#gemrb@irc.freenode.net logs for 28 Dec 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:00:33] <-- LK|30C3 has left IRC (Ping timeout: 245 seconds)
[00:01:34] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[00:05:12] --> edheldil has joined #gemrb
[00:05:13] --- ChanServ gives channel operator status to edheldil
[00:08:43] <-- dolio has left IRC (Quit: Leaving)
[00:15:02] <-- edheldil has left IRC (Ping timeout: 246 seconds)
[00:26:26] <kpedersen> hey again. Thought I would let you know that gemrb is working great on openbsd. I managed to complete the tutorial with no problems at all.
[00:27:17] <kpedersen> The only thing to note was that `make -j2` caused a few linker errors and `make` was used instead.
[00:28:13] <kpedersen> I will see if I can find a few fixes and report back. Until then, I have a Yoshimo to backstab!
[00:29:19] --> dolio has joined #gemrb
[00:50:25] --> LK|30C3 has joined #gemrb
[00:56:07] <-- raevol has left IRC (Quit: Leaving.)
[01:20:09] <brada> we dont support make so i doubt anybody cares
[01:39:19] <Pepelka> [wiki] pst_bugs - [Torment General Bugs] Introducing: Tormenthacks http://www.gemrb.org/wiki/doku.php?id=pst_bugs&rev=1388194719&do=diff
[01:40:46] --> chiv has joined #gemrb
[01:41:31] <chiv> ^ I've designated myself a sandbox to play in as I work my way through the torment bugs I find
[01:43:03] <chiv> I can't always dwell on finding a perfect solution to each problem, but this makes it easier for me to contribute in the long run
[01:43:44] <brada> Im hoping we can talk fuzzie into helping you in january
[01:44:13] <brada> many of your fixes are close but not quite
[01:44:43] <chiv> yeah i'm sorry about that, I want to do better but I am usually at my limit of how much I can comprehend
[01:45:09] <brada> i find that hard to believe
[01:45:18] <brada> what you have figured out is nontrivial
[01:45:29] <brada> and the remaining bits are much simpler
[01:45:49] <chiv> heh, it's 90% motivation based I assure you
[01:45:50] <-- LK|30C3 has left IRC (Ping timeout: 252 seconds)
[01:46:13] <brada> it doesnt matter it still counts :)
[02:18:44] --> LK|30C3 has joined #gemrb
[02:33:27] <-- LK|30C3 has left IRC (Ping timeout: 260 seconds)
[02:43:43] --> LK|30C3 has joined #gemrb
[03:36:43] <-- dolio has left IRC (Quit: Leaving)
[03:51:42] <-- brada has left IRC (Quit: brada)
[03:53:23] --> dolio has joined #gemrb
[04:01:08] <-- |Cable| has left IRC (Ping timeout: 246 seconds)
[04:14:04] --> |Cable| has joined #gemrb
[04:26:41] --> brada has joined #gemrb
[06:30:19] <-- brada has left IRC (Quit: brada)
[06:31:50] <-- LK|30C3 has left IRC (Remote host closed the connection)
[08:54:15] <fuzzie> chiv: the secret is, no-one understands it all :/
[09:29:18] --> Beholder has joined #gemrb
[09:30:00] <Beholder> hi
[09:31:44] <Beholder> anyone answer me, may be that few sprites with different palettes draws in one frame?
[09:32:12] <Beholder> or all sprites draws with one palette
[10:07:58] --> lynxlynxlynx has joined #gemrb
[10:07:58] --- ChanServ gives channel operator status to lynxlynxlynx
[10:15:21] <Beholder> anyone answer me, may be that few sprites with different palettes draws in one frame?
[10:18:01] <fuzzie> I odn't understand the question
[10:18:17] <fuzzie> every sprite has its own palette
[10:20:03] <fuzzie> you want to know how many have the same one?
[10:27:06] <Beholder> right, is there palette animation gemrb?
[10:27:10] <Beholder> in gemrb
[10:37:19] <fuzzie> yes
[10:37:38] <fuzzie> per-sprite
[10:45:44] <lynxlynxlynx> chiv: did you rest recently in pst?
