#gemrb@irc.freenode.net logs for 5 Aug 2009 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[01:55:31] --> pupnik_ has joined #gemrb
[02:21:13] <-- pupnik has left IRC (Read error: 110 (Connection timed out))
[03:10:19] <-- raevol has left IRC (Remote closed the connection)
[03:15:35] --> raevol has joined #GemRB
[04:14:49] --> Gekz has joined #GemRB
[04:14:57] <-- Gekz_ has left IRC (Read error: 104 (Connection reset by peer))
[06:31:06] <-- raevol has left IRC (Remote closed the connection)
[06:32:51] --> raevol has joined #GemRB
[06:48:54] --> lynxlynxlynx has joined #gemrb
[06:48:54] --- ChanServ gives channel operator status to lynxlynxlynx
[07:37:53] --> Gekz_ has joined #GemRB
[07:43:30] --> Gekz__ has joined #GemRB
[07:44:08] <-- Gekz has left IRC (Read error: 104 (Connection reset by peer))
[08:52:27] <-- raevol has left IRC ("Leaving.")
[08:57:47] --> raevol has joined #GemRB
[09:06:03] <fuzzie> morning
[09:07:19] <lynxlynxlynx> oj
[09:13:43] <Gekz_> Hi
[09:22:29] <pupnik_> wow, the guy who played 'Manuel' on Fawlty Towers was a german (and a very intelligent person in interviews)
[09:28:44] <fuzzie> ok, the way that the 'detect traps' button is never unselected is annoying, i wonder what's going on therew
[09:33:49] <Gekz_> pupnik_: good to hear that you're comfortable now :)
[09:33:56] <Gekz_> pupnik_: still fear anaesthetic?
[09:43:19] <fuzzie> oh dear, Taimon is saying terrible things in the forums
[09:46:31] <lynxlynxlynx> link?
[09:46:44] <pupnik_> Gekz_: YES
[09:46:49] <pupnik_> i feel... different
[09:47:22] <pupnik_> like someone killed me, made a copy of me, and woke that up
[09:47:42] <fuzzie> lynxlynxlynx: just replying to zefklop's thread on startcutscene, it is not very interesting
[09:48:02] --> Pygma has joined #gemrb
[09:48:04] <fuzzie> he's just saying that we have to do some much nastier things to handle cutscenes properly..
[09:48:14] <lynxlynxlynx> boring
[09:48:44] <Gekz_> pupnik_: lol.
[09:48:49] <fuzzie> i was looking through old gemrb forum posts with huge lists of problems, it's really nice to go through and see that it's all fixed now :)
[09:51:16] <Pygma> Hey, installing this to try get Planescape Torment working (won't work for me on Wine)... Is it worth getting latest SVN or is it very unstable or not very different?
[09:51:38] <fuzzie> latest svn is probably better
[09:52:20] <fuzzie> but you will probably have more luck with wine, PST in gemrb is very much a "it sometimes accidentally works" thing, i think you get stuck relatively quickly
[09:53:30] <Pygma> Well I've tried for ages to get it working in Wine and doesn't seem to, so I might as well try, getting the 2 BG games soon anyway so even if it doesn't I could still use it with them
[09:53:54] <fuzzie> you've seen the wine appdb entry? it says setting "Full Screen" in torment.ini to 0 should make it run
[09:54:22] <Pygma> Yeah I've tried that
[09:54:37] <Pygma> And everything else it's asked there and many combinations of all of them
[10:02:14] <lynxlynxlynx> try an older wine, it used to work for me
[10:03:43] <fuzzie> it would be annoying if it didn't work, i'll have to try it later myself
[10:03:57] <fuzzie> i still don't know what the pst round times are :(
[10:04:42] <fuzzie> i should ask these people who are disassembling it
[10:09:38] <Pygma> Actually running a Win98 virtual machine might work, since I believe I ran FF8 in that ad it's more taxing
[10:11:02] <fuzzie> planescape torment is surprisingly demanding on the graphics hardware
[10:11:31] <fuzzie> there's a lot of upset at nvidia recently because their modern drivers don't accelerate it correctly
[10:11:51] <fuzzie> that said, it works fine for me in a vmware virtual machine :)
[10:14:18] <Pygma> Well one of the Wine errors I get seems to say something about faulty driver when I google for it, so perhaps that's it, though I doubt NVidia will be releasing any updates for it
[10:15:07] <fuzzie> i think it's just a problem with the windows drivers and directdraw, it should be ok on linux
[10:18:05] <Pygma> Well PST seems to at least be starting up, or when you said it didn't work did you just mean it's unplayable for other reasons?
[10:18:44] <lynxlynxlynx> gemrb is not finished yet and pst is one of the least finished parts
[10:18:56] <lynxlynxlynx> due to it being so different
[10:22:45] <fuzzie> it's a lot better than it was, but there are so many things we don't handle
[10:26:07] <lynxlynxlynx> yet :)
[10:27:25] <fuzzie> i spent a while going through a pst walkthrough and fixed the bugs, but levelup etc is such a mess
[10:27:59] <Gekz_> lol
[10:28:26] <fuzzie> so i am a bit unmotivated in the absense of anyone else working on it too :)
[10:29:15] <fuzzie> today i am being amazingly productive by repeatedly refreshing exam results pages, even though i know the lecturer is on holiday for two weeks
[10:30:38] <fuzzie> the g3 forums recently are filled with pst stuff we could probably usefully implement, lots of good tasks for anyone new..
[10:31:07] <fuzzie> but the levelup/rules/etc are full of TNO hacks which no-one except scient is interested in
[10:32:40] <fuzzie> and scient is last-seen a month ago, alas
[10:35:02] <lynxlynxlynx> it's summer
[10:38:52] <fuzzie> for some reason, in r6268 i added an actor->TargetDoor hack to ToggleDoor, that needs removing before it causes crashes..
[10:53:24] --> dawid has joined #GemRb
[11:27:52] <-- dawid has left IRC (Remote closed the connection)
[12:20:14] <Pygma> Out of interest, would you guys say that sound is essential in PST? I have a windows install but without sound
[12:30:22] <lynxlynxlynx> sound is important for immersion
[12:30:52] <lynxlynxlynx> it's wierd that you don't have any though
[12:32:34] <fuzzie> i am having an interesting time trying to make gemrb work nicely on ubuntu 9.04
[12:32:57] <fuzzie> pulseaudio uses 100% cpu, pasuspender doesn't make a difference, killing pulseaudio doesn't work because it respawns when i run gemrb..
[12:35:22] <lynxlynxlynx> hehe
[12:35:27] <lynxlynxlynx> early adoption ftw
[12:38:12] <Gekz_> fuzzie: -> points to a member of a struct, right?
