#gemrb@irc.freenode.net logs for 12 Sep 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[02:10:02] --> raevol has joined #GemRb
[05:34:32] --> ditw has joined #GemRb
[05:38:18] <-- deepinthewoods has left IRC (Ping timeout: 240 seconds)
[06:14:17] --> edheldil_ has joined #GemRb
[06:15:32] --> Avenger has joined #GemRb
[06:15:32] --- ChanServ gives channel operator status to Avenger
[06:17:09] <Avenger> it looks like HoW selects not just expmap/worldmap based on the first 3 letters of an area resref. I moved to AR7001 from expmap, and it showed the location of ar7000 in the first one (didn't make it visible as i was never there)
[06:20:08] <Avenger> pst has a wmaplay.2da to decide which map to show as 'you are here', what about using that
[06:31:47] <edheldil_> porting it to other games?
[06:36:44] <Avenger> yes
[06:37:12] <Avenger> and it seems, HoW doesn't use the visible flag in the worldmap, it uses variables!
[06:37:29] <Avenger> if you set the global AR7000_visited = 1
[06:37:50] <Avenger> and look at the worldmap, it seems to be there, if you remove the variable, it goes away
[06:37:56] <Avenger> and the wmp file isn't changed
[06:38:19] <Avenger> we handled worlde only when one tries to leave in a direction
[06:38:33] <Avenger> i moved that a bit, now the worldmap is slightly better
[06:39:01] <Avenger> worlde.2da is the linking of variables to area visibility
[06:39:18] <Avenger> it works exactly the opposite as iesdp says :)
[06:43:52] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.9/20100824153629])
[07:38:16] <Lightkey> *facepalm*
[07:38:38] <Lightkey> I always thought wmaplay meant WMA play
[08:24:14] <-- edheldil_ has left IRC (Ping timeout: 276 seconds)
[08:37:38] --> edheldil_ has joined #GemRb
[08:49:31] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[08:53:30] <-- raevol has left IRC (Quit: Leaving.)
[08:54:08] --> edheldil_ has joined #GemRb
[09:04:12] --> lynxlynxlynx has joined #GemRb
[09:04:12] --- ChanServ gives channel operator status to lynxlynxlynx
[09:15:31] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[09:58:19] --> Avenger has joined #GemRb
[09:58:19] --- ChanServ gives channel operator status to Avenger
[10:03:06] <fuzzie> Avenger: good work on the HoW map stuff :)
[10:03:20] <Avenger> still working
[10:07:06] <CIA-93> GemRB: 03avenger_teambg * rc74b6c4793cb 10gemrb/gemrb/ (7 files in 5 dirs): optional flexible area location in worldmap
[10:07:11] <Avenger> ok, this works now, almost like the original, it cannot swap worldmaps if the area is not in the worldmap AND it needs to swap maps
[10:07:19] <Avenger> but that won't happen without hacking
[10:07:29] <Avenger> you start the expmap in ar9100 which is on the map
[10:08:01] <fuzzie> what about when going back? :) but it's not important
[10:08:15] <Avenger> going back is for kuldahar, right?
[10:08:19] <Avenger> that's also on the map
[10:08:50] <fuzzie> i don't know, because i only played how long enough to work out that it did this stuff based on the first digit of the area :)
[10:09:02] <Avenger> it will find ar9101 if you are already on the expmap, and will use ar9100 for it
[10:09:10] <fuzzie> no time to actually play games..
[10:09:11] <Avenger> it is not that simple
[10:09:32] <Avenger> i cannot simply round the areanumber to 1000
[10:09:41] <fuzzie> well, i didn't look into the details :P
[10:10:21] <fuzzie> same with worlde, i only looked at it enough to make it work for the bugs i found
[10:10:34] <Avenger> we will know how it exactly works when i peeked into it with IDA
[10:11:19] <Avenger> this shouldn't be as complex as scripting :)
[10:49:46] <xrogaan> heh
[10:49:49] <xrogaan> what is how ?
[10:51:07] <Lightkey> >_>
[10:51:23] <xrogaan> from the last commit
[10:51:57] <xrogaan> ' gemrb/override/how/'
[10:52:51] <fuzzie> Heart of Winter, iwd expansion
[10:53:01] <-- ditw has left IRC ()
[10:53:29] <xrogaan> okay
[10:53:32] <xrogaan> thanks :)
[11:05:55] <fuzzie> i think basically we have two sets of files, 'iwd' for the plain version and 'how' for the engine with the expansion, i guess because they're very different
[11:16:51] <lynxlynxlynx> yes, how is much better
[11:17:35] <lynxlynxlynx> iwd is very similar to bg1, while how has improvements similar to iwd2
[11:17:40] <xrogaan> and, is it possible to create a game from scratch with gemrb ?