[10:50:17] <chiv> no, but it was broken last time i checked
[10:50:37] <lynxlynxlynx> yeah, same here
[10:51:09] <lynxlynxlynx> do you remember if you got any of the memories back by resting?
[10:51:09] <chiv> it uses a different scheme to decide where you can rest
[10:51:27] <chiv> as far as I remember, no that doesn't happen
[10:51:46] <chiv> but I could be wrong and it just never happened to me...
[10:53:17] <chiv> memories are usually provoked by lookint at certain things or following certain conversation paths
[10:54:02] <chiv> I don't remember any bg1/2 style 'freebies'
[10:56:29] <lynxlynxlynx> ok
[10:56:50] <lynxlynxlynx> everyone seemed to go fine through the loop
[10:59:34] <lynxlynxlynx> yeah, it works, there's just no visual feedback
[11:02:06] <chiv> resting? although what was broken isn't that you can't reset, its just you can reset in all the wrong places
[11:02:34] <chiv> ugh. rest not reset
[11:15:03] <-- chiv has left IRC (Quit: Page closed)
[11:25:22] <kpedersen> brada, I meant when using `make` after the `cmake ..` has generated it.
[11:28:01] <Beholder> in DrawSprite x and y is a screen coordinate, and Region is a part of sprite that we need to blit? if Region == null we blit a full sprite. Right?
[11:29:48] <wjp> x,y are screen coordinates if anchor is true
[11:29:53] <wjp> otherwise they're viewport coordinates
[11:31:09] <wjp> assuming you mean BlitSprite
[11:31:45] <wjp> and if by Region you mean 'clip', then that is the clipping rect in screen coordinates
[11:44:36] <Beholder> computeClipRect computes a sprite region for blit?
[12:05:14] <-- Beholder has left #gemrb
[12:09:37] <-- WingedHussar has left IRC (Ping timeout: 272 seconds)
[12:10:42] --> WingedHussar has joined #gemrb
[12:20:05] <-- WingedHussar has left IRC (Quit: WingedHussar)
[12:48:55] <wjp> no, the final clipping rect, still in screen coordinates. It takes into account the passed 'clip' parameter, the global clipping rect set in the video driver, the screen, and the sprite dimensions
[12:49:15] <wjp> (for the logs)
[13:09:07] <Pepelka> [wiki] news - created http://www.gemrb.org/wiki/doku.php?id=news&rev=1388235916&do=diff
[13:25:24] <lynxlynxlynx> how do you feel about a bit of cheating? I'm considering copying one column from iwd2's original skills.2da to our skillstat.2da just to avoid having the core need to read the first.
[13:25:40] <lynxlynxlynx> it's only used in chargen and levelup otherwise and that is all on the guiscript side
[13:33:59] <fuzzie> denormalizinglynx!
[13:35:04] <wjp> does nobody ever override it?
[13:41:58] <lynxlynxlynx> not that i heard of
[13:42:15] <lynxlynxlynx> iwd2 mods are rare as it is and this is about trained/untrained skill distinction
[13:42:30] <lynxlynxlynx> not too sexy to play with
[13:47:50] <wjp> is Beholder working on a new renderer, by the way?
[13:49:51] <lynxlynxlynx> i think he's just trying to increase its performance
[14:06:41] --> chiv has joined #gemrb
[14:26:24] <fuzzie> improving BlitSprite would be an obvious place to start I guess
[14:55:44] <lynxlynxlynx> i'll take that as an "ok" and go ahead
[15:15:54] --> brada has joined #gemrb
[15:26:20] <fuzzie> you're in the best place to decide?