[12:39:20] <fuzzie> if you're referencing it via a pointer
[12:39:28] <fuzzie> '.' is normal access, '->' is access via a pointer
[12:39:37] <Gekz_> I see
[12:39:39] <Gekz_> I hate pointers
[12:39:40] <Gekz_> lol
[12:39:50] <Gekz_> my brain struggles to understand them
[12:39:53] <Gekz_> although I kinda get it now
[12:41:27] <lynxlynxlynx> pointers are cheap, so they get used a lot
[12:43:39] <fuzzie> everything is a pointer internally, so :)
[12:49:05] <fuzzie> pointers usually seems to be where people end up having a lot of trouble with C++ courses
[12:49:29] <fuzzie> at my uni the student assistants had a terrible time trying to explain them over and over to the poor physics/maths students in particular
[12:50:00] <fuzzie> who have to pass the course but aren't interested enough to already grasp the idea
[12:50:31] <lynxlynxlynx> it's pretty abstract if you don't use it yourself
[12:52:18] <fuzzie> my dad taught me when i was really still a little kid, which sounds a bit mad but really something i'm quite grateful for now :)
[12:53:16] <lynxlynxlynx> gemrb even has some threestar code :)
[12:54:40] <Pygma> I learned pointers when i was but a child too, I've never seen any explanation of them that actually gets across how useful they are which might explain why so many people hate them
[14:58:11] --> tombhadAC has joined #gemrb
[16:22:37] --> zefklop has joined #gemrb
[16:26:05] --> Avenger has joined #gemrb
[16:26:19] --- ChanServ gives channel operator status to Avenger
[16:27:04] <Avenger> hi
[16:27:05] <zefklop> hi everybody
[16:28:58] <Avenger> hey fuzzie, i work on pst :) just yesterday i added the death counters :P
[16:30:44] <fuzzie> will you add the spawn ini handling for them too?
[16:30:48] <fuzzie> or is that already done?
[16:32:26] <Avenger> not done
[16:32:50] <fuzzie> i am revising for tomorrow's exam, but i will catch up tomorrow evening :)
[16:34:07] <Avenger> do you know ANY ini file where any of those counters are used?
[16:34:26] <fuzzie> no, do you think they are unused?
[16:34:29] <Avenger> if not, then this feature is between irrelevant and unneeded :)
[16:34:31] <fuzzie> it is simple to check, let me do it
[16:35:25] <Avenger> besides, i don't know if it is a trigger/condition or an attribute modifier
[16:35:29] <fuzzie> obviously death_scriptname is used..
[16:35:36] <fuzzie> which sets an appearance flag bit
[16:35:47] <Avenger> ok, that is easy to set, i guess
[16:36:01] <fuzzie> otherwise pst's spawn inis set no other appearance bits
[16:36:56] <zefklop> uh uh, some interesting pieces of informations there :-)
[16:37:26] <Avenger> this code is not really important, most if not all of those cre's already got the bit set
[16:37:35] <zefklop> you will have them at g3, in the cutscene topic
[16:40:59] <zefklop> tha attack action will continue even if the target is inconscious right?
[16:41:49] <lynxlynxlynx> i think so
[16:42:27] <Avenger> oh i see, the spawn ini stuff clears the appearance flags in the original game
[16:42:40] <Avenger> this causes weird bugs qwinn just recently found
[16:43:03] <Avenger> what if we don't care about clearing the flags :)
[16:44:19] <Avenger> these pst developers always over complicated things
[16:44:59] <Avenger> first they set the flags in the .cre, then clear it in spawn ini, and then if they didn't forget, set some back
[16:46:18] <Avenger> i'm sure they would have loved weidu too :) all info in lousy text files, parsed more than once until you forget what did you set and where
[16:49:31] <fuzzie> Avenger: yes, we should just set the flags if needed
[16:50:01] <fuzzie> i expect we might end up adding the others for fixpack compatibility, but we'll see..
[16:50:33] <fuzzie> zefklop: the
[16:50:48] <fuzzie> zefklop: the SetInterrupt stuff is interesting, but implementing CutsceneId that way is probably not worth the hassle
[16:50:50] <Avenger> what about a test, we could check if the creature had that flag originally, and print some warning if spawn ini would set it differently
[16:51:04] <Avenger> cutsceneid is good as it is now, i think
[16:51:36] <fuzzie> yes, we shouldn't be trying to implement hacks that the original engine did just for the sake of having hacks
[16:52:03] <zefklop> where did I speak about setinterrupt?
[16:52:13] <fuzzie> zefklop: it is in the reply to your forum post
[16:52:16] <Avenger> it is an instant, which means, it is evaluated at the queue push time. It sets Sender->cutsceneid, which is used by the cutscene code
[16:52:29] <fuzzie> Avenger: that is not how it works in the original engine, though
[16:52:37] <fuzzie> but i think gemrb's code is better
[16:52:48] <Avenger> the cutscene code pushes all actions on the queue
[16:52:50] <zefklop> fuzzie: this is interesting, but incomplete
[16:53:05] <fuzzie> zefklop: yes, no-one has answered the question about clearing the queue yet, did your mod work it out? :)
[16:53:37] <zefklop> yes, I just ran it another time to double check, and I'll answer to my own post :)
[16:53:41] <fuzzie> the forums are often not a good place to ask, because they only care about how the original engine disassembly does it :(
[16:53:45] <Avenger> well, i guess, it is not the same as the original, but i think the difference shows only when one uses incorrect scripts?
[16:53:54] <Avenger> like not all cutscene blocks start with a cutsceneid
[16:53:55] <fuzzie> Avenger: yes, so we should keep our version and not use the hack :)
[16:54:09] <fuzzie> unless there are bad scripts out there?
[16:54:09] <Avenger> in that case ours seem to be more 'syntax correct'
[16:54:31] <zefklop> the one I'll give you is very ugly :-D
[16:54:33] <Avenger> we could add a warning that a block is not starting with cutsceneid?
[16:54:49] <fuzzie> sure
[16:55:07] <fuzzie> but warnings on spawn ini and cutsceneid and etc are only useful if gemrb stops outputting so many bad warnings
[16:55:14] <fuzzie> someone should perhaps work on that sometime
[16:55:57] <Avenger> well, either the warning needs to be removed, or the cause fixed
[16:56:14] <fuzzie> well, for instance, gemrb outputs errors if the actionoverride target doesn't exist
[16:56:23] <fuzzie> but that should be in script debug mode only
[16:56:28] <Avenger> you would faint if you find out how many warnings we suppress :)
[16:56:39] <Avenger> oh you know that
[16:56:41] <Avenger> k
[16:56:54] <fuzzie> yes, the script debug flags are very helpful!
[16:56:59] <-- |Cable| has left IRC (Remote closed the connection)
[16:56:59] <fuzzie> but only when i am actually debugging scripts :)
[16:57:19] <fuzzie> the pre-existing debug lines at the top of effects are great too
[16:58:00] <fuzzie> it's really nice to just be able to change them to 'if (1)'
[17:00:11] <Avenger> yes i agree the actionoverride failure is not needed anymore, UNLESS, you change cutscenes again :)
[17:00:40] <zefklop> take a look at g3... i'm too lazy to retype it
[17:00:47] <zefklop> not that it is big...
[17:01:14] <zefklop> this is much simpler than I thought
[17:02:03] <fuzzie> ok, so that is exactly what gemrb does
[17:02:14] <fuzzie> except we implement CutSceneId properly instead of their hack
[17:02:21] <Avenger> yes
[17:02:34] <zefklop> it lacks the freeze in scripts
[17:02:35] <fuzzie> our interrupt code needs some work anyway
[17:02:46] <fuzzie> zefklop: well, how do you mean 'freeze'?
