#gemrb@irc.freenode.net logs for 20 Feb 2012 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:04:15] <lostLinSoul> Maybe also use D_/DE_ and M_/MA_ for #defines and macros respectively to make them obvious from constants
[00:04:37] <lostLinSoul> ?
[00:10:45] <CIA-28> GemRB: 03edheldil * r577045c31d30 10gemrb/gemrb/ (3 files in 2 dirs):
[00:10:46] <CIA-28> GemRB: Add Button.SetPushOffset() and Button.SetAnchor()
[00:10:46] <CIA-28> GemRB: * new method for setting amount of label & pic movement
[00:10:46] <CIA-28> GemRB: when button is pressed, defaults to 2,2
[00:10:46] <CIA-28> GemRB: * GuiScript API for SetAnchor, still does not do anything
[00:21:14] <edheldil_> oops, forgot to add the files :)
[00:24:07] <edheldil_> argh, I have git reset --hard the changed files as well, woe on me!
[00:24:13] <lostLinSoul> LOL
[00:24:14] * edheldil_ facepalms
[00:25:05] <lostLinSoul> I guess that is going to be a pain
[00:25:38] <edheldil_> nah, it was not much code, but I feel so lame
[00:26:01] <lostLinSoul> :0)
[00:28:24] <gembot> build #554 of cmake clang++ is complete: Failure [4failed minimal test] Build details are at http://buildbot.gemrb.org/builders/cmake%20clang%2B%2B/builds/554 blamelist: edheldil@users.sourceforge.net
[00:30:04] <tomprince> gembot: mute
[00:30:04] <gembot> Shutting up for now.
[00:30:22] <tomprince> edheldil_: Do you want unmute gembot once its green again?
[00:34:20] <edheldil_> I will unmute it when I commit the missing files. Should take about 20 minutes. Should I do a new commit or should I do revert?
[00:35:32] <tomprince> New commit.
[00:40:57] <edheldil_> that will teach me next time not to be lazy and do git checkout FILE instead of git reset --hard :-)
[00:43:02] <tomprince> 'git checkout -p' is nice
[00:54:48] <edheldil_> gembot: unmute
[00:54:48] <gembot> I'm baaaaaaaaaaack!
[00:55:14] <CIA-28> GemRB: 03edheldil * r5410ae3daaf7 10gemrb/gemrb/ (3 files in 2 dirs):
[00:55:14] <CIA-28> GemRB: Add Button.SetPushOffset() and Button.SetAnchor() (add missing files)
[00:55:14] <CIA-28> GemRB: * new method for setting amount of label & pic movement
[00:55:14] <CIA-28> GemRB: when button is pressed, defaults to 2,2
[00:55:14] <CIA-28> GemRB: * GuiScript API for SetAnchor, still does not do anything
[01:13:23] <gembot> build #555 of cmake clang++ is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20clang%2B%2B/builds/555
[01:13:32] <gembot> build #512 of mingw32 is complete: Failure [4failed minimal test] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/512 blamelist: edheldil@users.sourceforge.net
[01:16:54] <gembot> build #167 of cmake g++-4.2 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.2/builds/167
[01:17:12] <gembot> build #175 of osx-test is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/osx-test/builds/175
[01:18:12] <lostLinSoul> Night All
[01:20:40] <gembot> build #168 of cmake g++-4.4 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/168
[01:28:02] <gembot> build #161 of cmake g++-4.5 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.5/builds/161
[01:51:22] --> nutron|w has joined #gemrb
[01:55:47] <edheldil_> hmm, do I have to do st. else for the mingw32 target?
[01:58:25] <CIA-28> GemRB: 03edheldil * r2803c55d298f 10gemrb/gemrb/GUIScripts/pst/FloatMenuWindow.py:
[01:58:25] <CIA-28> GemRB: Improve PST's FloatingMenu display and behavior
[01:58:25] <CIA-28> GemRB: * Open the right windows on single person actions
[01:58:25] <CIA-28> GemRB: * Make the buttons push more like in original
[01:58:25] <CIA-28> GemRB: * Fixed slot buttons when jumping between dialog / items / actions / ...
[01:58:26] <CIA-28> GemRB: * Use constants for control ids
[02:01:02] <tomprince> gembot: force build mingw32
[02:01:04] <gembot> build forced [ETA 15m42s]
[02:01:04] <gembot> I'll give a shout when the build finishes
[02:13:12] <gembot> build #514[None] of mingw32 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/514
[02:13:45] <edheldil_> oh, thanks
[02:13:56] <edheldil_> night
[02:17:27] --> joneirik has joined #gemrb
[02:18:10] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[02:40:28] <gembot> build #556 of cmake clang++ is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/cmake%20clang%2B%2B/builds/556 blamelist: edheldil@users.sourceforge.net
[02:40:34] <gembot> build #168 of cmake g++-4.2 is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.2/builds/168 blamelist: edheldil@users.sourceforge.net
[02:40:38] <gembot> build #169 of cmake g++-4.4 is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/169 blamelist: edheldil@users.sourceforge.net
[03:03:31] <gembot> build #230 of nmake-msvc++10 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B10/builds/230
[03:13:07] <-- lostLinSoul has left IRC (Quit: Page closed)
[03:21:57] <gembot> build #503 of msvc++6 is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/msvc%2B%2B6/builds/503 blamelist: edheldil@users.sourceforge.net
[06:27:07] <-- joneirik has left IRC (Remote host closed the connection)
[06:27:36] --> edheldil_ has joined #gemrb
[07:54:35] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[07:58:31] --> lynxlynxlynx has joined #gemrb
[07:58:31] <-- lynxlynxlynx has left IRC (Changing host)
[07:58:31] --> lynxlynxlynx has joined #gemrb
[07:58:31] --- ChanServ gives channel operator status to lynxlynxlynx
[08:47:36] --> edheldil_ has joined #gemrb
[08:53:32] <-- edheldil_ has left IRC (Ping timeout: 272 seconds)
[08:56:00] <edheldil> Grr, I see that buildbot blames on me everything including network problems :)
[08:56:16] --> wortek has joined #gemrb
[08:56:21] <wortek> Hi all
[08:56:25] <edheldil> hi
[08:56:34] <wortek> Hey guys can you help me out
[08:57:00] <edheldil> don't ask to ask, just ask :)
[08:57:10] <wortek> I'am going to buy Galaxy Note soon,did anyone tried to rung GemBR on Galaxy Note?