[15:29:48] <brada> im finding myself annoyed that you cant click+hold scrollbarbuttons in gemrb, and it seems non-trivial to implement :(
[15:30:43] <lynxlynxlynx> who clicks on those buttons to scroll? oO
[15:31:19] <lynxlynxlynx> if you don't have a wheel, clicking on the slider or even dragging the odd-word-i-wont-remember is more efficient
[15:31:50] <brada> dragging is annoying because if you leave the control bounds it stops
[15:31:52] <lynxlynxlynx> and i think we already have what we need, see the RK_ repeat flags
[15:32:06] <brada> i have seen it
[15:32:28] <wjp> if dragging stops outside of the control bounds we should probably fix that
[15:36:03] <brada> im not sure how
[15:36:27] <brada> ill look at it
[15:43:57] --> Beholder has joined #gemrb
[15:45:44] <Beholder> what format used for non-paletted images (such as background)? RGBA 8 bit per pixel?
[15:46:13] <Beholder> i mistakes 32 bit per pixel
[15:46:22] <Beholder> 8 bit per component
[15:46:47] <wjp> background tiles have a palette
[15:47:39] <wjp> (see BlitTile)
[15:50:48] <wjp> I think for all non-paletted images we currently force them to 32bpp on load?
[15:51:35] <wjp> yes
[15:52:03] <brada> you shouldnt rely on any specific format anyway.
[15:52:16] <brada> for non paletted sprites
[15:53:25] <brada> tho it does appear BlitGameSprite does require a specific format
[15:53:35] <brada> er at least bit depth
[15:53:52] <Beholder> BlitTile not used for drawing background in menu
[15:54:01] <Beholder> BlitSprite is used
[15:54:28] <wjp> ah, menu
[15:54:37] <wjp> that's MOS then I think? That's loaded as 32bpp
[16:00:57] <brada> what can i do to make this less terrible: http://pastebin.com/yFdQga2H
[16:00:58] <Pepelka> [Diff] diff --git a/gemrb/core/GUI/EventMgr.cpp b/gemrb/core/GUI/EventMgr.cpp index d4 - Pastebin.com
[16:01:05] <brada> and what did I break?
[16:01:57] <wjp> multiple mouse downs probably
[16:03:13] <wjp> hm, it's been too long since I looked at this code. A name like 'focusLock' or such is probably clearer than eventTarget
[16:03:32] <brada> last_win_over->OnMouseOver(x, y); seems better choice too
[16:03:51] <brada> works anyway :p
[16:04:32] <brada> double click still seems to work if thats what you mean
[16:04:57] --> Eli2 has joined #gemrb
[16:04:59] <wjp> I meant sequences like left-down right-down left-up right-up
[16:05:37] <wjp> that seems to confuse GTK quite a bit too though :-)
[16:06:22] <brada> i probably should limit this to left button anyway
[16:06:40] <wjp> will all OnMouseOver handlers properly handle mouse moves outside of their usual region?
[16:06:40] <Beholder> odd, i trying to draw menu background with GLES, but image displayed with wrong colors
[16:07:16] <wjp> Beholder: the MOS importer has its own pixel format. See MOSImporter.cpp
[16:07:35] <brada> wjp: i have no idea about "all"
[16:07:45] <brada> but this does uncover a small bug with the scrollbar
[16:07:55] <brada> if you drag above the control it jumps to the bottom
[16:08:09] <brada> unsigned
[16:08:14] <brada> i imagine
[16:08:35] <wjp> yes, that does look broken
[16:08:44] <Beholder> wjp: thank you
[16:09:03] <wjp> Beholder: that format should get properly set in the SDL PixelFormat for that sprite though
[16:13:54] <brada> exactly
[16:14:17] <brada> you can get the formato out of the SDL sprites
[16:15:36] <brada> since you need to implement shaders for game sprite effects you loose nothing by dropping support for BAM sprites
[16:15:50] <-- |Cable| has left IRC (Ping timeout: 245 seconds)
[16:24:33] <-- Eli2 has left IRC (Remote host closed the connection)
[16:29:00] <brada> wjp: I havent been able to identify any problems with events out of bounds other than the unsigned problem
[16:29:16] <brada> did uncover a regression i introduced :/
[16:29:34] --> |Cable| has joined #gemrb
[16:44:42] <brada> have we never supported the little buttons on the map
[16:45:02] <brada> I thought i broke them, but checking out to before my changes reveals they still dont work
[16:46:47] --> kpederse1 has joined #gemrb
[16:47:23] <brada> does anybody have a problem with this map control behavior change: http://pastebin.com/P6dfpiCm
[16:47:23] <Pepelka> [Diff] diff --git a/gemrb/core/GUI/MapControl.cpp b/gemrb/core/GUI/MapControl.cpp inde - Pastebin.com
[16:49:29] <-- kpedersen has left IRC (Ping timeout: 252 seconds)
[16:50:46] <lynxlynxlynx> you mean the directional arrows?