[17:02:52] <Avenger> yes, the interrupt/freeze stuff might be needed
[17:03:05] <zefklop> like freeze :-) no scripts are executed
[17:03:07] <fuzzie> you don't give enough detail to explain
[17:03:13] <fuzzie> gemrb doesn't execute new scripts in a cutscene
[17:03:19] <fuzzie> but obviously there are still actions in the queue
[17:03:26] <Avenger> if you change something about scripts, please comment, why did you do it
[17:03:37] <Avenger> and also heed similar comments already in the code
[17:03:40] <fuzzie> if you mean "no actions are executed", then that needs more research
[17:03:45] <zefklop> so, how do you explain that PC is killed by the ogre?
[17:03:55] <Avenger> zefklop: it isn't in my code
[17:03:58] <fuzzie> i think the ogre's AI script shouldn't be run, it is a timing bug
[17:04:12] <Avenger> it didnt' attack me for a long time
[17:04:18] <zefklop> Avenger: meh
[17:04:31] <Avenger> when cutscenes cleared all previous actions
[17:04:59] <fuzzie> but do you know whether it is "no new scripts run" or "no actions are executed" in a cutscene?
[17:05:05] <fuzzie> that seems the important information
[17:05:43] <Avenger> well, i thought it is : clear existing actions (and possibly no scripts run)
[17:05:44] <zefklop> hmmm... I'd have to create a new creature in the middle of the cutscene
[17:05:48] <fuzzie> if action queues are not cleared, and actions do not run in cutscenes, then all actions must resume after a cutscene?
[17:06:23] <fuzzie> that seems not hard to test, just queue some actions on a creature and then start a cutscene, see when they're executed
[17:06:47] <zefklop> wait, I'll give you the code, and then you'll comment/add watever you want -)
[17:06:49] <Avenger> well, how would you stuff the cutscene actions before them, and in order
[17:06:50] <zefklop> :-)
[17:06:53] <fuzzie> well, i cannot run bg1
[17:06:56] <fuzzie> all i can run is gemrb
[17:07:03] <fuzzie> so i can't actually test anything myself :)
[17:11:28] <fuzzie> but i don't think we should change anything until someone is sure of exactly how it works
[17:11:50] <fuzzie> sometime we need to fix ActionOverride to clear the action queue, for example, but i'm still not sure of exactly how that works, either
[17:12:30] <fuzzie> it is .. difficult
[17:23:35] <zefklop> meh, using DisplayString in a .d file of weidu is impossible
[17:26:21] <zefklop> sent to the patches section of sf
[17:26:48] <zefklop> hope everyone likes weidu :-)
[17:27:03] <fuzzie> i guess we will have to learn :-)
[17:27:08] <zefklop> btw, where is the bug reporting of weidu?
[17:27:31] <fuzzie> they have a forum
[17:28:25] <fuzzie> http://forums.pocketplane.net/index.php?board=44.0
[17:29:00] <Avenger> i'm pretty sure DisplayString in weidu works
[17:29:12] <fuzzie> a .d is compiled to dialogs, i'm not sure how those work
[17:29:35] <fuzzie> can you run DisplayString from inside a dialog at all?
[17:29:58] <Avenger> i think, yes
[17:30:14] <Avenger> you will have to scroll back
[17:35:21] <zefklop> no, it is specifying a string in an action from .d that does not work
[17:35:45] <zefklop> when using it with ~, it refuses to install
[17:35:59] <zefklop> and using quotes don't work
[17:38:52] <fuzzie> well, maybe it's just not supported
[17:39:05] <zefklop> yes, I"ll ask
[17:39:14] <fuzzie> does it work if you provide a strref?
[17:40:08] <zefklop> it will, I think, but I can"t figure out how to create an strref and get its number as a variable
[17:40:36] <fuzzie> i think you use a TRA file?
[17:43:38] <zefklop> no
[17:43:46] <zefklop> no reason to translate this
[17:44:25] <zefklop> weidu's forum is at SHS right?
[17:44:36] <fuzzie> i linked to weidu's forum above, it's at pocketplane
[17:44:51] <fuzzie> it's pretty quiet, if you're just asking for help then SHS is probably better
[17:45:26] <Avenger> hmm startcutscene didn't clear the action queue?
[17:45:40] <Avenger> that's terrible
[17:45:47] <zefklop> yes!
[17:45:56] <fuzzie> i seem to remember there are scripts which depend on that behaviour
[17:46:20] <Avenger> but they are executed after the cutscene, right?
[17:46:38] <zefklop> no, during the cutscene
[17:46:45] <Avenger> hmm?
[17:46:54] <Avenger> don't confuse me
[17:46:58] <fuzzie> the action queue is still processed while the cutscene is running
[17:47:10] <zefklop> thanks fuzzie
[17:47:35] <Avenger> ok, so the actions of the cutscene are added on top of any pending actions, but no new actions are added by scripts
[17:47:35] <fuzzie> i will try and help with zefklop's mod once i have time, it does seem best for checking all this
[17:47:47] <fuzzie> Avenger: i'm not sure that they're added on top
[17:47:55] <fuzzie> i don't think anyone has any evidence for that yet
[17:47:57] <Avenger> you guys just said that
[17:48:07] <zefklop> I do agree with avanger
[17:48:13] <zefklop> Avenger sorry
[17:48:22] <Avenger> -->but they are executed after the cutscene, right? zefklop> no, during the cutscene ->fuzzie> the action queue is still processed while the cutscene is running
[17:49:09] <zefklop> attack action is interrupted by walkto?
[17:49:26] <Avenger> attack action 'contains' walkto
[17:49:32] <Avenger> that part is easy
[17:49:45] <zefklop> say, walkto an other point
[17:49:48] <fuzzie> only one action ever happens at once
[17:49:51] <Avenger> oh hmm, well
[17:49:56] <fuzzie> so if you do another action like walking, yes, the attack is stopped
[17:50:19] <Avenger> fuzzie: depends on how the second action was given
[17:50:24] <fuzzie> Avenger: oh?
[17:50:36] <Avenger> if not via the gui, why would it be stopped
[17:50:38] <fuzzie> i mean, once the action is already executing, the attack is stopped
[17:50:48] <Avenger> oh, hmm
[17:50:50] <fuzzie> ActionOverride will certainly wipe the queue, and a new block will wipe the queue
[17:51:16] <fuzzie> and when the queue is wiped, the current action is stopped
[17:51:23] <Avenger> but it is possible to use several actionoverrides in a row, no?
[17:51:27] <fuzzie> yes
[17:51:34] <Avenger> so, it cannot wipe the queue
[17:51:36] <fuzzie> the first actionoverride clears the queue of the target
[17:51:48] <Avenger> but the second would wipe it again
[17:51:49] <fuzzie> i don't know exactly the circumstances there, i said that in here an hour ago or so :)
[17:51:59] <zefklop> that's terrible!