[11:21:41] <lynxlynxlynx> sure
[11:26:22] <Lightkey> :>
[11:34:53] <xrogaan> there is a little problem with stacked items
[11:36:16] <xrogaan> when i take a stack and drop it on another one, it takes items from the 2nd stack instead of dropping them
[11:39:10] <xrogaan> or something odd
[11:43:58] <lynxlynxlynx> yes, we know
[11:45:19] <xrogaan> okay
[11:45:22] <xrogaan> [ResourceManager]: Searching for error02... Tried error02.wav error02.ogg error02.acm error02.wav [ERROR]
[11:45:25] <xrogaan> uhuhuh
[11:55:58] <Avenger> hey lynx, did you try the worldmap improvements?
[11:57:36] <Avenger> xrogaan if i drop a stack of 5 arrows on a stack of 10 arrows, i get a stack of 15 arrows, nothing unexpected
[11:57:57] <Avenger> could you be more specific about the problem?
[11:59:46] <fuzzie> i imagine it is about full stacks
[12:00:02] <fuzzie> but just a guess
[12:02:19] <Avenger> yes, the original would swap them, we don't
[12:02:39] <Avenger> and item splitting doesn't work
[12:03:52] <xrogaan> Avenger: well, not right now. But sometimes it don't do what's expected
[12:04:56] <xrogaan> next time, i'll notes every details
[12:05:33] <Avenger> fuzzie: do you know why OpenItemAmountWindow is not called when i use shift click or double click on an item?
[12:05:47] <lynxlynxlynx> with high values the dragged item gets restocked, not vice-versa
[12:06:02] <lynxlynxlynx> Avenger: not the latest commit yet
[12:06:25] --> edheldil_ has joined #GemRb
[12:06:38] <Avenger> lynx: the question is for you too
[12:07:16] <lynxlynxlynx> i don't know
[12:07:40] <lynxlynxlynx> shouldn't check today either, i have an exam tommorow
[12:08:46] <fuzzie> Avenger: which game?
[12:08:54] <Avenger> how and bg2 too
[12:09:27] <fuzzie> it really doesn't get called, you put a print in there?
[12:10:01] <Avenger> yes
[12:10:21] <Avenger> but it is set, because the other callback is not called
[12:10:26] <Avenger> so, wtf
[12:11:49] <lynxlynxlynx> wasn't the doubleclick case disabled, since we were drinking potions instead of splitting them?
[12:12:04] <Avenger> i use shift
[12:12:07] <Avenger> not double click
[12:12:09] <fuzzie> yeah, the doubleclick was always broken
[12:12:11] <fuzzie> but the shift worked
[12:12:24] <fuzzie> but then there's all these hacks for portraits in there now, it seems
[12:13:39] <fuzzie> so if you're dragging an item or portrait, this code will disable itself
[12:14:13] <fuzzie> so, does it happen in a new game before anything got dragged?
[12:14:14] <Avenger> i tried to take an item in the inventory screen
[12:14:21] <Avenger> shift clicking on an inventory item
[12:14:27] <Avenger> no portraits, no double click
[12:14:36] <fuzzie> and you didn't do any portraits or dragging items in that game run?
[12:14:41] <Avenger> no
[12:14:58] <Avenger> ok, maybe dragging items?
[12:15:04] <fuzzie> well, this is why i as
[12:15:06] <Avenger> i might have clicked on the item earlier
[12:15:06] <fuzzie> k
[12:15:34] <fuzzie> just guessing at causes
[12:17:01] <lynxlynxlynx> try afresh and rule it out
[12:18:53] <Avenger> ok, a little new info: i just didn't install the script
[12:19:02] <Avenger> actually it is called
[12:19:06] <Avenger> just doesn't do anything
[12:19:10] <Avenger> so it is not as mysterious
[12:19:23] <fuzzie> ok. you get a silly hat now. :P
[12:19:52] <lynxlynxlynx> unrelated: i get double starting gear for a new tob game
[12:20:14] <Avenger> hehe, yes, that is expected
[12:20:21] <Avenger> you still have your old code in it?
[12:20:49] <Avenger> now, the game upgrade happens when it the game->Extension field goes from 3 to 5
[12:21:17] <Avenger> so, the item code in chargen is not needed :)
[12:21:21] <fuzzie> oh?
[12:21:25] <fuzzie> i mean, they are not identical cases
[12:21:41] <fuzzie> the game upgrade from 4->5 should happen after chargen
[12:21:51] <fuzzie> and then it shouldn't add bags if they already got added by chargen
[12:23:14] <Avenger> if you can fix it, feel free :) probably you know what to do better than i
[12:23:31] <fuzzie> well
[12:23:37] <fuzzie> where is GameExpansion() called?
[12:23:40] <Avenger> i will try to find the bug in the item amount stuff
[12:23:51] <fuzzie> the original engine hardcodes it for the first ToB area, i seem to remember
[12:23:53] <Avenger> the python script?