[08:57:31] <wortek> :)
[08:57:58] --> alx3apps has joined #gemrb
[08:58:11] <edheldil> it's an Android, right? Ask on or forum in the android thread
[08:58:21] <wortek> I just want to make sure it runs smoothly on Galaxy Note
[08:58:34] <wortek> Yeah it's Android
[08:59:35] <edheldil> You will get the best android advice there, the android dev does not hang here much
[09:00:56] <alx3apps> wortek: I just logged in - missed your previous questions. It should run fine on Galaxy note. I'm not an android dev (Beholder is), I just build gemrb for android and can answer questions about android setup
[09:01:18] <wortek> great i hope it will work
[09:01:42] <wortek> I'll try to run all games on it and make a video about succesfull ones
[09:02:30] <edheldil> great
[09:03:09] <-- wortek has left IRC (Quit: Page closed)
[09:13:13] --> haad has joined #gemrb
[11:24:57] <-- lynxlynxlynx has left IRC (Ping timeout: 252 seconds)
[12:21:22] --> lynxlynxlynx has joined #gemrb
[12:21:22] <-- lynxlynxlynx has left IRC (Changing host)
[12:21:22] --> lynxlynxlynx has joined #gemrb
[12:21:22] --- ChanServ gives channel operator status to lynxlynxlynx
[12:52:24] --> SiENcE has joined #gemrb
[13:36:15] <-- Gekz_ has left IRC (Ping timeout: 252 seconds)
[13:42:47] --> Gekz has joined #gemrb
[13:42:47] <-- Gekz has left IRC (Changing host)
[13:42:47] --> Gekz has joined #gemrb
[13:57:14] <edheldil> if anyone's interested: floating menu in ps:t uses a color filter as well, this one from color to grayscale
[13:58:56] <wjp> does the blitter need new features?
[14:01:45] --> lostLinSoul has joined #gemrb
[14:01:57] <edheldil> I am not sure whether blitter - it might just prepare the gray sprites in advance, like fog-of-war
[14:02:27] <edheldil> the blitter does the tints on the fly?
[14:02:38] <fuzzie> for BAMs
[14:02:46] <fuzzie> since that is required
[14:03:49] --> edheldil_ has joined #gemrb
[14:03:51] <edheldil> I can't check at the moment what it does
[14:04:25] <edheldil> is it in core or in SDLVideo?
[14:09:39] <edheldil> I will check it wgen I can
[14:12:46] <lostLinSoul> It have a try...except statement in GUICommonWindows.py which still prints '[GUIScript]: Runtime Error:' to the console
[14:13:42] <lostLinSoul> Can I ignore that, or do I either try to fix it or get it fixed by someone more capable than I?
[14:20:21] --> kettuz has joined #gemrb
[14:22:43] <edheldil> lostLinSoul: what's your problem?
[14:23:44] <lostLinSoul> JUst as stated, there is no body to the '[GUIScript]: Runtime Error:', just that text
[14:26:58] <edheldil> what's in the except clause?
[14:27:42] <tomprince> gembot: force build msvc++6
[14:27:44] <gembot> build forced [ETA 4m37s]
[14:27:44] <gembot> I'll give a shout when the build finishes
[14:27:52] <gembot> Hey! build msvc++6 #504 is complete: Failure [4failed]
[14:27:52] <gembot> Build details are at http://buildbot.gemrb.org/builders/msvc%2B%2B6/builds/504
[14:27:58] <tomprince> gembot: force build g++-4.4
[14:27:58] <gembot> no such builder 'g++-4.4'
[14:28:12] <lostLinSoul> Try to retied Button with window.getcontrol and if it fails, create the button as I have not already created one the specified control id
[14:28:20] <tomprince> gembot: force build 'cmake g++-4.4'
[14:28:21] <gembot> build forced [ETA 3m42s]
[14:28:21] <gembot> I'll give a shout when the build finishes
[14:28:24] <gembot> Hey! build cmake g++-4.4 #170 is complete: Failure [4failed]
[14:28:24] <gembot> Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/170
[14:28:46] <lostLinSoul> retied -> retrieve
[14:29:28] <edheldil> post the code somewhere or whisper it to me
[14:31:17] <tomprince> gembot: force build g++-4.4
[14:31:17] <gembot> no such builder 'g++-4.4'
[14:31:21] <tomprince> gembot: force build 'cmake g++-4.4'
[14:31:22] <gembot> build forced [ETA 3m42s]
[14:31:22] <gembot> I'll give a shout when the build finishes
[14:31:24] <gembot> Hey! build cmake g++-4.4 #171 is complete: Failure [4failed]
[14:31:24] <gembot> Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/171
[14:33:02] <lostLinSoul> edheldil: http://paste.debian.net/156911/
[14:34:14] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[14:35:06] <tomprince> Wierd. My buildslave seems blocked, for some reason.
[14:47:49] <edheldil> replace the try/except with if block. I suspect that the exception is thrown in the except: clause
[14:50:39] <wjp> lostLinSoul: are you sure those controlIDs are "free"?
[14:50:55] <wjp> (i.e., not already in use for something else)
[14:51:43] <wjp> but I should read the comments before saying anything I suppose :-)
[14:52:50] --> lynxlynxlynx has joined #gemrb
[14:52:50] <-- lynxlynxlynx has left IRC (Changing host)
[14:52:50] --> lynxlynxlynx has joined #gemrb
[14:52:50] --- ChanServ gives channel operator status to lynxlynxlynx
[14:55:53] <lostLinSoul> edheldil: If I do that the runtime error will cause the function/def or whatever it's called it python, from running and they prevents the portrait window from showing up
[14:58:05] <wjp> the idea is to fix the problem, not hide it :-)
[14:58:12] <lostLinSoul> wjp: From what I've seen, the control id's appear to start from zero for each window and for the portrait window, that stops at 8/9. Unless someone else has also been manipulating that window to add new controls, 15 should be safe
[14:59:53] <lostLinSoul> Not trying to hide it. Is there is no control defined and I have not yuet created a button for the given control id, then it should have and except as I try to get something that do not yet exist.
[15:00:22] <lostLinSoul> Is there is no -> If there is no
[15:02:24] <fuzzie> yes, but you shouldn't be trying to get buttons for controls which aren't there yet
[15:03:11] <edheldil> Ah, I see: that error message is done in RuntimeError(), even though the exception is caught. MoreOver, the printMessage() call in RuntimeError() and AttributeError() does not contain the message
[15:03:29] <lostLinSoul> '[GUIScript]: Runtime Error:' is a print statement in 'RuntimeError()' of GUIScript.cpp
[15:04:02] <lostLinSoul> Before PyErr_SetString( PyExc_RuntimeError, msg );
[15:04:18] <edheldil> so it's harmless
[15:04:37] <edheldil> but we should reduce the error spam
[15:04:46] <lostLinSoul> That's it!