[17:52:05] <fuzzie> but the first one wipes it, and if you do another one immediately after, the second one does not
[17:52:15] <fuzzie> but i don't know what happens if you Wait() and then try it again
[17:52:49] <Avenger> i keep to my projectiles
[17:52:59] <fuzzie> zefklop: it is all terrible :)
[17:53:23] <zefklop> let's try to avoid killing our brains!
[17:53:25] <Avenger> though pst projectiles are a nightmare too
[17:53:34] <fuzzie> zefklop: when you get weidu to create a .d, it only provides the strref
[17:53:40] <fuzzie> for example: IF ~~ THEN DO ~SetGlobal("PartyCured","GLOBAL",1)
[17:53:40] <fuzzie> DisplayString("Marek",20675)
[17:53:48] <fuzzie> ^- no string, only the strref
[17:53:53] <zefklop> I agree with that!
[17:54:05] <fuzzie> so there is probably no 'nice' way to do it
[17:54:05] <zefklop> Strings are fine in dialogs
[17:54:28] <zefklop> this strings in actions associated to dialogs tat don't work
[17:54:50] <zefklop> ouch, my keyboard is sick!
[17:55:15] <zefklop> this is strings in actions associated to dialogs that don' work
[17:55:29] <fuzzie> yes, i think that is probably just not supported
[17:55:49] <zefklop> ->feature request!
[17:55:54] <fuzzie> :)
[17:56:24] <fuzzie> Avenger: i guess you are keeping track of the item/spell threads on the forums, and will notice anything that gemrb doesn't implement?
[17:56:58] <Avenger> i didn't notice anything special about items or spells yet
[17:57:07] --> |Cable| has joined #gemrb
[17:57:29] <Avenger> the only thing that pst has hardcoded spell list for dead targeting: i know about that, just found it the same time
[17:57:40] <Avenger> i cursed a lot about it
[17:57:43] <fuzzie> and the hardcoded treasure list is already known about?
[17:57:53] <Avenger> we already provide the 2da :)
[17:58:06] <Avenger> it was found several years ago
[17:58:12] <fuzzie> ok :)
[17:58:30] <Avenger> i don't know if it is perfectly the same chances, but it works
[17:58:37] <fuzzie> i don't think anyone will notice
[17:59:12] <fuzzie> i have set weidu making a .d file for every .dlg in the game, it is a bit slow :)
[18:00:38] <Avenger> brb
[18:00:39] <-- Avenger has left IRC ("bye!")
[18:03:22] --> Avenger has joined #gemrb
[18:03:43] --- ChanServ gives channel operator status to Avenger
[18:04:37] <Avenger> fuzzie: i just noticed you complain about the detect traps button.
[18:04:49] <Avenger> NoAction should clear the modal action
[18:04:58] <Avenger> or any ClearActions
[18:05:21] <fuzzie> i think it does, but the UI doesn't seem to notice
[18:05:23] <Avenger> do you know if the modal action is cleared successfully and this is only a gui feedback problem?
[18:05:25] <Avenger> ok
[18:05:55] <fuzzie> maybe i'm wrong, though, because i couldn't find the code just now
[18:06:12] <Avenger> the code for feedback, i don't know if we have any :D
[18:06:25] <Avenger> the code to clear modal, should be there
[18:07:01] <fuzzie> you'd think it would be SetModal, but that doesn't get called from anywhere relevant..
[18:07:19] <Avenger> //not sure about this
[18:07:21] <Avenger> //act->SetModal(MS_NONE);
[18:07:23] <Avenger> oops
[18:07:44] <Avenger> void GameScript::ClearActions(Scriptable* Sender, Action* /*parameters*/)
[18:07:55] <fuzzie> well, that is in the wrong place i think :)
[18:08:04] <Avenger> though, i would prefer it called right inside Actor::ClearActions
[18:08:12] <fuzzie> yes, so i guess ClearActions should be virtual
[18:08:26] <Avenger> yep
[18:08:34] <Avenger> as modal is part only of actor
[18:08:42] <fuzzie> well, it will go on my todo list, unless you do it :)
[18:08:51] <Avenger> it is yours :)
[18:09:59] <fuzzie> if you read backscroll then you perhaps also see me talking about how i seem to have added a TargetDoor pointer to Actor for use in the ToggleDoor action, i don't remember why, it seems like it will crash
[18:10:34] <Avenger> don't use pointers
[18:10:36] <fuzzie> i think maybe i just forgot about GenerateActionDirect..
[18:10:37] <Avenger> for this
[18:10:48] <fuzzie> but that also needs fixing as soon as I have time :(
[18:11:09] <fuzzie> oh, right, GenerateActionDirect only works for actors
[18:11:33] <Avenger> doors could be targeted by their scriptname
[18:11:34] <fuzzie> hm, I wonder how to do that without a pointer
[18:11:41] <fuzzie> do they always have a unique scriptname?
[18:11:55] <Avenger> yes, i think doors are safe :)
[18:11:59] <Avenger> unlike actors
[18:12:20] <Avenger> though, it worths a try to test two doors with same name
[18:12:27] <Avenger> would the original IE get confused?
[18:13:00] <fuzzie> UseContainer has the same problem, i guess
[18:13:29] <fuzzie> with the added complication of piles
[18:13:37] <Avenger> oh yes, piles
[18:13:46] <Avenger> though, i think i generate a unique name for piles
[18:13:55] <Avenger> but the engine doesn't
[18:14:06] <Avenger> so saved games might be crappy
[18:14:37] <fuzzie> i think there are also unnamed piles in AREs
[18:15:20] <Avenger> yes
[18:15:29] <Avenger> i name them with this: sprintf(heapname,"heap_%hd.%hd",position.x,position.y);
[18:15:54] <fuzzie> that's only if the heap doesn't already exist
[18:16:09] <Avenger> yes, sadly
[18:16:17] <fuzzie> but sometimes AREs have them, i wonder if it's okay to patch those on load, or whether the original engine will crash
[18:16:28] <Avenger> exactly my thoughts :)
[18:17:03] <Avenger> wont' crash, but if there is some ground pile the original engine depends on, it will vanish after a gemrb session
[18:17:14] <fuzzie> oh, of course, saved games will just have AREs too
[18:17:17] <Avenger> not sure if it is still a goal to support switching
[18:17:21] <fuzzie> i forgot about that :)
[18:18:02] <fuzzie> it would be nice if we could make switching work, but there are many ways to make the actions work properly, so it's no problem
[18:18:53] <Avenger> well, i would prefer a solution which doesn't increase the action struct size. So any identifier should be squeezed into the current object size
[18:20:03] <Avenger> lets first go with the scriptname
[18:20:30] <fuzzie> someone needs to check that doors are uniquely named, but I think that is easy with ie_shell
[18:20:40] <Avenger> did you know, gemrb originally used it for all generated actions? :) it was fun when i tried to talk to some actor, and my pc walked to someone else
[18:20:52] <fuzzie> hehe
[18:21:05] <fuzzie> yes, that was a problem with the dialog code too
[18:21:33] <fuzzie> you would talk to an actor, and then the dialog would switch to someone else via a transition, then back to the original via scriptname, but it would be the wrong original actor!