[12:24:29] <Avenger> it is called after the game is loaded. When any game is loaded, i set EF_EXPANSION to 1
[12:24:42] <Avenger> then the eventflag handler calls into python
[12:24:55] <fuzzie> ok. so we don't do what the original does here.
[12:25:31] <fuzzie> but i'm confused because your code seems to call into chargen's GiveEquipment
[12:25:49] <fuzzie> you really found that in the RE?
[12:26:13] <Avenger> no, what i found in RE is in the wiki
[12:26:31] <Avenger> upgraded games and new games may be different
[12:26:39] <fuzzie> well, the wiki is correct
[12:26:51] <fuzzie> so i wonder why your added code calls GiveEquipment :)
[12:27:26] <Avenger> because it would work well with chargen, and 'amost' well with upgraded games
[12:27:43] <Avenger> you can try and improve it :)
[12:27:45] <fuzzie> it does the wrong thing for the SoA->ToB move
[12:28:30] <Avenger> i thought no item at all is worse than giving the same items as a fresh char gets
[12:29:06] <fuzzie> sure, i just wonder, since you RE-ed the correct item resrefs to give here :)
[12:31:58] <Avenger> haha, someone sabotaged openitemamountwindow
[12:32:21] <Avenger> it returns if isdraggingitem != 1 too
[12:33:12] <Avenger> wonder why :0
[12:34:08] <lynxlynxlynx> git blame
[12:34:42] <fuzzie> it was .. lynx :p
[12:35:14] <lynxlynxlynx> any clue in the commit msg?
[12:35:18] <fuzzie> see 5bf6f21f i think?
[12:35:24] <fuzzie> "fixed most of the item splitting gui bugs"
[12:35:37] <fuzzie> but just before that i messed with it too
[12:35:59] <lynxlynxlynx> i guess those were general runtime errors
[14:42:39] <Avenger> is it possible to print the call stack in python?
[14:43:08] <wjp> yes, there's a module for that. Let me see if I can find it
[14:43:24] <Avenger> google instant sucks
[14:43:29] <wjp> http://docs.python.org/library/traceback.html
[14:43:33] <Avenger> thanks
[14:45:05] <wjp> (print_stack is probably the one you want, unless you're in an exception handler)
[14:45:07] <Avenger> so i just use: traceback.print_stack() ?
[14:45:17] <Avenger> yeah, i'm in a normal function
[14:45:23] <wjp> I think that should do it, yes
[14:46:08] <Avenger> i just need to add import traceback ?
[14:46:14] * wjp nods
[14:50:39] <Avenger> you helped a lot :)
[14:51:32] <wjp> bug located? :-)
[14:54:39] <Avenger> fixed too
[14:54:48] <Avenger> killed, exterminated the whole swarm
[14:54:53] <wjp> whee :-)
[14:55:57] <CIA-93> GemRB: 03avenger_teambg * rd11690e051f4 10gemrb/gemrb/GUIScripts/bg2/GUIINV.py: fixed item splitting/doubleclick and all
[14:56:45] <Avenger> now i just need lynx or fuzzie to fix all the other games :) my brain almost melted from this
[14:57:43] <Avenger> looks like UpdateSlot brought the inventory window back front, while ItemAmountWindow was front (and modal). This is some bug in the core
[14:58:10] <Avenger> i couldn't fix that, but refreshed the modal state of ItemAmountWindow
[15:01:22] <wjp> hm
[15:01:45] <wjp> any idea which call to core causes this?
[15:02:21] <Avenger> something in UpdateSlot for sure
[15:02:41] <Avenger> probably a simple GetControl?
[15:02:52] <fuzzie> it's not doing ShowWindow?
[15:03:01] <fuzzie> or hacks with HideGUI/ShowGUI?
[15:03:08] <wjp> nothing jumps out
[15:03:13] <Avenger> or also SetEvent
[15:03:22] <Avenger> no fuzzie, there is no showwindow in it
[15:03:36] <Avenger> no hide/unhidegui
[15:04:09] <fuzzie> where do you see tis?
[15:04:39] <Avenger> got the newest changes?