[16:51:08] <brada> i mean the little button markers on the map control
[16:51:17] <brada> the location markers i guess
[16:51:44] <lynxlynxlynx> of course they worked, you can't travel without them in half of the games
[16:51:53] <lynxlynxlynx> what's wrong?
[16:52:13] <brada> wrong map
[16:52:20] <brada> not world map
[16:52:51] <brada> the ones that mark places of interest in the current area
[16:53:27] <lynxlynxlynx> less certain, but i think they did
[16:53:42] <lynxlynxlynx> btw, how would you navigate the worldmap without mouse dragging?
[16:54:12] <brada> that isnt for the world map ;)
[16:54:17] <brada> and it doesnt break dragging
[16:54:31] <lynxlynxlynx> oh yes, it definitely did work, now we don't even display the predefined ones
[16:54:51] <brada> it just makes it so the viewport moves on mouse down and initiates dragging
[16:55:06] <brada> did it break or was it removed?
[16:55:19] <brada> im assuming the prior
[16:55:24] <lynxlynxlynx> heh
[16:55:41] <lynxlynxlynx> don't test in bg1, since it didn't have them
[16:55:59] <lynxlynxlynx> we also have redrawing issues with the title now :s
[16:56:11] <brada> thats the regression i mentioned
[16:56:17] <brada> but that patch i posted fixes that too
[16:56:32] <brada> do you see a problem with that behavior modification?
[16:57:17] <lynxlynxlynx> not really
[16:58:36] <lynxlynxlynx> hmm, the main game viewport? i don't understand how it's different then
[16:59:34] <lynxlynxlynx> one click less, i guess
[17:00:16] <brada> its the behavior *I* expect
[17:00:59] <brada> i expect to be able to slick and drag anywhere on the map
[17:01:05] <brada> not just within the green rectangle
[17:01:52] <lynxlynxlynx> fine with me
[17:02:05] <lynxlynxlynx> who uses drag in there anyway? :P
[17:02:11] <brada> me ha ha
[17:03:25] <brada> lynx: so should i bisect the map markers? or do you know when/why the are gone?
[17:03:58] <lynxlynxlynx> they're just not drawn, so my guess would be your recent changes
[17:04:53] <brada> no
[17:04:59] <brada> I checked out before as I said
[17:05:02] <brada> they still dont work
[17:32:18] <lynxlynxlynx> they do, you can open them
[17:32:28] <lynxlynxlynx> also add them, though i can't verify
[17:32:41] <lynxlynxlynx> oh, as in when it failed
[17:34:19] <lynxlynxlynx> ok, they're not displayed in iwd2 either, so it must be something more general
[17:34:48] <lynxlynxlynx> no idea, bisect away
[17:51:06] <lynxlynxlynx> cool, none of the dialogs in DREAM_SCRIPT_FILE of pdialog.2da even exist
[17:55:03] <lynxlynxlynx> ... because they're scripts
[17:56:27] <lynxlynxlynx> that don't matter
[18:02:20] <lynxlynxlynx> eh, some actually exist
[18:35:58] <Pepelka> [wiki] pst_bugs - [Inventory] item restrictions are fun http://www.gemrb.org/wiki/doku.php?id=pst_bugs&rev=1388255360&do=diff
[18:39:08] <lynxlynxlynx> fuzzie: any comments? http://sprunge.us/NcZJ?diff
[18:40:19] <lynxlynxlynx> it's not optimised, but i'm seeking feedback on the general idea
[18:40:37] <lynxlynxlynx> is that the right way to inject a script?