[18:22:00] <Avenger> and how do you know which is the right one
[18:22:12] <Avenger> there could be a long chain
[18:22:14] <fuzzie> i made gemrb remember the original actor of the dialog, for now
[18:22:29] <fuzzie> i didn't find any dialogs which required anything more complicated, yet
[18:22:46] <Avenger> bg2 has all those interjections
[18:23:03] <Avenger> especially when flooded with npc mods :)
[18:23:05] <fuzzie> yes, pst has a lot of them too
[18:23:55] <Avenger> though i don't know many ambiguous scriptnames involved with interjections
[18:24:57] <fuzzie> in pst they are everywhere, thanks to spawns
[18:25:06] <Avenger> hehe
[18:25:29] <fuzzie> but it works fine now, it's really great
[18:25:39] <fuzzie> i was really surprised to find all the spawn ini already implemented!
[18:25:48] <Avenger> yeah, almost all
[18:25:54] <Avenger> without it , pst was very deserted
[18:26:07] <Avenger> so i had to do it, especially when i found iwd uses it too :)
[18:48:32] <lynxlynxlynx> another thing about strings and strrefs
[18:48:59] <lynxlynxlynx> can we generate a strref for an arbitrary string at runtime somehow?
[18:49:33] <lynxlynxlynx> it is needed for the saving of the biography, at least the way it is currently set up (setplayerstring)
[18:50:00] <Avenger> not right now, i left the tlk override halfway implemented
[18:50:27] <Avenger> it is a difficult thing
[18:50:59] <Avenger> from one side: we don't want the stale strings overcrowd the save game, from the other side: we need multiple custom strings
[18:52:09] <Avenger> ahh, and i forgot, we should not free up a string slot while it is being used
[18:52:32] <Avenger> most likely they are referenced only in one point
[18:52:35] <Avenger> i hope...
[18:52:39] <fuzzie> we could write some kind of reference count system?
[18:52:59] <fuzzie> where code is responsible for saying "i need this string" at load time, and "i don't need this string any more" when necessary
[18:53:08] <Avenger> the reference count is not needed, i think
[18:53:44] <Avenger> custom strings are used for: 1. biography, 2. area mapnotes
[18:54:47] <Avenger> i think the system would allocate a new string for every instance, even if it has the same string
[18:55:33] <Avenger> so, you need to do only the following: 1. request a new string id, 2. modify an existing string id
[18:55:53] <Avenger> maybe 3. kill a string for good
[18:56:01] <lynxlynxlynx> ok, i'll leave it be
[18:56:07] <Avenger> :(
[18:56:12] <Avenger> why don't you implement it :)
[18:56:21] <lynxlynxlynx> because i suck
[18:56:56] <Avenger> we can't let fuzzie to do everything :D
[19:08:51] <zefklop> hi again
[19:09:16] <-- lynxlynxlynx has left IRC (Remote closed the connection)
[19:16:27] --> zipi has joined #GemRB
[19:24:47] <fuzzie> we should maybe start a list of code that people could do, together with an explanation of what would need done
[19:25:46] --> Gekz has joined #GemRB
[19:27:13] <Avenger> yep
[19:27:23] <-- Gekz_ has left IRC (Read error: 104 (Connection reset by peer))
[19:27:26] <-- Gekz__ has left IRC (Read error: 104 (Connection reset by peer))
[19:27:35] <Avenger> a more detailed todo
[19:30:11] <zipi> hi
[19:30:36] --> Gekz_ has joined #GemRB
[19:32:02] <zipi> can you tell me if this is correct or not: http://www.pastebin.be/20192
[19:34:17] <zefklop> what are you trying to do?
[19:34:20] <Avenger> it seems correct, but it should be part of the main script.
[19:35:04] <zipi> which one?
[19:35:22] <zefklop> GuiCommonWindows I think
[19:35:44] <Avenger> i mean, keymap.py should be included by messagewindow
[19:35:55] <Avenger> i think it is MessageWindow
[19:36:01] <fuzzie> Avenger: but then the keystrokes don't work anywhere else
[19:36:03] <Avenger> but i'm a bit out of the loop :)
[19:36:09] <Avenger> oh hmm
[19:36:12] <fuzzie> it is complicated, and i think trying to put the functions in a global file does not work
[19:36:25] <Avenger> i think it is a good thing :D
[19:36:27] <zipi> so i need to put it in for each game?
[19:36:29] <fuzzie> because you have to think about character generation, for example
[19:36:38] <fuzzie> and you have to think about when you're in a window that you can't easily exit
[19:36:44] <Avenger> you need a different set of keystrokes outside the main game script :)
[19:37:08] <fuzzie> but we haven't seen the rest of zipi's patch, so we don't know how it would work
[19:37:08] <Avenger> besides, what keystrokes should work in the chargen???
[19:37:26] <fuzzie> Avenger: none, but we shouldn't be hard-coding that in the core
[19:37:53] <zipi> should i paste it without the python part?
[19:38:17] <Avenger> maybe you don't even need a python part
[19:38:31] <Avenger> it would be less complex, for sure
[19:38:47] <fuzzie> the keystrokes are too complicated to handle otherwise, i think
[19:39:15] <fuzzie> you would have to write all kinds of complicated code to pass *some* keys onto the python, and have lookup tables for the spells, and etc
[19:40:11] <Avenger> well, the only part i see now is key_Inventory
[19:40:28] <Avenger> that would be simpler without
[19:40:47] <zefklop> (silly question) what about attaching a button to a key?
[19:40:56] <zefklop> instead of a function?
[19:41:06] <fuzzie> zefklop: that would work, but again, only for some of the keystrokes :)
[19:41:10] <fuzzie> so you end up coding everything twice
[19:41:29] <fuzzie> the python seems the simplest method, you just pass everything to the guiscript and the core doesn't worry any more
[19:41:52] <Avenger> you will still code everything twice, or more
[19:41:56] <zefklop> like gemrb.AttachButtonKeyStrike(button,"keymapName")
[19:42:07] <Avenger> you just delegated the problem to python
[19:42:11] <fuzzie> Avenger: well, there are a lot more coders for the guiscript
[19:42:16] <fuzzie> and the guiscript has to run most of the code anyway
[19:42:53] <zefklop> you don't have to code much... handling the key in game control and pass find the button associated with it
[19:43:01] <fuzzie> zefklop: there are often no buttons present
[19:43:07] <zipi> :) anyone interested in the code?
[19:43:11] <zefklop> yes
[19:43:15] <fuzzie> zefklop: so you have to have two pieces of code, one button handling, and another for where there is no button :)
[19:43:18] <fuzzie> zipi: yes
[19:43:29] <zipi> i merge it with the latest svn and upload the patch file somewhere
[19:43:40] <Avenger> use sf, please
[19:44:15] <zefklop> fuzzie: in which case is there nno button?
[19:44:23] <zefklop> or any gamecontrol
[19:44:24] <Avenger> the bug tracker is good because it could be commented, and changes could be uploaded the same place
[19:44:31] <zipi> ok
[19:45:19] <Avenger> all keys should be processed by gamecontrol, at least i don't know of any other
[19:45:22] <fuzzie> zefklop: for spells, for quicksaves, for 'display location', for hard pause, for selecting multiple chars at once, for mac bindings
[19:46:01] <Avenger> there are no keymap keys outside of the main game screen, are there?
