#gemrb@irc.freenode.net logs for 18 Oct 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:36:07] <-- Maighstir has left IRC (Quit: ~ Trillian Astra - www.trillian.im ~)
[04:20:08] <-- deepinthewoods has left IRC (Ping timeout: 245 seconds)
[04:37:06] --> deepinthewoods has joined #GemRb
[04:47:57] <-- Gekz has left IRC (Remote host closed the connection)
[07:54:05] --> lubos has joined #GemRb
[08:00:39] <-- |Cable| has left IRC (Remote host closed the connection)
[08:06:12] --> Gekz has joined #GemRb
[08:06:40] <-- Gekz has left IRC (Remote host closed the connection)
[08:09:42] --> Gekz has joined #GemRb
[08:37:42] --> lynxlynxlynx has joined #GemRb
[08:37:42] --- ChanServ gives channel operator status to lynxlynxlynx
[08:39:23] <edheldil> Hi, Lynx
[08:52:32] <lynxlynxlynx> oj
[10:15:45] --> SiENcE has joined #GemRb
[10:32:13] <-- deepinthewoods has left IRC (Ping timeout: 245 seconds)
[12:16:02] --> pupnik_ has joined #GemRb
[12:17:51] <-- pupnik has left IRC (Read error: Operation timed out)
[13:04:46] --- pupnik_ is now known as pupnik
[13:55:07] <-- edheldil has left IRC (Ping timeout: 272 seconds)
[14:13:18] --> edheldil has joined #GemRb
[14:13:18] --- ChanServ gives channel operator status to edheldil
[15:22:35] --> Maighstir has joined #GemRb
[15:48:33] <-- lubos has left IRC (Quit: Leaving.)
[15:53:11] <-- SiENcE has left IRC (Quit: @all: cya)
[16:41:10] --> deepinthewoods has joined #GemRb
[17:11:09] --> Avenger has joined #GemRb
[17:11:09] --- ChanServ gives channel operator status to Avenger
[17:11:18] <Avenger> hello people
[17:11:51] <Avenger> fuzzie: i started looking into HoW too, as the new version's requirement seems to hang on SetDialogRange
[17:12:48] <Avenger> the action itself seems to simply set a field in the actor (but i didn't see how it is used yet). Anyway, it seems we can simply add a zero field, which could be set by the action, then subtract that value from the personaldistances
[17:12:56] <Avenger> i mean, distancesquares
[17:20:13] <Avenger> the HoW actor struct looks similar to the ToB one with the following differences: no object copies for the trigger recreation, the objects are huge, even bigger than in pst. There are separate memory pieces for every possible overlay in every creature (no pointers, but whole objects embedded). This makes HoW creatures almost as big as the ToB ones, but with much less fields.
[17:47:00] <fuzzie> weird
[17:51:04] <Avenger> if you want, you or lynx can implement what i said, and check if this fixes the iwd problems. it may very well be the last blocker for HoW
[17:51:29] <Avenger> even if it isn't i will try to find out the next problem that comes up :)
[17:58:38] <Avenger> fuzzie: another weird thing in HoW: the destroyself action seems to remove some curious items from the inventory before setting the self destruct flag. Items named like: REG1HP2, MIN1HP, RINGTOM... etc
[17:58:53] <Avenger> looks like this means self destruct wouldn't work with minhp1 (in HoW)
[17:59:16] <Avenger> i guess we can safely ignore this, and just make self destruct work regardless of minhp
[17:59:22] <fuzzie> there's something (maybe the death opcode?) which doesn't work in ToB with minhp set
[17:59:39] <Avenger> this is the self destruct opcode, which should still work
[17:59:44] <fuzzie> there's a few threads somewhere because someone removed a minhp item and the game broke
[18:00:06] <Avenger> err, not even opcode, but the action 'DestroySelf'
[18:00:13] <Avenger> which simply sets a flag
[18:00:24] <Avenger> before removing a select list of items
[18:00:29] <Avenger> it is totally crap
[18:00:33] <fuzzie> the same flag in ToB?