[18:42:35] <fuzzie> does it work?
[18:43:27] <lynxlynxlynx> haven't tried yet
[18:44:47] <fuzzie> it seems fine if it does
[18:45:38] <fuzzie> but you'll end up with the actions being executed too late I guess
[18:47:30] <fuzzie> is that ok?
[18:47:46] <lynxlynxlynx> you mean in the "morning"?
[18:48:03] <lynxlynxlynx> it either doesn't work or none of my saves have nice enough npcs
[18:48:12] <fuzzie> yes, in the morning
[18:52:35] <lynxlynxlynx> it is looking up the right script, so yeah, i'll have to fake something
[18:52:43] <lynxlynxlynx> the only thing that i see are romances
[18:52:56] <lynxlynxlynx> and some of that needs to happen before rest to make sense
[19:02:08] <lynxlynxlynx> is there a better way to get it done than putting this bit into another guiscript function and running the actual restparty with settimedevent a tick later?
[19:03:18] <lynxlynxlynx> the other idea is using the restparty script action with a wait
[19:03:51] <lynxlynxlynx> but that would also require special casing and not be particularly nice
[19:39:51] --> dindolo has joined #gemrb
[19:42:07] <-- dindolo has left IRC (Client Quit)
[20:47:46] <lynxlynxlynx> yep, works fine the python way
[20:55:54] --> Yoshimo has joined #gemrb
[21:13:59] <-- Beholder has left #gemrb
[21:28:57] <chiv> torment hacks branch doubled in size today, many problems zeroed in on :)
[21:33:50] <lynxlynxlynx> :)
[21:33:58] <chiv> fixes are mostly duct tape and crazy glue, but I very fast approaching having a working game
[21:34:17] <lynxlynxlynx> anything to look at?
[21:34:42] <chiv> https://github.com/chilvence/gemrb/compare/tormenthacks <emphasis on 'hacks'
[21:34:45] <Pepelka> Comparing gemrb:master...chilvence:tormenthacks · chilvence/gemrb · GitHub
[21:34:46] <Pepelka> »gemrb - Engine Made with preRendered Background«
[21:41:43] <lynxlynxlynx> the dakkon blade and the door trigger one are ok as-is
[21:42:10] <chiv> ok, I'll send them
[21:42:38] <lynxlynxlynx> we already generate all the various death variables, so something else is probably wrong there
[21:43:26] <chiv> the problem seems the variable is "X_DEAD" and it only looks for "X" in the trigger
[21:44:22] <lynxlynxlynx> what is the problem with the worldmap palette?
[21:44:34] <chiv> it's all inverted
[21:44:41] <lynxlynxlynx> we have something in other games too, but ... not that
[21:44:43] <-- Yoshimo has left IRC (Ping timeout: 245 seconds)
[21:44:45] <chiv> and so are lots of actors
[21:45:41] <chiv> i don't know if that is the files or the code though
[21:47:20] <lynxlynxlynx> maybe some extra flag somewhere
[21:47:47] <lynxlynxlynx> usual problems would be clown colors or nothing at all, so this is new to me
[21:48:01] <chiv> are the worldmap icons the same format as the actor sprites?
[21:48:35] <-- PKodon has left IRC (Quit: The Rodent Tracker 8000, just like on TV ... Because household pests never build up an immunity to bullets. (Tex Murphy))
[21:50:04] <lynxlynxlynx> i think so
[21:50:22] <lynxlynxlynx> you were right about disguises, for some reason this is handled via global vars
[21:51:19] <chiv> I've not actually tested wether the disguises are respected by the dustmen yet
[21:51:54] <chiv> there's also the thing where you aren't supposed to run in a disguise, because they didn't make the sprites...
[21:56:23] <chiv> looks like dustman robes do at least
[21:58:37] <chiv> yep. that all seems to work
[21:59:55] <chiv> which means the first level is almost 100% faithful
[22:02:44] <lynxlynxlynx> apparently it's only called for setting the zombie disguise, never to cancel
[22:03:18] <lynxlynxlynx> how did you end it in the original?