[19:46:09] <fuzzie> Avenger: there are
[19:46:29] <Avenger> one example?
[19:46:31] <fuzzie> for switching screens, switching between PCs (which does not change the selection, so must be in python!)
[19:46:51] <fuzzie> someone also mentioned something else that i can't remember
[19:46:55] <zefklop> meh, this is quite complete
[19:47:10] <Avenger> switching screens?
[19:47:18] <fuzzie> Avenger: 'I' for inventory, for example
[19:47:25] <Avenger> oh, i see
[19:48:09] <zefklop> ok, let's look at what zipi did :-)
[19:48:15] <zefklop> zipi?
[19:48:18] <Avenger> got anything outside of MessageWindow too?
[19:48:42] <Avenger> mostly chargen i guess
[19:48:59] <fuzzie> i think key bindings in chargen/startup might be useful, but there are none in the original gam
[19:49:03] <fuzzie> e
[19:49:18] <fuzzie> since we handle escape/enter automatically
[19:50:07] --> zipike has joined #GemRB
[19:51:47] <Avenger> see you later
[19:51:49] <-- Avenger has left IRC ("ChatZilla 0.9.85 [Firefox 3.5.1/20090715094852]")
[19:59:53] --> zipp has joined #GemRB
[20:03:32] --> Edheldil has joined #gemrb
[20:03:32] --- ChanServ gives channel operator status to Edheldil
[20:05:16] <-- zipi has left IRC (Read error: 110 (Connection timed out))
[20:05:46] --> zipi has joined #GemRB
[20:05:57] <zipi> i uploaded it
[20:07:08] <fuzzie> hm
[20:07:20] <zefklop> you should use the Message() functions
[20:07:26] <fuzzie> putting the keypresses in Control::OnKeyPress is quite clever
[20:07:42] <fuzzie> parsing the ini file manually is a lot of work, though!
[20:08:58] --> zippi has joined #GemRB
[20:09:59] <zippi> sorry but i have a very bad internet connection
[20:10:20] <-- zipike has left IRC (Read error: 110 (Connection timed out))
[20:10:31] <Edheldil> Looks like my unfinished patch is getting obsolete :)
[20:11:02] <zefklop> why Key-32 ?
[20:11:54] <fuzzie> maybe to convert case or similar?
[20:12:34] <zefklop> hmmm? tolower() should be fine there
[20:13:39] <fuzzie> we probably want to convert the code to use Message() and INIImporter, and not use fixed-length string allocation, but the patch is definitely the right idea, i think
[20:14:10] <fuzzie> i really don't know how to do the python side, though
[20:14:33] <fuzzie> functions like key_Inventory sound like the right idea
[20:14:47] <fuzzie> but maybe we want different ones for different screens..
[20:15:03] <zefklop> why not calling openinventory directly?
[20:15:10] <zefklop> I'm ot very fond of wrappers
[20:15:26] <fuzzie> because often we have to close other windows first
[20:15:44] <fuzzie> although actually, the buttons must already do that..
[20:16:00] <fuzzie> oh, right, the buttons are disabled when we have another window open, i guess
[20:16:00] <zefklop> well, openinventory and similars already do that
[20:16:11] <zefklop> too
[20:16:27] <zefklop> that's why I like the associate key to button approach
[20:16:38] <fuzzie> but then you don't have the buttons :)
[20:17:23] <zefklop> well, having no buttons often means that we can't and shouldn't click it :-D
[20:17:35] <fuzzie> but often it means that it should work anyway
[20:17:35] <zefklop> and then the keymap should not work
[20:17:38] <fuzzie> so that kind of spoils it
[20:18:00] <zefklop> meh, we disagree there
[20:18:12] <fuzzie> we would end up creating a large number of fake buttons, and that is a lot more work than a function lookup table
[20:18:28] <fuzzie> or else we end up writing codepaths for both buttons and no-buttons
[20:18:31] <-- zipp has left IRC (Read error: 110 (Connection timed out))
[20:18:37] <zefklop> yeah I know
[20:18:50] <zefklop> ok, you spoke about fake buttons
[20:18:52] <fuzzie> we can merge a patch which at least reads the keymap entries, though
[20:18:58] <zefklop> I spoke about true buttons
[20:19:08] <fuzzie> yes, but for a lot of things there are no true buttons
[20:19:29] <fuzzie> so you have to have some way to do those, either python functions or fake buttons?
[20:19:31] <zefklop> well, for each of them there is one I think, but not accessible directly
[20:19:49] <fuzzie> there's no button for hard pause, or quicksave, or location, i'm fairly sure
[20:20:02] <zefklop> hard pause? Space bar?
[20:20:08] <fuzzie> spacebar is normal pause
[20:20:13] <fuzzie> which you can get by clicking the timer
[20:20:19] <fuzzie> hard pause means "always pause"
[20:20:21] <Edheldil> there will be lots of functions for things like cast: fireball, cast: magic missile ....
[20:20:39] <fuzzie> Edheldil: well, my original thought was to simply get the python to dynamically create those functions
[20:20:51] <zefklop> wow!
[20:20:59] <fuzzie> or we could have a guiscript for KeyPressed
[20:21:03] <Edheldil> that should not be hard
[20:21:13] <fuzzie> but, i mean, i don't hate the idea of associating things with buttons!
[20:21:16] <fuzzie> i just don't know what is best
[20:21:31] <zefklop> wait...
[20:21:39] <Edheldil> I like the direct mapping better tahn some marshalling function in python
[20:21:54] <zefklop> almost all of the keymap things is accessible via GUI
[20:21:57] <fuzzie> and i know that we can't implement all keystrokes with existing buttons
[20:22:04] <fuzzie> and i know that we can't implement all keystrokes in the core
[20:22:14] <fuzzie> so we have to call python sometimes anyway
[20:22:37] <zefklop> what about processing ALL input from python?
[20:23:02] <Edheldil> will be a bit slow, I think
[20:23:05] <fuzzie> sure, we could pass keystrokes to python, and then python could have a dictionary
[20:23:21] <fuzzie> if we implement scrolling and textareas in the core, i think the rest would be fine in python
[20:23:26] <zefklop> I mea, the keyborad one
[20:23:29] <zefklop> mean
[20:23:55] <zefklop> keyboard
[20:24:06] <zefklop> mes
[20:24:21] <fuzzie> but if we're going to do that, the core might as well pass the looked-up keymap entries to python instead of the pressed key..
[20:25:22] <Edheldil> don't forget python will need the actual keystrokes for keys config window
[20:25:30] <-- zipi has left IRC (Read error: 110 (Connection timed out))
[20:25:45] <fuzzie> but python could have some kind of table, which mapped the keymap entries to a button or a function, whichever is available
[20:26:00] <zefklop> is there one in BG?
[20:26:09] <Edheldil> although it will have to lack the keymaps there anuhow
[20:26:13] <zefklop> isn't it in bgconfig?