[15:05:09] <fuzzie> RefreshInventoryWindow does showwindow, and is called from UpdateInventoryWindow which is called from everywhere
[15:05:11] <Avenger> if you comment out the showmodal from UpdateSlot, it will break when you try to doubleclick on an item
[15:05:32] <Avenger> OH
[15:05:41] <Avenger> cool
[15:05:46] <fuzzie> is that it? :P
[15:05:48] <Avenger> then the Showmodal is not needed
[15:05:50] <Avenger> yes
[15:05:50] <fuzzie> i didn't look at your new code
[15:06:07] <Avenger> UpdateInventoryWindow was called by DropDraggedItem or whatever
[15:06:10] <Avenger> OnDragItem
[15:06:19] <Avenger> that was what i needed wjp's help for
[15:06:23] <fuzzie> ah :)
[15:06:32] <Avenger> so the showmodal is not needed
[15:06:38] <Avenger> i killed that call anyway
[15:06:50] <fuzzie> but yes, i have an eternal battle with window ordering
[15:07:04] <Avenger> in OnDragItem
[15:07:08] <fuzzie> everything tries intefering with it, especially the hidegui/showgui stuff breaks the game
[15:07:22] <Avenger> that if ItemAmountWindow != None: return is crucial then ;)
[15:07:25] <fuzzie> because things do hidegui when they aren't hiding the gui, they just want to change something :(
[15:08:36] <Avenger> this is a whole chain of hacks, the first click in doubleclick picks the item up, so when i detect a doubleclick, i force it to be dropped back
[15:09:03] <Avenger> but the forced drop back refreshed the windows, and brought the underlaying window back active
[15:09:10] <Avenger> even if it was all gray :)
[15:09:17] <Avenger> that's still some bug
[15:09:30] <Avenger> Showwindow shouldn't make the window active if it is not modal
[15:09:50] <fuzzie> well, that is what the existing code relies on
[15:10:22] <fuzzie> (ShowWindow always bringing a window to the front of the draw order)
[15:10:36] <Avenger> well, except for a modal
[15:10:49] <Avenger> modals should trump this
[15:10:53] <fuzzie> yes, they should
[15:11:11] <Avenger> anyway, the bug is squished by avoiding the showwindows call :)
[15:11:20] <Avenger> and it isn't damaging anything
[15:12:14] <fuzzie> hmm maybe it is more buggy than i thought
[15:12:22] <fuzzie> i forget how this works, and i am pretty sick
[15:12:51] <fuzzie> was first week of university and all the students spent the week coughing and sneezing and they seem to have managed to infect me :(
[15:14:00] <CIA-93> GemRB: 03avenger_teambg * rf0a89cec946f 10gemrb/gemrb/GUIScripts/bg2/GUIINV.py: no need of the showmodal hack
[15:14:38] <Avenger> the doubleclick stuff now works perfect :)
[15:14:51] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[15:15:16] <Avenger> there is one small graphical glitch though, when you leave the split window, the yellow border is not updated
[15:15:34] <Avenger> it is refreshed if you drop the item and pick it back up
[15:29:20] <Avenger> haha, i bet this is because after killing the modal window, the top window doesn't become active
[15:36:51] <CIA-93> GemRB: 03avenger_teambg * r65a85f4bf025 10gemrb/gemrb/GUIScripts/bg2/GUIINV.py: fixed yellow border glitch
[16:13:28] <lynxlynxlynx> was the last updateslot removal necessary?
[16:14:03] <lynxlynxlynx> i haven't checked before, but now this stuff seems to work fine, except the stack number isn't updated immediately after the split
[16:21:22] <Avenger> isn't it?
[16:22:50] <Avenger> hmm, yeah,
[16:23:03] <Avenger> this fixes the yellow bug though :)
[16:23:09] <Avenger> so we need something midway
[16:41:16] <Avenger> uh i took out the wrong showmodal too
[16:44:35] <fuzzie> heh
[16:50:51] <Avenger> ok, i gave up on it, i fixed what i can
[16:51:01] <CIA-93> GemRB: 03avenger_teambg * r31b0ce3585a6 10gemrb/gemrb/GUIScripts/bg2/GUIINV.py: so far the best guiinv :)
[16:56:13] --> Maighstir has joined #GemRb
[16:59:01] --> edheldil_ has joined #GemRb
[17:18:34] <spike411> Meh those audio crashes (StereoDeinterleave32_X86) again: http://paste.jabbim.cz/4863 Seems like this Torque GB guy has the same problem: http://www.torquepowered.com/community/forums/viewthread/115115 ‘Decreasing the number of sound effects prevents the crash.’ :/
[17:18:59] <spike411> Mac OS X libs are PITA. :(
[17:50:34] --> kettuz has joined #GemRb
[18:21:35] --> tombhadAC has joined #GemRb
[18:41:34] <-- spike411 has left IRC (Quit: Manga & anime pokec na Jabberu: manga.cz@conf.netlab.cz)
[20:09:43] <-- Avenger has left IRC (Quit: bye!)
[20:10:05] --> pupnik has joined #GemRb
[20:14:41] <-- pupnik_ has left IRC (Ping timeout: 276 seconds)
[20:21:40] <-- kettuz has left IRC (Quit: Leaving)
[22:22:57] --> raevol has joined #GemRb
[22:48:00] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:32:08] <-- edheldil_ has left IRC (Ping timeout: 276 seconds)