[15:05:36] <lostLinSoul> I'm guessing the printMessage is used for colouring the text
[15:06:12] <lostLinSoul> Otherwise it could be included in the PyErr_SetString
[15:07:23] <lostLinSoul> But as you say it's spamming unnecessary messages of '[GUIScript]: Runtime Error:'
[15:09:29] <lostLinSoul> Or at least it will once I created the portrait button for that control id, it will con complain for it again - but for any other new portrait not having a control id in CHU it will give the message again
[15:09:51] <lostLinSoul> con - > not
[15:10:40] <lostLinSoul> And not having had a button created for it yet (by that code), that is
[15:12:24] <lostLinSoul> fuzzie: It's a check to see if I have previously created a button for that I'd or not
[15:12:34] <fuzzie> you should do that somewhere else, though
[15:13:08] <lostLinSoul> It's for new portrait buttons that are not existing in CHU
[15:14:04] <lostLinSoul> It's being done in GUICommonWindows.py where the portrait panel is being created.
[15:14:17] <lostLinSoul> If I don't do it there, where should I do it?
[15:16:20] <lostLinSoul> fuzzie: ---^
[15:16:35] <fuzzie> why can't you do it there?
[15:17:15] <lostLinSoul> You just said I should be doing it somewhere else!
[15:17:24] <fuzzie> oh, sorry
[15:17:32] <fuzzie> i mean, you shouldn't be doing the check by trying to get the control
[15:17:57] <fuzzie> you should be using HasControl first, before you try anything which might cause an error
[15:18:45] <lostLinSoul> HasControl? Will that cech for id's not in CHU?
[15:18:58] <fuzzie> sure
[15:20:10] <tomprince> ..... although, if we had more refined errors, it would be more pythonic to try, and then catch the error.
[15:20:20] <fuzzie> would it?
[15:20:31] <tomprince> But if we were going to be doing that, something like KeyError or IndexError would be appropriate.
[15:21:15] <fuzzie> I'm just confused as to why the controls need to be checked at *all* though, since presumably they're going to all get destroyed every time you reload the chu. Do we really not do that?
[15:22:19] <tomprince> http://python.net/~goodger/projects/pycon/2007/idiomatic/handout.html#eafp-vs-lbyl
[15:22:58] <lostLinSoul> In all hosnesty I don't know. My logic is sime if the button is there return it, if not create it, the check is essentially the assignment to the return value
[15:23:27] <lostLinSoul> which is a button
[15:23:47] <fuzzie> tomprince: yes, but "don't bother keeping track of state" is not "use try/except when you might have to add a lot of code for unexpected cases".
[15:24:00] <fuzzie> and it makes things a huge pain to debug.
[15:25:58] <lostLinSoul> It's no problem, thought it would be more efficient that way and didn't know of HasControl
[15:26:19] <lostLinSoul> I can jsut amend the code to use that instead
[15:26:32] <fuzzie> ideally you would just do this when the CHU is loaded
[15:26:57] <lostLinSoul> Even if the CHU is loaded they will not exist
[15:27:09] <lostLinSoul> in changing party size
[15:27:20] <tomprince> Well, but if you do when the chu is loaded, then after that you will know they are always there.
[15:27:20] <lostLinSoul> I'm changing party size
[15:27:56] <lostLinSoul> the control id's I'm looking for will not exist unless I edit the CHU file
[15:28:33] <lostLinSoul> i.e. control id's part portrait button over six are not in CHU
[15:28:46] <lostLinSoul> part -> for
[15:28:55] <tomprince> yes, we know.
[15:29:43] <tomprince> But that doesn't mean you can't add them *immedieatly* after. And then when this function is called you will know they are there, since they were already created.
[15:30:33] <lostLinSoul> Oh right now I'm with you
[15:31:06] <lostLinSoul> That makes dynamic party size impossible though
[15:31:42] <lostLinSoul> unless i add about 30 new id's straight off the bat
[15:32:56] <lostLinSoul> I want gamers to be able to choose if it is enforced or they can just let the party keep growing if they so desire....
[15:36:48] <lostLinSoul> Someone also mention something about including familiar, so I'm just thinking ahead - althouigh I have even started to lokk at how that might work
[15:37:37] <fuzzie> well, the portrait buttons will be hidden for people who don't exist anyway?
[15:38:09] <lostLinSoul> Yes
[15:39:37] <lostLinSoul> But isn't creating an unnecessary memory hog if they are not need (thoug I don't know how much it would equate to)
[15:40:47] <lostLinSoul> Sorry I just can't type
[15:41:07] <wjp> I would probably try to first get things working before worrying too much about that
[15:42:06] <lostLinSoul> Well the basics are already working apart for the message...
[15:43:26] <lostLinSoul> Although I need to work on other bits because I run at 1280, so I need to manage how it works at lower res.
[16:28:47] <-- haad has left IRC (Quit: Leaving.)
[16:29:56] --> brad_a has joined #gemrb
[16:30:53] <brad_a> alx3apps: about that patch i gave you. you just need an #ifdef ANDROID #define main SDL_main block
[16:31:05] <brad_a> in GemRB.cpp
[16:34:43] <alx3apps> brad_a: haven't tried it since yesterday, i'll try now
[16:38:51] <brad_a> i admit i just glossed over this chu id stuff, but cant we write a python generator like GetNewControlId to incrementally return ids starting with the last one from the CHU?
[16:49:12] <lynxlynxlynx> what's the benefit of having it dynamic? you have to look at the chu manually anyway
[16:49:27] <brad_a> well i have no idea how any of this works :-P
[16:50:10] <brad_a> and if this is to work with other mods doesnt it have to be dynamic?
[17:01:14] <-- lynxlynxlynx has left IRC (Ping timeout: 260 seconds)
[17:15:51] --> lynxlynxlynx has joined #gemrb
[17:15:51] <-- lynxlynxlynx has left IRC (Changing host)
[17:15:51] --> lynxlynxlynx has joined #gemrb
[17:15:51] --- ChanServ gives channel operator status to lynxlynxlynx
[17:16:59] <lynxlynxlynx> existing gui mods don't dare touch ids or they'd crash in the original
[17:18:13] <brad_a> ah
[17:21:33] <-- SiENcE has left IRC (Quit: @all: cya)
[17:24:18] <brad_a> still if we are going to keep modding the GUI...
[17:25:42] <lostLinSoul> Not quite true, the char gen does
[17:25:44] <tomprince> I want something more expressive that chu. Perhaps something that can reference a chu, and then extend apon it, also adding descriptive ids, rather than numberic ones.
[17:27:03] <fuzzie> well, it's difficult to do that and not make it hugely painful to mod
[17:28:38] <tomprince> How come?
[17:29:19] <fuzzie> because you have to avoid it being yet another level of redirection, right?
[17:29:31] <lostLinSoul> In GUIScripts/bg2/GUICG22.py ln 94
[17:29:44] <lynxlynxlynx> that's our code :P
[17:29:55] <fuzzie> I imagine e.g. you could build some kind of visualisation/editing tool for them.
[17:31:55] <alx3apps> brad_a: I got the same error even when I change main to SDL_main manually. I'm sure I missed something.