[18:00:41] <Avenger> well, 'same-ish' :)
[18:00:59] <fuzzie> i wouldn't be so sure that it is safe to ignore such things, but i guess if we find a bug caused by that then we know what caused it
[18:01:00] <Avenger> same action, but the actor structs are different
[18:01:42] <fuzzie> but you've seen that it is lucky that the HoW scripts work at all, since they messed so many of them up, i wouldn't be surprised to find them relying on stupid things by accident :)
[18:01:42] <Avenger> yep, that's why i said this
[18:01:50] <fuzzie> yes, thankyou
[18:02:17] <Avenger> i just wouldn't want to add another list with these stupid item names
[18:03:13] <Avenger> the action list seems to be exactly the same up to faceobject
[18:03:29] <Avenger> even that some of these actions are missing from actions.ids
[18:03:33] <Avenger> they are used internally
[18:03:36] <Avenger> so, MEH
[18:03:45] <Avenger> we will need to use the action override feature
[18:06:34] <fuzzie> lynx encountered some problems with that, i think generate action doesn't pay attention
[18:06:48] <fuzzie> easy enough to fix though
[18:07:28] <Avenger> another thing: we might want to get rid of our Nids... hacks, as the engine seems to use existing actions for the gui generated actions
[18:07:37] <Avenger> like PlayerDialog is used when a player wants to talk
[18:07:52] <fuzzie> sure
[18:08:00] <fuzzie> and UseDoor and UseContainer and etc
[18:08:05] <Avenger> yep
[18:08:10] <Avenger> they are all used for that
[18:08:14] <fuzzie> just no-one knew what they did at the time :)
[18:08:21] <Avenger> LeaveAreaName for travel regions
[18:08:29] <Avenger> that is used in all games
[18:08:40] <Avenger> so an action.ids override would work
[18:09:10] <Avenger> well, when i did that hack i didn't know the original does the same, with its own actions :)
[18:10:11] <Avenger> they also shoehorn the objects into the actions, though not that way as we do. They simply force the object in the actions in their own version of generateaction
[18:10:23] <Avenger> they simply got different versions of the action constructor
[18:10:59] <Avenger> i use some slightly obscure marker in the ids list (-1) means an object is needed
[18:11:22] <Avenger> hope you already know that part
[18:14:27] <fuzzie> i don't think i ever saw that
[18:15:19] <fuzzie> i see it, though
[18:15:34] <fuzzie> but i only really ever touch this code to fix bugs
[18:18:05] <Avenger> but you wanted to add the universal object stuff. if objects in scripts have the globalid as they should, this hack isn't needed
[18:18:55] <Avenger> i basically hacked the global id into an used field, but the original has it in the object always
[18:24:00] <fuzzie> yeah, we really need that
[18:31:27] <Avenger> hmm, HoW seems to have a changestat lua command that can set any stat, i didn't know that (it is almost identical to our SetPlayerStat command)
[18:31:57] <Avenger> maybe it was disabled somehow thats why i didn't notice it
[18:33:26] <Avenger> HoW seems to have 86 stats, but only the first 70 are in stats.ids
[18:35:10] <Avenger> it is good they got this code, it gives a perfect mapping to the stats
[18:53:22] <Avenger> haha, almost perfect mapping, it seems they flunked the cut&paste code
[19:30:39] <-- pupnik has left IRC (Remote host closed the connection)
[19:36:58] <Avenger> huh, the ForceHide action has an object parameter, but it forces the Sender (only actors can do this) to do the hide modal action
[19:37:17] <Avenger> if the forcehide object doesn't exist, the action fails
[19:37:25] <Avenger> i wonder how this is supposed to work :)
[20:17:18] --> SiENcE has joined #GemRb
[21:04:57] <lynxlynxlynx> i'm back
[21:06:09] <lynxlynxlynx> Avenger: there's only a few (three?) how blockers right now and i'm not sure about one's validity
[21:07:05] <lynxlynxlynx> yeah, 3
[21:07:21] <lynxlynxlynx> http://gemrb.sourceforge.net/wiki/doku.php?id=todo#icewind_dale
[21:13:49] --> CIA-102 has joined #GemRb
[21:14:03] <Lightkey> eh
[21:16:48] <-- CIA-22 has left IRC (*.net *.split)
[21:19:54] <Avenger> what's up with the helmet?
[21:20:09] <Avenger> there are shoulder plates on that girl
[21:20:42] <Avenger> and how is this a blocker :)
[21:23:02] <Avenger> about hjolder: it has destroyself in it. Would minhp1 disable destroyself?
