36:07] <chiv> I cant really find anything about these infopoints anyway...
[19:42:01] <traveler__> re chiv 'or perhaps you can, but they jump to the 'too far away to use' conclusion for some reason': that's what i have with durlag's tower levers in bg1, nothing happens until you went away, and then they display 'too far away' (clicking was definitely near them)
[19:43:08] <chiv> hmm, in torment there is a slight difference in that the text appears instantly, but that might be the difference with the floating text
[19:43:53] <traveler__> ah
[19:44:09] <traveler__> back under the rock for me, then
[19:44:27] <chiv> well, its probably the same thing, just different symptoms
[19:44:36] <-- miha has left IRC (Ping timeout: 240 seconds)
[19:46:45] --> miha has joined #gemrb
[19:47:49] <traveler__> may be, but remember that most infopoints works in bg1 while the opposite is true for pst, i believe?
[19:47:59] <traveler__> just those durlag ones are giving me problems
[19:48:12] <edheldil_> chiv: ah, ar1000 crashed, but ar1001 works
[19:48:21] <chiv> yup
[19:48:26] <traveler__> well i hope so, i've not tested any other in a while
[19:48:34] <chiv> its the same bug i found before i think
[19:48:43] <chiv> i have a trace, 2 secs
[19:48:58] <lynxlynxlynx> you described a simple range problem, not sure why that wouldn't be accurate
[19:49:31] <lynxlynxlynx> not sure / no reason
[19:50:04] <chiv> http://pastebin.ca/2302103
[19:50:06] <Seniorita> pastebin - AR1000 - post number 2302103
[19:50:34] <edheldil_> chiv: but loading ar1000 does not crash
[19:50:54] <chiv> thats true, just something instantly after load crashes
[19:53:00] <-- miha has left IRC (Ping timeout: 264 seconds)
[19:53:53] --> brada_ has joined #gemrb
[19:53:53] <-- brada has left IRC (Read error: Connection reset by peer)
[19:53:54] --- brada_ is now known as brada
[19:54:22] --> miha has joined #gemrb
[19:54:29] <wjp> right; confirmed that BG1 behaves like BG2 areaanim-height-wise, and IWD2 like IWD1
[19:54:41] <wjp> haven't found a good anim to test with in PST
[19:55:10] <chiv> what kind of anim? pharods court is loaded with glowing stuff
[19:55:46] <chiv> some houses have glowing stoves
[19:56:02] <chiv> the art gallery has an animated piece...
[19:57:01] <chiv> can it be any of them or do you need something specific?
[19:57:10] <wjp> hm, one you can stand behind would be easiest
[19:57:21] <lynxlynxlynx> did you notice background hiding actors anywhere they shouldn't
[19:57:22] <wjp> or alternatively one you can stand on
[19:57:27] <chiv> ok, 2 secs
[19:57:47] <lynxlynxlynx> re bt, seen that before, i guess if you print actions it is bad?
[19:57:59] <lynxlynxlynx> a run under valgrind may be more useful
[19:58:42] <lynxlynxlynx> valgrind --track-origins=yes --suppressions=../testing/python.supp gemrb/gemrb -c pst.cfg &> val.log # if running from the build dir
[19:59:08] <chiv> for one you can stand on, AR0110, AR0608 for the gallery
[19:59:41] <wjp> which one in AR0110? I looked around there earlier
[20:00:14] <wjp> let's see if I can locate a savegame without rebooting back to linux
[20:02:16] <chiv> wjp: sorry, i thought you were just looking for places with anims, I don't actually know what you are testing...
[20:02:37] <wjp> well, anims you can stand on :-)
[20:03:13] <wjp> such as the floor glow caused by fireplaces in iwd1/2
[20:03:44] <chiv> i could have sword I saw one earlier, now i cant find it
[20:04:29] <chiv> do the big purple glowing pits not count?
[20:07:33] <chiv> I am actually looking at the area list, the game is still quite fresh in my memory, and I can't think of a single underfoot animation like the iwd fireplaces
[20:08:51] <wjp> ones that you can stand behind would also work
[20:09:03] <wjp> such as butterflies in the BG1 starting map :-)
[20:09:52] <wjp> if they don't exist that's even easier, because then it doesn't matter what we do :-)
[20:10:04] <chiv> there is the giant glowing gem in the trial of impulse, AR1202, but that is really messed up...
[20:11:11] <chiv> grr... annoyingly, it is the only damn infopoint in the game that seems to work as well...