[17:31:55] <alx3apps> Also this output seems not logical to me (only SDL_main is in apple code):
[17:31:56] <alx3apps> ~/projects/android/gemrb$ git checkout v0.7.0
[17:31:56] <alx3apps> HEAD is now at 412a486... set the version number for 0.7.0
[17:31:56] <alx3apps> ~/projects/android/gemrb$ grep -nr "SDL_main[^.]" .
[17:31:56] <alx3apps> ./apple/CocoaWrapper/ios/CocoaMain.m:28:extern int SDL_main(int argc, char *argv[]);
[17:31:56] <alx3apps> ./apple/CocoaWrapper/ios/CocoaMain.m:31:int SDL_main (int argc, char **argv)
[17:31:56] <alx3apps> I'll try to figure it out on simple examples first, but not today
[17:32:22] <brad_a> no SDL_man is not apple
[17:32:34] <brad_a> if you look at SDL.h you see it does that same define
[17:32:43] <tomprince> Well, I was thinking it would be easier to do mods like this. And to support pst vs. others. Since the GUIScripts can just use the same name for everything, since the mapping is done at a lower level.
[17:32:49] <brad_a> thats how SDL gets the entry point to the old main
[17:33:14] <fuzzie> tomprince: but then how would new stuff be added?
[17:34:22] <fuzzie> Some mapping code could be done pretty quickly in python..
[17:34:22] <lostLinSoul> not forgetting that chu file are game files not gemrb files
[17:34:25] <brad_a> alx3apps: what is the exact error?
[17:35:14] <alx3apps> brad_a: yes, I used SDL_main earlier, but something different about it in gemrb
[17:35:17] <alx3apps> sdl_main/sdl_main.c:53: undefined reference to `SDL_main'
[17:35:23] <tomprince> Sure, python is more expressive than chu. :)
[17:35:52] <brad_a> and you replaced main wiht SDL_main? maybe you need to extern it like i do for apple
[17:39:08] <edheldil> tomprince: I prefer using libglade than writing gtk code in programs directly ;)
[17:39:12] <-- edheldil_ has left IRC (Ping timeout: 272 seconds)
[17:40:40] <tomprince> I did have something like that in mind. At least seperation of code from the description of the gui.
[17:42:41] <edheldil> what might have merit would be CHU v2 with strrefs for strings and object names allowing automatically connecting signals with python classes, plus some flags we usually have to supply by hand
[17:43:51] <edheldil> I think that what is REALLY cumbersome is GUIScripts API :)
[17:44:11] <tomprince> That is what I was suggesting. Though I also wanted to transparently support referencing the original chu, for handling modded and multiple resolutions.
[17:44:44] <tomprince> I'd love to have a better api. I just don't do any gui programming, so I have no idea what it should look like.
[17:44:51] <edheldil> now that starts to sound nasty :)
[17:45:39] <edheldil> well, not necessarily something pythonic ... regular would be enough :)
[17:45:59] <tomprince> I did suggests awhile ago to have a more standard gui api, but didn't get any interest then.
[17:46:05] <edheldil> yesterday I have stumbled over a gem of Set*Icon functions
[17:46:31] <lostLinSoul> Well I've thought of a rework for the original code I wrote, which I'm trying to implement now
[17:46:42] <edheldil> Each of them works different, setting different things on buttons
[17:47:23] <lostLinSoul> I'm sorry for having dug up a hornets net... :0P
[17:47:26] <edheldil> Maybe I have originally wrote some of them, too be honest :)
[17:47:37] <edheldil> naah
[17:47:48] <lostLinSoul> nest even
[17:48:13] <tomprince> edheldil: If you have suggestions for improving the api, write them down somewhere, and I'll fix things up, when I get a chance.
[17:48:33] <tomprince> At least the C++/Python interface, not neccesairly the code using that interface.
[17:50:09] <alx3apps> brad_a: it work's with extern "C" prefix (it also stated in SDL.h). Audio works on device, but audio continues when app go's to background. SDL 1.2. I think this case explained here https://github.com/pelya/commandergenius/blob/sdl_android/readme.txt#L227
[17:50:39] <edheldil> well, nothing interesting enough except ofmaking all the button images and pics into a sparse array
[17:51:35] <brad_a> alx3apps: it aslo says this "Also, the usual event SDL_ACTIVEEVENT with flag SDL_APPACTIVE will be sent when that happens"
[17:52:07] <brad_a> makes it sound like if you add al_android_pause_playback() and al_android_resume_playback() to the sdl event handler it should work
[17:52:48] <brad_a> would you try it :)
[17:54:48] <brad_a> you will need to include Audio.h i guess
[17:54:56] <brad_a> you probably knew that tho :)
[17:54:59] <tomprince> edheldil: Would a dict work there? (I'm not sure exactly what code you're referring to)
[17:56:05] <tomprince> alx3apps: C++ mangles function names (due to overloading. 'extern "C"' tells the compiler not to mangle the name of the given function, so it can interoperate with C or other code which doesn't do mangaling.
[17:58:00] <alx3apps> tomprince: yes, I know this in theory (and used it once long ago doing python bindings), but it's not easy to keep it in the mind without practice )
[17:59:17] <tomprince> alx3apps: I misread your message, I had thought it said "whats with extern "C"" :)
[17:59:53] <alx3apps> tomprince: :)
[18:00:28] <edheldil> tomprince: no worry. I was just thinking that it would be more elegant and generic to replace 4*sprite + picture or picturelist on buttons with just 4*sprite[] ...
[18:00:34] <edheldil> nothing urgent
[18:01:35] <edheldil> later
[18:10:25] --> edheldil_ has joined #gemrb
[18:10:27] --> SiENcE has joined #gemrb
[18:15:52] <alx3apps> tomprince: about android buildslave (TL;DR I'm slow, it'll take long). I think android build process should be moved to gemrb repository to make it work right with gemrb. It's too generic now in pelya's repo and hard (for me) to change without breaking something. So I see my next steps to setup android buildslave: 1) take some helloworld SDL project from pelya's repo and make one-command build for it with max simple scripts (to understand bu
[18:15:56] <-- edheldil_ has left IRC (Ping timeout: 272 seconds)
[18:16:40] <alx3apps> brad_a: sorry, cannot do it right now. promise to try it, but not today
[18:16:52] <brad_a> its fine :)
[18:20:56] <brad_a> tomprince: how does this look? http://dl.dropbox.com/u/13866402/cocoacore.patch
[18:23:26] <brad_a> min and max are defined like this: #define MAX(A,B) ({ __typeof__(A) __a = (A); __typeof__(B) __b = (B); __a < __b ? __b : __a; })
[18:23:47] <brad_a> i think thats fine, but maybe its not?
[18:24:03] <fuzzie> um, __typeof__?