[21:23:37] <Avenger> because hjolder has it, and the original seems to remove minhp1.itm, i just found that recently :)
[21:25:34] <Avenger> hmm well, we don't care about minhp, so that should work
[21:25:47] <lynxlynxlynx> the bold ones are blockers
[21:28:23] <lynxlynxlynx> i don't remember exactly what the problem there was, but he definitely didn't self-destruct
[21:30:02] <Avenger> i think i know what's wrong with hjolder escape
[21:30:11] <Avenger> clearallactions clears the actions
[21:30:21] <Avenger> but... cutscene actions are probably not clearable
[21:30:39] <Avenger> at least, that's what i suspect
[21:31:45] --> SiENcE_ has joined #GemRb
[21:31:58] <Avenger> hmm wait, clearallactions wouldn't clear the sender's queue
[21:33:11] <-- SiENcE_ has left IRC (Client Quit)
[21:33:25] <Avenger> heh, i wonder when did we write this down, our code is supposed to do this
[21:34:19] <-- SiENcE has left IRC (Ping timeout: 240 seconds)
[21:35:57] <lynxlynxlynx> iirc both the wait and the spell fired too
[21:36:14] <lynxlynxlynx> iirc it also isn't an instant problem like with cespenar
[21:36:41] <Avenger> did he vanish?
[21:36:48] <Avenger> because the spell would do an animation removal
[21:36:49] <lynxlynxlynx> no
[21:36:59] <Avenger> spin133
[21:36:59] <lynxlynxlynx> i'll go recheck
[21:37:08] <Avenger> has opcode 260
[21:37:42] <Avenger> did the windwalk sound play?
[21:37:48] <Avenger> eff_m106
[21:38:20] <Avenger> or the portal appear?
[21:39:10] <lynxlynxlynx> now it just got stuck in the scene
[21:39:27] <Avenger> did it start the script?
[21:39:28] <lynxlynxlynx> i'm not sure if i reverted my fix correctly though
[21:39:36] <lynxlynxlynx> do you have a how install handy?
[21:39:43] <Avenger> yes
[21:39:46] <Avenger> i'm in windows
[21:39:47] <lynxlynxlynx> does it have hjescape.bcs in the override?
[21:39:50] <Avenger> yes
[21:39:54] <lynxlynxlynx> oh
[21:40:25] <lynxlynxlynx> i'll recompile it to be sure
[21:41:06] <Avenger> cutsceneid(myself) is a bit odd ;)
[21:44:16] <lynxlynxlynx> huh, the same thing again
[21:44:59] <fuzzie> problem with action clearing is, you break other games unless you implement everything properly
[21:45:18] <Avenger> yea
[21:45:35] <Avenger> but atm, it seems to be 'almost' exactly what i read in how
[21:45:52] <Avenger> in how, the sender's queue is not cleared, but everyone else in the same area is
[21:46:02] <fuzzie> for clearallactions? that sounds right
[21:46:03] <Avenger> we don't clear the sender's queue if it is an actor
[21:46:17] <Avenger> in the how code, i don't see any actorcheck
[21:47:10] <fuzzie> but that's just because we only clear actors
[21:47:13] <Avenger> anyway, this shouldn't be a problem, i'm pretty sure the sender is hjolder even if it is a cutscene stuff, i hope
[21:47:35] <fuzzie> the ValidTarget/Schedule checks seem strange too
[21:47:47] <Avenger> i cannot tell if this list clears all actor or all scriptables of the area
[21:48:10] <Avenger> it iterates through a list of area, but i didn't decode the area struct yet
[21:49:25] <lynxlynxlynx> i get the same result now
[21:49:35] <Avenger> same result as? freeze?
[21:49:35] <lynxlynxlynx> so i'm puzzled
[21:49:43] <lynxlynxlynx> yes
[21:49:54] <Avenger> you restarted gemrb?
[21:50:13] <lynxlynxlynx> i'm not sure
[21:50:26] <Avenger> lol if not, then it surely cached the script
[21:50:39] <lynxlynxlynx> it did find it in the right place
[21:50:55] <Avenger> yep, thought you surely recompiled it
[21:50:58] <lynxlynxlynx> reproducible
[21:51:04] <Avenger> ok, maybe the recompile didn't work?
[21:51:15] <lynxlynxlynx> weidu did create a bcs
[21:51:19] <Avenger> lol
[21:51:25] <Avenger> it compiled with bg2 flags
[21:51:37] <Avenger> you need to supply some flag to compile how script
[21:51:45] <Avenger> do you use it with dltcep?