[20:26:32] <Edheldil> there is definitely one in pst
[20:26:50] <fuzzie> i had forgotten about pst
[20:27:04] <zefklop> meow
[20:27:14] <zefklop> I always thought it could be a featurre of gemrb
[20:27:24] <zefklop> black isle guys beat me
[20:27:27] <zefklop> :-)
[20:27:35] <fuzzie> wow, it displays something in gemrb, too
[20:27:51] <fuzzie> ok, it quickly crashed in gemrb
[20:27:59] <Edheldil> somebody has optimized those keystrokes out, though :)
[20:28:05] <fuzzie> but i am impressed anyway :)
[20:28:34] <Edheldil> it 's just astatic text
[20:28:40] <Edheldil> in gemrb
[20:28:57] <fuzzie> but, well, it is difficult to merge a patch unless we work out how this should work
[20:29:05] <fuzzie> since if we don't even know where the ini file must be read..
[20:29:44] <Edheldil> first of all, I think the keymap should be a different class for Interface
[20:29:51] <Edheldil> from
[20:29:57] <fuzzie> a couple of these (Right Click and Display Info) would probably be much less painful in the core
[20:30:28] <Edheldil> what do you mean?
[20:30:30] <fuzzie> Edheldil: well, the proposed patch just adds the keymap entries to Interface..
[20:31:02] <fuzzie> Edheldil: 'Right Click' simulates a right-click (since Macs don't have them), and 'Display Info' (tab normally) displays party HP above heads
[20:31:17] <fuzzie> both of those would probably be difficult to make work in guiscript
[20:32:12] <fuzzie> of course, both of them could also be hard-coded
[20:32:15] <fuzzie> i am just thinking
[20:32:28] <Edheldil> actually, they should be possibly configurable to allow swapping rmb x lmb
[20:33:26] <Edheldil> and there are now more endowed mice than just 3 mb + wheel :)
[20:33:50] <zefklop> Edheldil: nothing to do there, but did you see the "mod" I uploaded in the patches section of sf?
[20:34:24] <Edheldil> the test one? not yet
[20:34:30] <zefklop> yes
[20:34:33] <Edheldil> grrr, new form of spam ...
[20:34:34] <fuzzie> we could easily catch extra mouse buttons and use them in the keymap system too
[20:34:53] <zefklop> spam?
[20:35:06] <Edheldil> look at the comment to your keymap patch
[20:35:52] --> zipp has joined #GemRB
[20:36:19] --- zipp is now known as zipi
[20:36:37] <fuzzie> Edheldil: but, do you think the keymap is fine in Interface, or you think it should be elsewhere?
[20:36:38] <Edheldil> hi, zipi
[20:36:45] <zefklop> clever, they don't use forbidden words, only some physics related
[20:37:19] <Edheldil> I think that perhaps keymap should be it's own class. Possibly attachable to a window
[20:37:27] <-- zippi has left IRC (Read error: 110 (Connection timed out))
[20:37:43] <zipi> re
[20:37:48] <zipi> i've read back what you wrote
[20:38:17] <fuzzie> Edheldil: and the class should be responsible for reading the ini?
[20:39:20] <zefklop> fuzzie: apart from mutiselect, all non "cast *** spell" keystroke have a button attached
[20:39:35] <Edheldil> the reading is better in the interface until the INI reading is factored out as well
[20:40:09] <zefklop> this mmakes it easier
[20:40:15] <fuzzie> zefklop: where are they?
[20:40:36] <fuzzie> i mean, for hard pause and location and quicksave and quickload
[20:40:48] <zipi> in the cpp
[20:40:52] <zipi> in gamecontrol
[20:40:53] <fuzzie> and if you don't count the spell keystrokes, then you probably can't count any of the PS:T or IWD2 buttons either..
[20:41:48] <zipi> the other thing is that you can call the cpp functions from python, or am I wrong?
[20:42:17] <fuzzie> zipi: we are wondering if we can make the python simple
[20:42:35] <fuzzie> so that we don't have to have key_ functions for everything
[20:42:46] <zipi> ok
[20:42:59] <zipi> so we have to solve in cpp
[20:43:15] <zipi> i would suggests some mapping table mentioned above by somebody
[20:43:29] <zipi> like Inventory->OpenInventoryWindow
[20:43:41] <zipi> Return To Game->ReturnToGame
[20:43:42] <zipi> etc.
[20:43:58] <zipi> simple txt or anything should do it
[20:44:40] <Edheldil> the txt should also define the keystroke groups, like in pst
[20:45:36] <fuzzie> and that lookup table could easily be in python
[20:46:01] <zipi> :)
[20:46:26] <fuzzie> but using the buttons is also a good idea, we can include the button details in the lookup table, and disable the key when the button is disabled
[20:46:35] <fuzzie> i don't know quite what zefklop was thinking, though
[20:47:17] <Edheldil> aren't the buttons often not disabled, just obscured?
[20:47:30] <zipi> yes, in a way all keymap commands are made up of pressing buttons
[20:48:10] <fuzzie> Edheldil: i'm not sure..
[20:49:01] <Edheldil> is the goal of that getting buttons visibly pressed on a keystroke? Because else just as well the buttons could be dependent on the handler functions
[20:49:54] <Edheldil> or it's possibly to control the running of the handler, right?
[20:50:07] <fuzzie> i think the idea is to stop the handler from running when it shouldn't run
[20:50:12] <Edheldil> the ability to run it
[20:50:20] <Edheldil> ah, sure
[20:50:20] <fuzzie> for example, if you have an 'item information' window open, then the handlers shouldn't work
[20:50:30] <fuzzie> but i think this can all go in python..
[20:50:50] --> zipike has joined #GemRB
[20:51:00] <fuzzie> and i can change zipi's patch a little and then use that for the C++ side, just calling a python function instead
[20:51:12] <zipike> not necessarily visible
[20:51:36] <fuzzie> but zefklop has gone quiet, so i wait
[21:03:26] --> zipp has joined #GemRB
[21:09:49] <-- zipi has left IRC (Read error: 110 (Connection timed out))
[21:10:34] --> zipi has joined #GemRB
[21:10:53] <zefklop> yes?
[21:13:14] <zefklop> what I'd do is (I have only BG1 there)
[21:13:46] <zefklop> I can see only three kinds of key mapping
[21:13:59] <zefklop> general ones, cast priest spell, cast wizard spell
[21:14:33] <zefklop> for the first ones, attach them to the button thy're associated with (lacks hrd pause, multiselect and a few others...)
[21:15:12] <zefklop> for the next ones, look into the selected character spell book and go on with waht is in there
[21:15:54] <fuzzie> but do you think it's okay to do it all in python?
[21:16:30] <zefklop> that way, only the button part has to be done from python
[21:16:49] <zefklop> the rest is on cpp part IMHO
[21:16:58] <fuzzie> ok.
[21:17:08] <zefklop> know which char is selected, etc... is easier from core
[21:18:38] <zefklop> Edheldil, zipi: what do you think?
[21:19:36] <-- zipike has left IRC (Read error: 110 (Connection timed out))
[21:19:41] <zefklop> searching through spell books for names present in keymap.ini is easier than producing some functions from python.