[20:11:14] <wjp> ah, there are some glowing bits in Pharod's court between tiles that you can stand on
[20:11:21] <wjp> 'really messed up'?
[20:11:39] <chiv> i thinks its probably best to look...
[20:11:39] <-- brada has left IRC (Read error: Connection reset by peer)
[20:12:15] --> brada has joined #gemrb
[20:13:55] <wjp> oh, dltcep crashes on this savegame from edheldil
[20:15:16] <edheldil_> which one?
[20:15:39] <wjp> technically speaking one I made myself after walking around a little bit
[20:16:02] <wjp> although mixing languages like this probably isn't a good idea
[20:16:21] <chiv> pourquoi pas?
[20:24:00] <wjp> hm, 1202 is interesting
[20:25:10] <wjp> missing alpha most likely
[20:25:21] <wjp> and actors should probably be behind the smoke
[20:27:41] <wjp> so those flags may mean something else entirely in PST
[20:27:57] <lynxlynxlynx> ed's saves work fine here
[20:27:58] <edheldil_> 1202 -- nice :)
[20:29:45] <wjp> definitely misinterpreted flags
[20:31:09] <wjp> setting A_ANI_BLEND and clearing A_ANI_BACKGROUND and A_ANI_PLAYONCE fixes the smoke; or at least most of it
[20:31:47] <wjp> some of it is still frozen
[20:32:12] <wjp> ah, A_ANI_FROZEN :-)
[20:32:17] <wjp> clearing that too fixes it
[20:32:48] <wjp> so looks like we'll have to figure out what the flags mean in PST...
[20:33:10] <wjp> but 1202 will be a very good one to test flag effects with in the original
[20:34:51] <chiv> its a nice small area...
[20:40:56] <chiv> http://pastebin.ca/22
[20:40:58] <Seniorita> pastebin - slePP - post number 22
[20:41:04] <chiv> http://pastebin.ca/2302212
[20:41:23] <chiv> http://pastebin.ca/2302112
[20:41:28] <Seniorita> pastebin - baator valgrind - post number 2302112
[20:41:28] <chiv> arg, my damn fingers
[20:41:59] <chiv> it definitely looks like the same error to me
[20:45:15] <brada> is this not the same scripting bug as before?
[20:46:02] <chiv> thats what I think anyway
[20:48:59] <Seniorita> [commit] Jaka Kranjc: Actor::GetAttackStyle: fix valgrind warning http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=34f4e50dd22e452f405bf633e40da35ff620a0b9
[21:08:39] --> fizzle has joined #gemrb
[21:15:04] <lynxlynxlynx> oh, now that i think of it, wasn't this already investigated? ChangeAIScript + continue
[21:15:56] <Seniorita> [commit] Jaka Kranjc: InfoPoint::InfoPoint: init all the members http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=d253b0bf04f2e4aae92ea9fb0512110dd2ebd7a4
[21:15:57] <Seniorita> [commit] Jaka Kranjc: MapNote::MapNote: init all members http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=ca6e31826fd74c85ec015ff5e2b63b1a95410ab6
[21:24:11] --> nutron|w has joined #gemrb
[21:25:38] --- ermo^ is now known as ermo
[21:27:20] <Seniorita> [commit] Willem Jan Palenstijn: Fix drawing of floor area animations in IWD/HOW/IWD2 http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=e32189c3bd4ab2efcd94a89d5454fcbe0779a698
[21:27:58] <-- traveler__ has left IRC (Ping timeout: 245 seconds)
[21:29:47] <wjp> chiv: did you already note this pst/ar1202 issue somewhere?
[21:30:33] <chiv> i cant remember actually, I might have forgotten... i did sort of rush through the end bit
[21:30:54] <fizzle> eh, I think we need a GA_PARTY flag or similar
[21:31:07] <fizzle> I can't select my thief anymore...
[21:31:59] <chiv> oh, it seems i made a note yes
[21:33:10] <chiv> right at the end of pst_quests wiki page
[21:36:36] --> traveler__ has joined #gemrb
[21:36:56] <-- miha has left IRC (Quit: Lost terminal)
[21:38:18] <traveler__> i would appreciate fixing this off by one font/icon bug, it hinders my gameplay pleasure ;) (joking aside, stacked items without counter are nuisance)
[21:39:23] <wjp> for now I would just use a slightly older version
[21:39:56] <wjp> (7b75c73834 is the commit right before this font merge)
[21:43:13] <traveler__> thanks for answer, it's not that important (indeed i can stay on version before merge) just would like to not detach myself from tip too much (e.g. lynx recent infopoint commit)
[21:44:04] <lynxlynxlynx> fizzle: jpw cp,e*
[21:44:11] <lynxlynxlynx> wow
[21:44:16] <lynxlynxlynx> how come?