[21:51:48] <lynxlynxlynx> no
[21:52:39] <Avenger> ok, i don't know what to do
[21:52:46] <Avenger> iwd2 has --script-style=iwd2
[21:53:03] <Avenger> but for how, dltcep just does --script-style=bg
[21:53:06] <Avenger> so i'm helpless
[21:53:32] <lynxlynxlynx> i'll do something else
[21:53:37] <Avenger> pst also has a script style
[21:53:43] <Avenger> but now iwd1/how
[21:53:47] <Avenger> now->not
[21:55:03] <Avenger> wait, i don't have any hjescape.bcs in override
[21:55:33] <Avenger> so that wasn't it
[21:55:40] <Lightkey> oh right, I ordered UK imports for the newly supported SCI games in ScummVM 1.2.0 and the shop also had the IE series ;-)
[21:55:44] <lynxlynxlynx> lynxlynx.info/ie/000000000-Auto-Save.zip
[21:55:53] <lynxlynxlynx> i didn't have it either in the first today's try
[21:56:03] <lynxlynxlynx> just go out of the cave there to start
[21:56:29] <Lightkey> fuzzie: what did you say about PST being hard to come by? there's lots on Amazon, I got it for 15
[21:57:54] <Avenger> meh it crashes in original
[21:58:28] <Avenger> recompiling gemrb
[21:58:55] <fuzzie> Lightkey: someone did a re-release fairly recently
[21:59:11] <fuzzie> and since it is on gog.com nowadays, i don't think anyone can claim it is hard to come by any more :P
[22:00:08] <fuzzie> i can only guess that the lawyers finally sorted out the rights
[22:01:54] <Avenger> i'm almost sure gog sells recompiled versions, so the sources were not lost
[22:02:56] <fuzzie> well, the only claim i've seen from bioware is that they don't have useful ToB source
[22:03:04] <fuzzie> i guess you heard more than that
[22:03:32] <Avenger> i was told that they 'don't have the sources anymore' :)
[22:03:40] <fuzzie> but things like the PST source are not going to be held by Bioware anyway..
[22:04:08] <Avenger> it might be that 1. it wasn't true 2. they passed the source on (bg1 is also on gog)
[22:04:34] <fuzzie> i can't think why they'd need source for bg1 or iwd, though :)
[22:05:11] <fuzzie> but i guess if the widescreen mod needed changes then that is interesting
[22:06:08] <Lightkey> oh, right
[22:06:08] <Maighstir> GOG's IE games contain at least a few of the same bugs as my earlier disc versions do; BG1 needs to be in windowed mode to start a network game on WinVista and Win7, PST wants 16-bit colour. I doubt they're recompiled, but merely patched up to the latest patch, and completed with a hex edit to no longer require the disc.
[22:06:41] <Lightkey> but I meant physical copies of course, I'm not interested in download only ;p
[22:06:42] <fuzzie> they're more different than that, it seems
[22:07:28] <fuzzie> maybe not recompiled, but not the same compile as any previously known exes
[22:07:42] <Avenger> lynx: hjescape.bcs is loaded, but doesn't complete for me
[22:08:20] <Avenger> i didn't even able to choose '1 -farewell' in the dialog!
[22:08:29] <Avenger> apparently the cutscene hits before i could end the dialog
[22:08:34] <Avenger> fuzzie?
[22:08:41] <lynxlynxlynx> it was a complete freeze here
[22:08:52] <Avenger> i typed: EndCutSceneMode()
[22:08:54] <lynxlynxlynx> even the animations stopped
[22:08:57] <Avenger> that brought back the gui
[22:09:05] <Avenger> sure, because in dialogs everything is frozen
[22:09:12] <lynxlynxlynx> oh, right
[22:09:16] <Avenger> and it is because the cutscene starts before the dialog ends
[22:09:30] <fuzzie> oh, well, there are new bugs causing dialogs to fail
[22:09:33] <Avenger> cutscene starts, but the dialog blocks nonblocking scripts
[22:09:40] <fuzzie> you can reproduce that with Avenger's loadscreen stuff too
[22:09:47] <Avenger> err, i mean, nonblocking actions run, but
[22:10:12] <Avenger> the problem is that nonblocking actions run in the dialog, one of the nonblocking actions is startcutscene :)
[22:10:23] <Avenger> then it disables the dialog
[22:10:27] <fuzzie> that should be a message?
[22:10:27] <Avenger> hehe
[22:10:43] <fuzzie> not sure that helps though
[22:10:49] <Avenger> dunno, i don't quite see why the actions are run before i hit '1-farewell'
[22:11:00] <fuzzie> which dialog?