[22:04:11] <chiv> have to test. usually, it stops when you run, and probably doesn't last outside mortuary.
[22:06:41] <lynxlynxlynx> thanks
[22:07:20] <lynxlynxlynx> i think we're missing something ugly from the original for this
[22:07:27] <chiv> you can also give yourself away if you talk to any of the dustmen
[22:07:46] <chiv> i've just noticed you can actually run with the robes on
[22:08:28] <lynxlynxlynx> we have some code to deal with the robe and that affects the appearance
[22:09:40] <lynxlynxlynx> ah, the dialogs themselves play with it too
[22:10:05] <lynxlynxlynx> i doubt there are area restrictions
[22:10:55] <lynxlynxlynx> several check if you are disguised
[22:12:05] <chiv> I will have to analyse it on my next playthrough, but for now the actual quest side of the disguises seems to all work
[22:12:17] <chiv> just the visual sprite is messed up
[22:13:12] <chiv> oh and vhailor turns into your clone because of this too
[22:24:17] <lynxlynxlynx> try this for me instead: http://sprunge.us/FHCj?diff
[22:29:52] <lynxlynxlynx> interesting, we already have a note in the importer about pst differences for usepoint
[22:37:57] <chiv> ok that works for both disguises ^, but vhailor still changes to TNO
[22:41:42] <lynxlynxlynx> that's odd, is he always like that?
[22:42:01] <lynxlynxlynx> btw, re infopoint talking, did you try with the talkpos too?
[22:42:20] <chiv> no but he initially starts with an inactive animation state and then switches
[22:44:38] <chiv> I tried every value I could think of on the infopoints, the only one that works for ragpicker square is calculating the middle
[22:45:03] <lynxlynxlynx> so did that include it or not - if you remember?
[22:45:17] <lynxlynxlynx> pst is the only one with it set and right next to the dialog ref
[22:45:41] <lynxlynxlynx> it's odd that vhailor changes at all, we don't seem to have PolymorphEx mapped to anything
[22:47:45] <lynxlynxlynx> and one call is with 0 in place of the animation id
[22:47:49] <lynxlynxlynx> g-reat
[22:49:58] <chiv> it really isn't helping me that dltcep labels things differently to ni, but if you look at from_portal_trigger in ar0101, both of the points are on the unreachable side of the hitmap
[22:50:45] <lynxlynxlynx> if it's even displaying that point
[22:51:08] <chiv> well according to ni and dltcep
[22:51:08] <lynxlynxlynx> i'll just commit and we'll see what happens
[22:51:49] <lynxlynxlynx> i would translate from_trigger to traplaunch
[22:52:01] <chiv> 1236.168. which reminds me, when you press L, it gives the screen coordinates and not the map coordinates
[22:52:02] <lynxlynxlynx> too lazy to check either source
[22:52:17] <lynxlynxlynx> there's a button for that too
[22:52:35] <lynxlynxlynx> x? ctrl-j will print it too
[22:54:37] <lynxlynxlynx> PolymorphEx("""""(cl 1) Wererat""""",1) <-- wtf, i hope this is weidu being silly
[22:55:18] <chiv> anyway I tested it by actually moving the whole polygon, which produced the expected result in torment.exe
[23:00:41] <chiv> doh, I actually meant to type 1236.628 looking back at it
[23:04:57] <chiv> are you looking at 1500soeg?
[23:05:30] <lynxlynxlynx> at nothing
[23:05:59] <chiv> wheres the wierd polymorph trigger from?
[23:06:25] <lynxlynxlynx> several scripts use it, 4 of which are from vhailor
[23:07:34] <lynxlynxlynx> ../../scripts/pst/0702vhai.baf ../../scripts/pst/1204vhai.baf
[23:07:34] <lynxlynxlynx> ../../scripts/pst/1202vhai.baf ../../scripts/pst/pcvhail.baf
[23:08:16] <lynxlynxlynx> two more actually
[23:29:17] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:58:46] <-- chiv has left IRC (Ping timeout: 272 seconds)