[21:44:39] <fizzle> eh?
[21:45:36] <lynxlynxlynx> you can't select your thief
[21:45:45] <lynxlynxlynx> wjp: i can confirm the fix works great in iwd2
[21:45:48] <fizzle> oh, just hid in shadows
[21:46:04] <fizzle> wasn't overly difficult :)
[21:46:48] <lynxlynxlynx> ah, i see, you can't click directly
[21:46:55] <lynxlynxlynx> select by rect works though
[21:47:03] <fizzle> true
[21:47:14] <fizzle> and by portrait, too
[21:49:02] <lynxlynxlynx> we used to check ea there, but didn't really understand why the original did it
[21:49:16] <lynxlynxlynx> then the only known reason was removed
[21:49:59] <lynxlynxlynx> i guess you can simply shortcircuit Actor::IsInvisibleTo if both are party members
[21:50:14] <fizzle> not sure about that
[21:50:27] <fizzle> what if somebody charms your thief?
[21:51:02] --> Avenger has joined #gemrb
[21:51:02] --- ChanServ gives channel operator status to Avenger
[21:51:10] <lynxlynxlynx> mhm
[21:51:11] <fizzle> besides, the scriptable seems to be NULL there
[21:51:18] <Avenger> hi
[21:51:32] <fizzle> because we're not in a script, obviously
[21:51:40] <lynxlynxlynx> yeah, gamecontrol i guess
[21:53:29] <wjp> lynxlynxlynx: cool; thanks for checking
[21:54:37] <fizzle> so likely need to bring the EA check back?
[21:55:49] <lynxlynxlynx> sounds like it
[21:56:05] <lynxlynxlynx> iirc it was reversed, but pay no mind
[21:56:40] <fizzle> hm, looks like their was a flag to skip the check
[21:56:50] <fizzle> do we need that again, too?
[21:57:51] <lynxlynxlynx> i doubt it
[21:58:18] <lynxlynxlynx> it was for the SeeCore path
[21:59:03] <lynxlynxlynx> maybe it would be better to solve this higher
[21:59:58] --> CamDawg1 has joined #gemrb
[22:00:20] <-- CamDawg has left IRC (Ping timeout: 248 seconds)
[22:00:36] <-- Cuvieronius has left IRC (Quit: ChatZilla 0.9.89 [Firefox 18.0/20130104151925])
[22:01:35] <fizzle> hm, if we just return true on EA state then suddenly everyone can see your thief
[22:02:03] <lynxlynxlynx> of course
[22:04:35] <-- CamDawg1 has left IRC (Ping timeout: 255 seconds)
[22:05:24] <lynxlynxlynx> http://sprunge.us/QKRE?diff <-- something like this?
[22:06:50] <fizzle> having to call it twice isn't very nice
[22:07:42] <brada> i think i finally figured out how to reconcile chinese/other bams with standard ones so the status icons should be fixed shortly
[22:08:20] <fizzle> just check for EA <= EA_CONTROLLABLE && checker == NULL?
[22:09:46] <-- Avenger has left IRC (Quit: bye!)
[22:10:02] <lynxlynxlynx> i'm not sure that'd always be ok
[22:10:20] <lynxlynxlynx> would need to check all the callers and think about it
[22:10:48] <fizzle> I agree that solving it in GameControl would be nice
[22:11:19] <lynxlynxlynx> http://sprunge.us/URWV?diff
[22:11:30] <lynxlynxlynx> why didn't i think of that before
[22:12:03] <lynxlynxlynx> i know why, it's broken :)
[22:14:35] <lynxlynxlynx> http://sprunge.us/HZDh?diff <-- still a hack, but a bit better
[22:15:25] <lynxlynxlynx> there's no way to get that awesome stat set for party members, save for mods
[22:17:17] <fizzle> there aren't actually that many callers to IsInvisibleTo
[22:18:32] <lynxlynxlynx> there are many to ValidTarget though
[22:18:33] <traveler__> brada: cool. i'm sure there are some potential chinese gemrb users :) or there will be...
[22:18:46] <brada> im pushing the fix now
[22:18:57] <traveler__> no need to hurry
[22:19:01] <chiv> heh, i can go push it at the trow.cc site :)
[22:19:07] <traveler__> well, if you feel ppushed, that is
[22:19:22] <chiv> they will probably laugh at my chinese, but oh well...