[18:24:25] <brad_a> fuzzie: its not mine its defined in NSObjCRuntime.h
[18:24:37] <fuzzie> oh, I see
[18:24:54] <fuzzie> well that's fine then I guess
[18:25:16] <brad_a> i think so and i havent seen anything break in my limited testing on ipad simulator
[18:28:11] <fuzzie> the namespace thing on the other hand is horrible :-p
[18:28:42] <fuzzie> but if you do it then please don't intend the contents, in any case
[18:28:56] <fuzzie> also keep the comment above the class..
[18:29:12] --> Avenger has joined #gemrb
[18:29:23] <brad_a> i did indent...
[18:29:29] <Avenger> Hey! what is this PushOffset thing? why it was needed?
[18:29:30] <fuzzie> yes, *don't* indent
[18:29:35] <brad_a> oh
[18:29:47] <brad_a> my eyes are still tired :-P
[18:29:54] <Avenger> If it was only for this: Button.SetPushOffset (0, 0), then it is not needed at all
[18:30:10] <fuzzie> Avenger: no?
[18:30:16] <fuzzie> how would you do it?
[18:30:20] <Avenger> buttons can be locked (not depressed) and active
[18:30:28] <brad_a> i dont know that the namespace is needed but its the only way i knew how to mesh with it
[18:30:47] <fuzzie> i mean that's all edheldil's code so you have to ask edheldil :p
[18:31:03] <Avenger> yeah, i know, i hope he is awake
[18:31:21] <fuzzie> but locking disables PRESSED?
[18:31:45] <fuzzie> no, I guess that's useless
[18:31:51] <fuzzie> what's your suggestion?
[18:32:23] <Avenger> no, locked doesn't disable pressing/messages
[18:32:26] <Avenger> just the image change
[18:32:34] <fuzzie> LOCKED seems to disable the image change, yes
[18:32:57] <Avenger> I guess SetPushOffset(0,0) is just for that, no?
[18:32:59] <fuzzie> no
[18:33:23] <fuzzie> i mean, doesn't this code *want* the pressed image, just not the offset change?
[18:35:27] <fuzzie> or are all the images/etc identical?
[18:35:44] <Avenger> i don't know, but i'm pretty sure this can be coded without the new feature (If it is ever used with 0,0)
[18:35:51] <fuzzie> well
[18:36:01] <fuzzie> "disable the offset when you push a button" seems like a useful feature to have
[18:36:12] <Avenger> there are state, like 'SECOND' and 'THIRD' too
[18:36:13] <fuzzie> you could add yet another flag but .. we have how many billion flags already? :)
[18:36:27] <Avenger> i don't think you need to add any new
[18:36:35] <Avenger> the current featureset should be enough :D
[18:36:36] <fuzzie> ah, SECOND is good?
[18:37:03] <Avenger> it is a 'pressed' state
[18:37:04] <fuzzie> maybe if you know how it works, then add comments
[18:37:56] <fuzzie> but, I guess this code is unfinished
[18:37:59] <Avenger> well, first i check what this new stuff wants to accomplish, then i'll try to work it out without the new call
[18:38:08] <fuzzie> so, maybe ask before removing anything
[18:38:16] <fuzzie> although harmless to change the guiscript
[18:39:17] <Avenger> it's not like i didn't consider this, i just thought i can live without it as the whole stuff was like either 0,0 or 2,2
[18:40:11] <fuzzie> well .. it's difficult for anyone else to understand when this SECOND thing has no comment about what it does :)
[18:41:57] <Avenger> it is to highlight items in guiinv. Same function as ed used in the floatmenu
[18:44:29] <fuzzie> sure, I can see now what it's doing
[18:44:37] <fuzzie> but only because you suggested it and I poked through the code..
[18:48:17] <Avenger> yeah, i know it is not documented too much except in SetButtonState
[18:48:28] <Avenger> but there is definitely no 'recommended practices' and such :)
[18:49:13] <Avenger> the whole code works perfectly with LOCKED used in place of ENABLED
[18:51:29] <CIA-28> GemRB: 03avenger_teambg * rbae39793630a 10gemrb/gemrb/GUIScripts/pst/FloatMenuWindow.py: no need of SetPushOffset (if it is planned to only ever 0,0 or 2,2 then it is not necessary)
[18:51:39] <fuzzie> no UI difference at all?
[18:51:44] <Avenger> nothing
[18:51:49] <Avenger> you can try it
[18:51:50] <fuzzie> and it matches original?
[18:51:57] <fuzzie> well, I can't because I have no pst here
[18:52:04] <Avenger> i don't know, it just matches the one i replaced
[18:52:20] <fuzzie> well I just think, maybe you sabotaged poor ed's code before it is done, if it doesn't match original :-)
[18:52:24] <Avenger> the difference is that while you press the item slot while filled, the item is not shifted
[18:52:41] <fuzzie> and there's no change in rendering at all?
[18:52:55] <Avenger> it gets a pink highlight
[18:53:34] <fuzzie> aha
[18:53:43] <Avenger> or rather, a border...
[18:54:20] <Avenger> after i release mouse
[18:54:34] <Avenger> i'm pretty sure it didn't get the highlight before that
[18:54:43] <Avenger> with his code either, did it?
[18:54:45] <tomprince> Well, I'd rather more but othogonal options, than less but special purpose options.
[18:55:31] <Avenger> i would rather not have a new function if it works with old options that are already used 10+ places through the code
[18:56:19] <tomprince> Well, me and edheldil where discussing simplifying the python api earlier.
[18:56:35] <Avenger> a new function means a new thing to learn, document, maintain. And doing the same thing differently in different places confuses people. Also confuses lynx when he tries to merge code that does the same thing.
[18:56:42] <tomprince> brad_a: I'd rather not add namespaces just for a single header.
[18:57:09] <brad_a> k. what are my options?
[18:57:13] <tomprince> I'd not be opposed to adding namespaces everywhere.
[18:58:03] <tomprince> I don't know about others, though.
[18:58:20] <fuzzie> well I don't see the harm in adding it in the one place brad needs it
[18:58:31] <gembot> build #557 of cmake clang++ is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20clang%2B%2B/builds/557
[18:58:50] <Avenger> why does he need namespace?
[18:58:50] <brad_a> yes i dont really feel like editing 500 files just to use 1...
[18:59:06] <brad_a> becasue point is defined in some apple library :-/
[18:59:14] <tomprince> Point is too generic a name.
[18:59:16] <Avenger> hmm
[18:59:21] <Avenger> then rename it
[18:59:37] <fuzzie> three lines to add namespace seems less annoying than renaming it :-p
[18:59:46] <brad_a> agin tho that means editing 100 files :-P but if thats what we need to do...
[18:59:49] <Avenger> i'm not against renaming, but adding namespace, just to avoid this collision
[19:00:23] <Avenger> well, there was some typo in some class or something earlier :) it wasn't as widespread as point, but eventually it was fixed
[19:00:44] <tomprince> Well, what's the issue with namespaces?