[22:11:02] <Avenger> the actions should be in the response block of that dialog node, no?
[22:12:18] <Avenger> DHJOLLDE?
[22:13:04] <fuzzie> could be buggy parsing on our part, then
[22:13:07] <Avenger> starts with node #33
[22:13:27] <fuzzie> i hate the action parsing code, very buggy
[22:14:25] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:15:34] <Avenger> no, the cutscene starts in a middle node
[22:15:35] <Avenger> lol
[22:15:39] <fuzzie> but that's really strange
[22:15:52] <fuzzie> the dialog seems to parse fine
[22:15:56] <Avenger> node #116 has the actions
[22:16:00] <fuzzie> you're sure it's not being run by an area script?
[22:16:21] <Avenger> node #116 has the actions for startcutscene :)
[22:16:43] <Avenger> then hjolder says he shall defy banishment etc
[22:16:55] <Avenger> and you tell him 'farewell'
[22:16:59] <Avenger> i wonder how this works
[22:17:09] <Avenger> he basically destroys self, in middle of dialog
[22:17:24] <Avenger> ok, there is a wait code in it
[22:17:45] <Avenger> but how this is supposed to run a cutscene in middle of dialog then resume?
[22:18:09] <Avenger> i cannot run this in the original
[22:18:26] <Avenger> maybe, in the original you never get the chance to say farewell???
[22:18:46] <Avenger> i cannot see any way how this could be able to resume dialog after a cutscene
[22:19:42] <fuzzie> no?
[22:19:54] <Avenger> anyway, even if i do endcutscene, hjolder casts his little spell, but doesn't vanish
[22:20:14] <Avenger> if you make 'startcutscene' a blocking action
[22:20:23] <Avenger> then it will work
[22:20:28] <fuzzie> yes, but it is not
[22:20:29] <Avenger> it will wait till the dialog ends
[22:20:42] <fuzzie> so that just breaks other code :P
[22:20:50] <Avenger> hmm
[22:21:06] <Avenger> because other code would not wait to end dialogs?
[22:21:41] <fuzzie> i don't know about dialogs
[22:22:09] <fuzzie> oh damn, does iwd have different blocking for dialogs and non-dialogs?
[22:22:10] <Avenger> startcutscene is not in the instant.ids list
[22:22:13] <fuzzie> i know iwd2 has something like that
[22:22:25] <fuzzie> oh, well, surely it shouldn't run, then
[22:22:33] <Avenger> so it is supposed to be a blocking action
[22:22:44] <Avenger> do we handle instant.ids?
[22:22:46] <fuzzie> yes
[22:23:02] <fuzzie> if it's not in instant.ids, it doesn't get run until the dialog is done
[22:23:46] <fuzzie> but startcutscene is in my how instant.ids.
[22:23:49] <Avenger> we do it the wrong way maybe
[22:24:08] <Avenger> instead of doing af_instant, we probably should set af_blocking for non instants???
[22:24:11] <fuzzie> but as far as i remember, startcutscene is just a message?
[22:25:01] <fuzzie> oh, maybe not, the action looks huge
[22:25:26] <fuzzie> but i'm pretty sure the instant handling for dialogs is correct for bg1/bg2
[22:25:35] <Avenger> there are 2 messages in it, at least
[22:25:48] <fuzzie> i didn't check any other game.
[22:26:02] <Avenger> looking in HoW now
[22:26:43] <fuzzie> but i don't see the problem here
[22:27:10] <Avenger> ok, i should go sleep
[22:27:14] <Avenger> i didn't notice the time
[22:27:19] <fuzzie> isn't DestroySelf just delayed?
[22:27:25] <fuzzie> but ok, me neither, see you tomorrow :)
[22:27:39] <Avenger> destroyself didn't run even if i end the dialog
[22:27:42] <Avenger> the spell starts
[22:27:46] <Avenger> but doesn't run correctly
[22:27:49] <Avenger> bye
[22:28:03] <-- Avenger has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.10/20100914125854])
[22:32:01] <-- deepinthewoods has left #GemRb ("Leaving")
[23:10:06] --> pupnik has joined #GemRb
[23:34:06] --> |Cable| has joined #GemRb
[23:34:13] <-- |Cable| has left IRC (Remote host closed the connection)
[23:34:27] --> |Cable| has joined #GemRb