[22:19:35] <traveler__> fixing is always appreciated though
[22:20:53] <traveler__> chiv: who knows, maybe you even score some new chinese dev ;p
[22:21:34] <brada> chiv: yo do realize that mb strings dont yet work in master right?
[22:21:54] <chiv> yes, im not going yet...
[22:23:48] <-- duckpunch has left IRC (Ping timeout: 248 seconds)
[22:24:22] <Seniorita> [commit] Brad Allred: BAMFont: fix offset states and initials + numeric fonts http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=7017a69bd7f436c8a9173faa0950c7781c3ebb59
[22:24:23] <Seniorita> [commit] Brad Allred: Merge branch 'master' of ssh://gemrb.git.sourceforge.net/gitroot/gemrb/gemrb http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=a1b5a0f7fd22b2c620cabe6c92e66089c57d8ce7
[22:24:59] <traveler__> clang 3.2 going into my branch of freebsd at this moment... i hope you will see me on irc today again after reboot :P
[22:25:09] <brada> interesting that seniorita only seems to see new commits that happen after the last one she reported
[22:25:17] <brada> ie doesnt see what i did before wjp
[22:25:37] --> duckpunch has joined #gemrb
[22:25:39] --> Avenger has joined #gemrb
[22:25:40] --- ChanServ gives channel operator status to Avenger
[22:26:20] <fizzle> lynxlynxlynx: where and how does the engine decide on whether to draw the avatar?
[22:26:54] <lynxlynxlynx> actor::draw or somesuch
[22:28:29] --> CamDawg has joined #gemrb
[22:29:51] <fizzle> hm, it uses the EA check there
[22:30:11] <fizzle> no help
[22:35:28] <lynxlynxlynx> wjp: in case you have enough willpower for another animated tile problem, teleport to ar0205 in bg2. The pulsating thing at the entrance is drawn above our container highlight, even when paused
[22:36:34] <lynxlynxlynx> brada: the demo still crashes due to fonts
[22:36:51] <brada> demo or min test?
[22:37:14] <lynxlynxlynx> demo
[22:37:24] <brada> ok ill run the demo
[22:37:38] <lynxlynxlynx> TTFFontManager.cpp:367 dest is destroyed
[22:38:36] <-- edheldil_ has left IRC (Ping timeout: 252 seconds)
[22:39:18] <lynxlynxlynx> and unrelated, it'd be great if the ttf plugin also searched system freetype paths by default
[22:40:25] <brada> excelent idea
[22:40:31] <brada> i was not aware of such a thing
[22:41:49] <lynxlynxlynx> should be $prefix/share/fonts/TTF where prefix is from freetype-config
[22:42:05] <lynxlynxlynx> but i'm sure there are c api ways to get that info
[22:42:44] <wjp> lynxlynxlynx: I think highlights are just drawn after tiles and before everything else?
[22:43:23] <lynxlynxlynx> could be, but it doesn't look like it for animated tiles
[22:43:35] <wjp> (if that's what DrawHighlightables does, anyway)
[22:43:36] <lynxlynxlynx> they are rare, that's why it is not so noticeable
[22:43:38] <wjp> hm
[22:44:03] <wjp> not for tonight anyway :-)
[22:44:04] <wjp> good night
[22:44:58] <-- brada has left IRC (Quit: brada)
[22:45:05] <lynxlynxlynx> nn
[22:46:31] --> brada has joined #gemrb
[22:46:37] <-- brada has left IRC (Client Quit)
[22:47:40] --> brada has joined #gemrb
[22:49:35] <fizzle> can't think of a way without calling GetActor twice
[22:49:43] <fizzle> how about this: http://nopaste.me/paste/76269180850f48b5de3ef3
[22:49:51] <Seniorita> @@ -1426,7 +1426,10 @@ void Ga - Diff - NoPaste.me - Secure and Anonymous
[22:50:22] <fizzle> maybe the InParty check is redundant?
[22:50:26] <-- brada has left IRC (Client Quit)
[22:51:37] --> brada has joined #gemrb
[22:57:38] <-- brada has left IRC (Read error: Connection reset by peer)
[22:58:03] <lynxlynxlynx> yeah, don't think it would be possible
[22:58:11] <lynxlynxlynx> add a comment and ship it
[23:00:41] --> brada has joined #gemrb
[23:02:25] <fizzle> hey, we can just call ValidTarget the second time, right?