[19:01:03] <Avenger> namespace is just one more obfuscating tools
[19:01:19] <Avenger> 'high level abstraction' meeh
[19:01:35] <Avenger> nice surface, weird stuff under hood
[19:01:36] <fuzzie> well brad was just using it as a hack
[19:01:44] <brad_a> hack?
[19:01:47] <fuzzie> wrap Point in namespace GemRB, then use GemRB::Point globally
[19:01:47] <tomprince> Well, all you need to do is open and close the namespace in the header files, and use it in the cpp files.
[19:02:02] <fuzzie> but namespace in general does seem like simply useless boilerplate for gemrb
[19:02:05] <gembot> build #169 of cmake g++-4.2 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.2/builds/169
[19:02:08] <Avenger> all you need is a different name for point :D
[19:02:08] <fuzzie> brad_a: death to using statements :)
[19:02:17] <brad_a> i dont know any better :)
[19:02:23] <fuzzie> Avenger: that is a lot more ugly than just a quick fix though :P
[19:02:27] <brad_a> jsut enought to get it to work i guess
[19:02:40] <tomprince> Death to using in header files, certainly.
[19:02:55] <tomprince> Not neccesairly in implementation files.
[19:02:55] <Avenger> heh, fuzzie, why? It is ugly only for the repository and the mailing list. The code remains nice.
[19:03:02] <fuzzie> Avenger: really?
[19:03:06] <fuzzie> i mean, i guess if you have a suggestion..
[19:03:11] <fuzzie> i just can't think of another name which wouldn't be ugly
[19:03:50] <Avenger> well, Point is a nice short word
[19:04:09] <fuzzie> yes, so it would be nice to keep it :p
[19:04:15] <Avenger> Position is too long?
[19:04:44] <brad_a> renaming will confuse people.
[19:04:59] <fuzzie> I don't really care since I probably won't be committing again until April.. ;p
[19:04:59] <brad_a> weve been using point and now sudenly cant. its annoying i feel
[19:05:03] <brad_a> he he
[19:05:31] <Avenger> yes, it is annoying
[19:05:37] <gembot> build #172 of cmake g++-4.4 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/172
[19:06:16] <Avenger> Still, using GemRB::Point is actually uglier than using IEPoint
[19:06:29] <fuzzie> yes
[19:06:31] <fuzzie> but we wouldn't do either
[19:06:51] <fuzzie> you just add 'namespace GemRB {' before Point, '}' after Point', then 'using GemRB::Point;' in some header
[19:06:55] <fuzzie> and that's three lines change
[19:07:11] <Avenger> i still have bad feelings about this
[19:07:15] <fuzzie> it is totally a hack though.
[19:08:02] <Avenger> i do hacks to overcome extra code in the binary, not because i fear of a bloated email :D
[19:08:18] <tomprince> I still suggest just moving everything to a namespace. Feels less hackish, to me.
[19:08:38] <fuzzie> yes, but then you have namespace boilerplate everywhere
[19:08:49] <fuzzie> just consider my vote to be that I hate all options
[19:08:52] <fuzzie> I will go slink back into my corner.
[19:09:09] <Avenger> yeah, this is ugly either way
[19:10:13] <Avenger> and all because of some silly apple api. Not like i don't hate windows apis too. Just added 'SetIcon' in IDA and it turns out it is some api function with a totally different prototype. It messed up the db by forcing the wrong type all over the db
[19:11:03] <tomprince> Or silly us, for using the global namespace.
[19:11:19] <brad_a> ^ this i think
[19:11:21] <Avenger> how much we need to change to move everything to gemrb namespace?
[19:11:31] <fuzzie> just add the namespace lines at top/bottom of every file
[19:11:45] <Avenger> can't you add it to global.h?
[19:11:52] <Avenger> or some other tricky position
[19:12:03] <fuzzie> no, has to be everywhere, since you have to have start/end
[19:12:13] <Avenger> well, that sucks
[19:12:29] <brad_a> yes thats why i went with the hack option :D
[19:12:31] <Avenger> if i have to modify a crapload of stuff, i would rather go with IEPoint :D
[19:13:18] <Avenger> there is no way to do something like #if APPLE #undef Point ????
[19:14:17] <fuzzie> well i don't care about how much has to be changed
[19:14:34] <fuzzie> just that everyone will glare in your direction every time they have to type 'IEPoint', for the rest of time, maybe
[19:14:43] <tomprince> Well, you don't need to change it.
[19:15:21] <brad_a> avenger: i did that originally a few weeks ago and tomprince said no ;-)
[19:15:47] <Avenger> meh, why? if it is an apple problem, deal with it only in apple
[19:16:03] <fuzzie> doesn't happen on QNX?
[19:16:08] <tomprince> It is a problem that has only occured with apple so far.
[19:16:30] <brad_a> yes its only apple thus far but many other potential platforms or 3rd party libs
[19:17:28] <Avenger> i dislike that a platform feature forces us to have extra code for other platforms. Though, i'm not sure how much an extra namespace will cause. Probably only more debug info?
[19:17:55] <fuzzie> well, larger names in debug info
[19:17:59] <tomprince> It won't affect the code at all. Just symbol names.
[19:18:00] <fuzzie> it is 'transparent', you just get 'GemRB::'
[19:18:30] <Avenger> yeah, all over the place ;)
[19:18:41] <Avenger> just symbol names, i know
[19:19:17] <Avenger> in low level, all these things look ugly
[19:20:12] <Avenger> well, i guess it is the simplest thing to do on source level
[19:20:59] <Avenger> but i fear i will wake up one day and have 10+ namespaces jumping up and down all over the place
[19:21:06] <fuzzie> then you apply big stick
[19:21:53] <tomprince> well, what's wrong with namespaces, if it makes where code is comming from clearer.
[19:22:04] <fuzzie> well, 'clearer' is kind of opinion :)
[19:22:10] <fuzzie> and, that is where it goes wrong
[19:22:13] <fuzzie> as usual
[19:22:38] <Avenger> it is not clearer, there is a header file far far away from the actual code i peer at. Hidden in that header is a namespace declaration
[19:22:38] <tomprince> Sure.
[19:22:41] <-- kettuz has left IRC (Quit: Leaving)
[19:23:05] <Avenger> suddenly one day, the Point object is something else
[19:23:24] <Avenger> and i will not know what the fsck is it, because i won't notice the namespace thingie
[19:23:44] <fuzzie> yes well then you shout at people :-p
[19:24:04] <Avenger> just like the bitmap plugin trying to unzip an area :D
[19:24:10] <fuzzie> heh
[19:24:47] <Avenger> except that the plugin system is cool, and namespaces are not :D
[19:26:43] <Avenger> i'm really happy with c++, i wouldn't want to go back to C and use only structs and like that. But using as few features as possible is still a goal for me
[19:30:29] <brad_a> its not like i WANT to use namespaces :-P
[19:31:03] <brad_a> but i do want to use the new logger and eventually the new config interface from objective-c code
[19:31:48] <Avenger> yeah, if the #if APPLE thing doesn't work, do the namespace hack
[19:32:56] <brad_a> i can do waht i did before and do #if apple #define Point GemPoint
[19:33:04] <brad_a> but tomprince said no :-P
[19:34:10] <brad_a> i agree with him tho
[19:34:19] <brad_a> this really really isnt an apple thing
[19:34:51] <brad_a> jsut i am the only one currently bolting on a gui config interface
[19:35:44] <Avenger> gui config interface?