[21:19:42] --> zipike has joined #GemRB
[21:20:37] <fuzzie> zefklop: http://fuzzie.org/nfs/gemrb/keymap/ by the way
[21:22:10] <Edheldil> calling generic python/gs hook could be even a possibility in the mapping file
[21:22:47] <zefklop> fuzzie: thanks
[21:23:08] <zefklop> the bg2 one is the most complete
[21:23:48] <fuzzie> they're all different in some way, though
[21:24:20] <fuzzie> so i thought it would be helpful
[21:25:37] <fuzzie> PS:T in particular is more of a problem because a lot of buttons are on the floating menu
[21:25:40] <-- zipp has left IRC (Read error: 110 (Connection timed out))
[21:26:07] <zefklop> oh yeah, I remember ths thing now
[21:26:52] <zefklop> that said, grom a performance POV, the floating menu should be hidden, and a right click should just unhide it
[21:27:27] <zefklop> that way, there woould be not much loading, and buttons would always be there :-)
[21:27:37] <fuzzie> i think it is rarely enough used that that would not be an optimisation :)
[21:27:51] <fuzzie> but it might be a fix for this kind of thing
[21:28:10] <Edheldil> fuzzie, you are joking.. floating menu is used a lot
[21:28:13] <zefklop> oh... it is not as useful than in nwn then?
[21:28:20] <zefklop> ah
[21:28:30] <fuzzie> Edheldil: i just use the keystrokes :)
[21:28:37] <zefklop> meh!
[21:28:49] <Edheldil> I do not :)
[21:29:05] <fuzzie> but obviously it will be vital for some people
[21:29:25] <fuzzie> another good task for someone would be to fix the floating menu ..
[21:29:28] <zefklop> I should find a copy of this game, that would help in this kind of discussions
[21:29:31] <fuzzie> unfortunately very few people seem to have PST
[21:29:44] <Edheldil> mine has disappeared :(
[21:29:59] <zefklop> :-(
[21:30:01] <fuzzie> zefklop: i think it would be no problem to keep it hidden, if it made keymap easier
[21:30:27] <zipike> i will check these tommorrow, sorry but'm a bit tired
[21:30:39] <zipike> and my connection drops recently
[21:30:40] <zefklop> god night :))
[21:30:49] <zipike> gn
[21:30:49] <Edheldil> gn, zipi
[21:31:02] <-- zipike has left IRC ()
[21:31:19] <zefklop> fuzzie: I think that a way to hide controls should be implemented anyway
[21:31:41] <zefklop> don't know f floating menu is a contrl or a window though
[21:31:58] <zefklop> and I really can't get used to this kkeyboard
[21:32:19] <fuzzie> it seems to be a window
[21:32:37] <fuzzie> in any case, we could simply let the python use buttons when available or functions when not, for now
[21:32:39] <Edheldil> window
[21:34:09] <zefklop> fuzzie: I think we should agree on a good design first, because changing it after would be painful
[21:35:18] <fuzzie> well, i think that something working is better than something not working, but i do not have much time myself
[21:36:21] <zefklop> nor do I, and I'd prefer to do more script testing...
[21:36:56] <fuzzie> more script testing would be very good
[21:37:50] <-- zipi has left IRC (Read error: 110 (Connection timed out))
[21:38:28] <zefklop> yeah. I'll introduce a third NPC, and I'll call him tester3
[21:38:38] <zefklop> the more, the funnier
[21:38:38] <fuzzie> i am mostly interested in just being able to play the games :)
[21:38:47] <zefklop> me too :-)
[21:38:54] <fuzzie> so it's nice if other people can work out what needs fixing!
[21:39:15] <fuzzie> bg1 was playable quite far before this ogre thing
[21:39:40] <zefklop> yeah, was really surprised myself
[21:43:05] <fuzzie> the viewport thing needs fixing, where cutscenes are shown in the wrong place
[21:43:49] <fuzzie> i don't know what the problem is, maybe not adjusting the coordinates for a hidden GUI?
[21:44:40] <zefklop> no
[21:44:53] <zefklop> vanilla BG1 does not hide the GUI
[21:45:00] <zefklop> it just greys it
[21:45:18] <zefklop> the PC is under the GUI
[21:45:37] <zefklop> gemrb vp is right
[21:45:39] <fuzzie> i thought that in vanilla BG1, the fight is also in the middle of the screen
[21:45:46] <fuzzie> is the fight just in the wrong place?
[21:46:49] <zefklop> it's in the center
[21:46:54] <zefklop> at least i think
[21:49:10] <zefklop> i'll check
[21:49:40] <fuzzie> maybe the dialog causes the problem..
[21:49:48] <fuzzie> it's okay until the dialog starts, but then it's off-centre
[21:50:28] <zefklop> hmm
[21:50:48] <zefklop> we might deactivate this center on speaker thing
[21:50:56] <zefklop> when in cutscene
[21:51:18] <fuzzie> another thing to test
[21:53:11] <zefklop> oh dear!
[21:53:39] <zefklop> tester2 killed tester1
[21:54:22] <zefklop> Something on thaco calculation
[21:55:01] <zefklop> do we handle malus/bonus from weapon proficiency somewhere?
[21:55:06] <fuzzie> yes
[21:55:13] <zefklop> even for fist?
[21:55:26] <fuzzie> i think fists only happen for monks?
[21:55:34] <fuzzie> lynx added some special 2da files for that
[21:55:56] <zefklop> eh, you can always fight with fist even if you are not a monk
[21:56:34] <fuzzie> but you don't get bonuses otherwise
[21:56:42] <zefklop> and I think that you have some muluses there
[21:56:46] <zefklop> maluses
[21:57:01] <zefklop> aas with any other weapon that you are not specialized in
[21:58:16] <fuzzie> i think the engine just defaults to the not-specialised numbers, but lynx and mattinb are the experts there
[21:59:16] <zefklop> I think I read somewher that it was the contrary
[21:59:33] <zefklop> something like one star gets no malus, nor bonus
[21:59:41] <zefklop> and a econd one gets bonus
[21:59:44] <zefklop> second
[22:00:00] <zefklop> and none gets malus
[22:00:07] <fuzzie> well, the current code matches bg2 behaviour
[22:00:13] <zefklop> should be verified though
[22:00:17] <fuzzie> but no-one checked anything else
[22:00:59] <zefklop> look at the string displayed in skills
[22:01:04] <zefklop> at chargen
[22:01:50] <zefklop> if he attacks with a weapon that he is not proficient with, he suffers a minus to hs attack roll
[22:01:55] <fuzzie> hm
[22:06:12] <-- |Cable| has left IRC (Read error: 110 (Connection timed out))
[22:06:14] <fuzzie> ok, well, i don't know :) ask lynx
[22:06:44] <zefklop> :-)
[22:07:02] <zefklop> ok, I'll poke him tomoroow
[22:07:08] <zefklop> rrow!
[22:07:20] <zefklop> good night e verybody
[22:07:27] <-- zefklop has left IRC ("ChatZilla 0.9.85 [Firefox 3.0.8/2009032609]")
[22:17:30] <-- Gekz has left IRC ("leaving")
[22:27:41] --- pupnik_ is now known as pupnik
[23:17:28] <-- Edheldil has left IRC ("Really?")