[23:02:57] <fizzle> no need to go through the entire list again
[23:05:28] <lynxlynxlynx> can target_types be only GA_DEFAULT|GA_NO_HIDDEN there?
[23:06:22] <fizzle> default seems to be NO_DEAD|NO_HIDDEN
[23:08:50] <fizzle> okay, so with DEAD there can be multiple actors at the same point I guess?
[23:09:47] <fizzle> and there can be others, too
[23:09:52] <brada> interesting that a chinese ttf font works here
[23:16:55] <brada> glyph count is extraordinarily high for this font so possibly the calloc fails
[23:16:56] <brada> anyway
[23:17:18] <brada> im going to patch this by forcing the extended ascii charmap for now
[23:17:45] <lynxlynxlynx> fizzle: maybe revisit my third suggestion: http://sprunge.us/HZDh?diff
[23:18:35] <fizzle> why would an invisible party member not be invisible to the PC, too?
[23:20:23] <brada> lynx: is the demo functional at all?
[23:20:33] <brada> i have never been able to do anything with it
[23:22:15] <brada> also min test seems to still fail, but i cant see why in the stdio
[23:23:06] <Seniorita> [commit] Brad Allred: TTF: temporarily limit fonts to the FT_ENCODING_APPLE_ROMAN (extended ascii) range http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=5873fb4fc5d6455a8d511f492d0630e99471dd34
[23:23:24] <lynxlynxlynx> fizzle: s/he is, this is only for normal targetting
[23:23:31] <-- nutron|w has left IRC (Changing host)
[23:23:31] --> nutron|w has joined #gemrb
[23:23:50] <lynxlynxlynx> brada: demo loads an area and an actor, that's it right now
[23:24:02] <brada> right but i cant do anything at all
[23:24:34] <lynxlynxlynx> you can scroll the text area or crash it by waiting for the tooltip
[23:24:44] <lynxlynxlynx> or scroll the area in general
[23:25:02] <lynxlynxlynx> the actor is not added to the party for some reason :)
[23:25:18] <brada> the overlaping text area seems to conflic with the area in some way
[23:25:26] <fizzle> lynx: then I don't think I understand what that snippet is trying to do
[23:25:29] <brada> i just get rapid flashing of the text over the area
[23:25:50] <fizzle> if lastActor is invisible you'll just reset it to NULL
[23:26:30] <lynxlynxlynx> the only minor glitch i get is the actor's avatar flashing a bit on viewport move
[23:26:32] <brada> og ill bet its maybe my reolution setting
[23:27:40] <lynxlynxlynx> fizzle: clearly it is missing the ea check
[23:28:24] <fizzle> oh, so you were talking mainly about the ~NO:HIDDEN bit?
[23:28:49] <fizzle> I was thinking of using NO_DEAD instead
[23:29:24] <lynxlynxlynx> the bit and how to readd the check without checking all the flags again
[23:30:09] <lynxlynxlynx> but since this is in gamecontrol, out of any hot path, i have no problem with two GetActor calls
[23:31:04] <lynxlynxlynx> so far, we haven't managed to make a better approach that is still clean and always correct
[23:31:19] <traveler__> bbl
[23:32:19] <fizzle> http://nopaste.me/paste/59462103250f49574c9e7b?
[23:32:20] <Seniorita> @@ -1426,7 +1426,14 @@ void Ga - Diff - NoPaste.me - Secure and Anonymous
[23:33:56] <fizzle> looks a bit cleaner than removing and (possibly) readding the HIDDEN bit, I think
[23:34:10] <fizzle> since it might not even be set?
[23:35:28] <-- traveler__ has left IRC (Ping timeout: 245 seconds)
[23:37:31] <Seniorita> [commit] Brad Allred: BAMFont: fix minimal test http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=85109b66479944384ee7403c9ad9f6dea713d2ba
[23:37:38] <brada> thought i has already pushed that
[23:37:43] <brada> must have messed up in the rebase
[23:37:59] <-- brada has left IRC (Quit: brada)
[23:38:15] <-- Avenger has left IRC (Quit: bye!)
[23:38:30] --> brada has joined #gemrb
[23:39:03] <-- brada has left IRC (Client Quit)
[23:43:05] <lynxlynxlynx> fizzle: looks good
[23:44:36] <fizzle> cool
[23:44:40] <fizzle> will add that later
[23:44:44] <fizzle> off to sleep now
[23:45:17] <-- fizzle has left #gemrb
[23:50:25] <Seniorita> [wiki] known_problems http://www.gemrb.org/wiki/doku.php?id=known_problems&rev=1358207178&do=diff