[19:36:04] <Avenger> you mean gui not using the current chu stuff actually in gemrb?
[19:37:19] <brad_a> yes
[19:37:44] <brad_a> an eventual alternative to gemrb.cfg
[19:37:48] <Avenger> i can only say, huh :D
[19:38:15] <Avenger> but why do you need something other than guiopt?
[19:38:23] <Avenger> and use the existing guiscripts
[19:38:32] <brad_a> i have no idea what those are
[19:38:49] <brad_a> but i need an interface to run prior to core being initialized
[19:39:43] <Avenger> i'm not a fond of using some platform specific windowing system directly in gemrb
[19:39:54] <Avenger> to be precise, i don't want that.
[19:40:17] <Avenger> it is fine to have a launcher, that is platform specific
[19:41:08] <gembot> build #505 of msvc++6 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/msvc%2B%2B6/builds/505
[19:41:16] <Avenger> either that, or just use chu like the original games
[19:41:30] <Avenger> you CAN design new chu, of course, it needs an initialized system
[19:41:43] <Avenger> but the original games can live without external aid too
[19:41:46] <brad_a> for all intents and purposes it is a launcher
[19:42:19] <Avenger> but if it is a separate executable, then gemrb doesn't need any namespace and you don't need any of gemrb's Point
[19:42:29] <brad_a> and has been around for quite some time ;-) im just trying to take it to the next level now
[19:42:52] <Avenger> you only have to write the config, and start gemrb
[19:42:57] <brad_a> well its not a separate exe but it does its thing before calling GemRB-Main
[19:43:08] <fuzzie> two processes is painful on some platforms
[19:43:12] <brad_a> yes
[19:43:25] <brad_a> and grossly unnessicary and even more hacklike
[19:43:30] <fuzzie> android port certainly doesn't do two processes, it just has all the hacky extra code in another repository
[19:43:38] <fuzzie> of course that is in java, so :-p
[19:43:45] <Avenger> but now we come back to one platform enforcing something on all the overall architecture
[19:44:15] <brad_a> i wouldnt say arcitecture
[19:44:27] <Avenger> fuzzie, how can it be single process with java?
[19:44:44] <fuzzie> Avenger: gemrb is built as a library, the java loads the library
[19:45:15] <Avenger> ok, but brad can do that too, without the point hack, no?
[19:45:17] <fuzzie> (all applications using the display/input on android must be java)
[19:45:49] <fuzzie> well
[19:46:00] <fuzzie> i think if we talk about one platform enforcing things, msvc6 is the highest candidate, really :)
[19:46:10] <brad_a> i don't see anybody making a cHU configuration interface anytime soon...
[19:46:16] <fuzzie> i don't see why we spend so much time recently arguing about how to make things really really more difficult
[19:46:22] <Avenger> ok fuzzie, but it is for me to actually debug this thing
[19:46:43] <Avenger> i would happily dump it if there is a competing debugger
[19:47:11] <fuzzie> yes, it's just .. this is the real world, things are annoying
[19:47:25] <fuzzie> i don't see why we have to have everything perfect before anything is committed :-p
[19:48:36] <Avenger> i think we argue now, because more people do things, there is nothing surprising in that
[19:48:43] <brad_a> i would be more than happy to live with an #if apple hack that wont affect anybody :)
[19:49:35] <brad_a> i know its not ideal but its 1 simple hack that enables me to take advantage of the OS specific features im trying to use
[19:49:45] <Avenger> the namespace is ok, i'm just not sure how much of this apple juice will trickle down here :D
[19:50:46] <Avenger> i mean, platform specific windowing is obviously not working, and the more close it is tied to gemrb, the more of the collateral damage it causes
[19:50:49] <brad_a> that should be it. i cant see any other problems. all i need to be able to do is use logger/config/core from objective-c code and this is the only thing stoppping me
[19:51:45] <Avenger> on the other hand, if you manage to put the quicktime movie window into a gemrb object ....
[19:51:47] <brad_a> i think you are worying too much :)
[19:52:15] <Avenger> either that, or we need a quicktime plugin
[19:52:28] <brad_a> i am thinking of doing a vlc plugin
[19:52:43] <brad_a> that would play quicktime and just about anything else
[19:54:14] <Avenger> it is not extremely important to have it directly in the object (ie moving with the window), just have a surface correctly sized and placed over gemrb when playing the movie.
[19:55:45] --> edheldil_ has joined #gemrb
[19:57:37] --> haad has joined #gemrb
[19:59:58] <-- brad_a has left IRC (Remote host closed the connection)
[20:00:06] --> brad_a has joined #gemrb
[20:17:23] --> brad_a_ has joined #gemrb
[20:19:12] <lynxlynxlynx> that was long
[20:20:01] <-- lostLinSoul has left IRC (Ping timeout: 245 seconds)
[20:20:08] <-- brad_a has left IRC (Ping timeout: 240 seconds)
[20:20:08] --- brad_a_ is now known as brad_a
[20:39:51] <-- alx3apps has left #gemrb
[20:45:20] <gembot> build #156[namespaces] of osx-xcode-binary is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/156 blamelist: edheldil, alx3apps <alx3apps@gmail.com>, brad_a, tomprince, Avenger
[20:46:55] <gembot> build #177[namespaces] of osx-test is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/osx-test/builds/177 blamelist: edheldil, alx3apps <alx3apps@gmail.com>, brad_a, tomprince, Avenger
[20:53:13] <tomprince> https://github.com/tomprince/gemrb/commit/namespaces
[20:53:58] <tomprince> I had to wrap the .cpp files in namespace too.
[20:54:12] <brad_a> really?
[20:54:48] <brad_a> i dont think avenger would be pleased :-P
[20:54:55] <tomprince> I'd have preferred just using using there, but that would require marking the globally visible definitions of functions/structures, which I guessed wouldn't fly.
[20:55:01] <fuzzie> 20:49 <Avenger> the namespace is ok, i'm just not sure how much of this apple juice will trickle down here :D
[20:55:18] <fuzzie> so, can always get a review
[20:55:28] <fuzzie> if it's all nice and hands-off
[20:56:14] <edheldil_> Avenger: it's NOT the same
[20:56:23] <edheldil_> grrr :)
[20:56:28] <brad_a> i dont mean to cause so much trouble :-P
[20:56:54] <tomprince> brad_a: You're not. Just a good excuse. ;)
[20:57:04] <brad_a> :D
[20:57:30] <fuzzie> edheldil_: well, as it is, your code seems like it does what SECOND does
[20:57:49] <edheldil_> it just look like that
[20:57:52] <edheldil_> looks
[20:58:11] <fuzzie> (specifically, not doing the offset)
[20:58:42] <Avenger> how is it different?
[20:59:12] <edheldil_> SECOND, LOCKED efectively disable the button, meaning that it does not change sprite on press and does not make click sound
[20:59:20] <fuzzie> no, SECOND changes the sprite on press
[20:59:22] <edheldil_> they are hacks
[20:59:45] <edheldil_> what Avenger used does not
[21:00:04] <Avenger> but the sprite doesn't need to be changed
[21:00:10] <edheldil_> it does
[21:00:14] <Avenger> what is the change?
[21:00:29] <Avenger> if you mean the pink outline, then it is there after you release the button
[21:00:58] <edheldil_> the spells and abilities highlight
[21:01:04] <edheldil_> not the outline
[21:01:20] <fuzzie> in the youtube videos, the spells/abilities seemed to get a little flashy highlight when held down
[21:02:11] <edheldil_> anyway, setting some strange states does confuse what the button does
[21:02:17] <Avenger> flashy highlight? you mean an animation?
[21:02:30] <fuzzie> hm ok, and indeed SECOND is a hack, since OnMouseDown does SetState(IE_GUI_BUTTON_PRESSED)
[21:02:45] <edheldil_> it's NORMAL button, it just does not shift the content.
[21:03:25] <edheldil_> and I thought setting it to st. else than 0,0 or 2,2 COULD be useful for mods with big or shallow buttons
[21:03:41] <fuzzie> it is much nicer to have it moddable
[21:03:51] <Avenger> if it is used, yes
[21:04:02] <gembot> build #232[namespaces] of nmake-msvc++10 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B10/builds/232 blamelist: edheldil, alx3apps <alx3apps@gmail.com>, brad_a, tomprince, Avenger
[21:04:08] <fuzzie> there's way too much of the UI hardcoded right now, so that it can't be modded
[21:04:11] <Avenger> i just don't see the difference
[21:04:25] <Avenger> i tested this with the items, and i didn't see ANY difference
[21:04:30] <fuzzie> the items don't do the anim
[21:04:45] <fuzzie> well I am going by youtube videos, but I assume that is normal :)
[21:05:01] <Avenger> fuzzie: SetPushOffset(0,0) won't do any animation
[21:05:20] <edheldil_> and moreover items and anims set different images on the buttons - picture vs. sprites
[21:05:22] <Avenger> it just disables the shifting
[21:05:23] <fuzzie> isn't it just a different Pressed sprite?
[21:05:29] <Avenger> no
[21:05:46] <fuzzie> hm, I don't see any gemrb code for doing an actual anim
[21:06:06] <edheldil_> Avenger: it's anim. Spelss are anim. Items are not
[21:06:38] <edheldil_> moreover, audio feedback
[21:06:40] <lynxlynxlynx> we have an animation bit already, it's used for the pause button
[21:07:03] <fuzzie> we seem to have gone a bit overboard with the button flags
[21:07:17] <fuzzie> my poor head
[21:08:04] <Avenger> which bam we talk about?
[21:08:40] <-- DrMcCoy has left IRC (Quit: Server migration)
[21:09:15] <Avenger> because 'amgens' is a single frame
[21:09:34] <gembot> build #517[namespaces] of mingw32 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/517 blamelist: edheldil, alx3apps <alx3apps@gmail.com>, brad_a, tomprince, Avenger
[21:09:48] <Avenger> maybe ampsll
[21:10:13] <edheldil_> SP001.BAM, for example
[21:10:14] --> DrMcCoy has joined #gemrb
[21:12:35] <gembot> build #506[namespaces] of msvc++6 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/msvc%2B%2B6/builds/506 blamelist: edheldil, alx3apps <alx3apps@gmail.com>, brad_a, tomprince, Avenger
[21:13:43] <Avenger> ok, revert the change
[21:14:00] <gembot> build #174[namespaces] of cmake g++-4.4 is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/174 blamelist: tomprince
[21:14:08] <gembot> build #174[namespaces] of cmake g++-4.6 is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.6/builds/174 blamelist: tomprince
[21:14:10] <Avenger> i never seen this highlight thing
[21:14:15] <gembot> build #165[namespaces] of cmake g++-4.5 is complete: Failure [4failed] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.5/builds/165 blamelist: tomprince
[21:14:16] <Avenger> never noticed :)
[21:14:53] <tomprince> It's fairly recent. It uses unmerge code to assoicate git names with irc names.
[21:14:57] <edheldil_> the menu has lots of other problems :)
[21:16:55] <Avenger> do you have iwd2?
[21:17:07] <Avenger> i meant Ed :D
[21:17:20] <gembot> build #279[namespaces] of nmake-msvc++6 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B6/builds/279 blamelist: edheldil, alx3apps <alx3apps@gmail.com>, brad_a, tomprince, Avenger
[21:17:35] <CIA-28> GemRB: 03edheldil * r5acddcc20aa5 10gemrb/gemrb/GUIScripts/pst/FloatMenuWindow.py:
[21:17:35] <CIA-28> GemRB: Revert "no need of SetPushOffset"
[21:17:35] <CIA-28> GemRB: This reverts commit bae39793630a85d2ea35d2eebec29b5d8d15ff70.
[21:17:58] <edheldil_> Avenger: yes
[21:18:02] <Avenger> iwd2 quickslots is much more complicated, and since you don't shy away from adding some new guiscript function, i'm pretty sure it will need one or two
[21:18:24] <edheldil_> hehe
[21:19:22] <edheldil_> Avenger: Button:Anchor is relative to control's 0:0 ?
[21:20:59] <Avenger> i don't know, but i would guess so, but i don't even see how is it used
[21:21:42] <Avenger> it might be not used, just loaded from chu
[21:21:43] <edheldil_> I have not find a single use of it, but I am not sure I have checked all data
[21:22:02] <Avenger> it isn't used
[21:22:13] <Avenger> well, at least, not implemented
[21:22:22] <Avenger> i didn't find any use for it either
[21:22:46] <Avenger> it is only loaded because i saw it documented somewhere
[21:23:01] <edheldil_> I know. It would be good for setting label position for item buttons
[21:23:18] <Avenger> maybe test it first
[21:23:32] <Avenger> i mean, test on original engine, with some changed anchor
[21:23:50] <Avenger> heh, i don't know if dltcep can edit this field at all
[21:24:00] <gembot> build #175[namespaces] of cmake g++-4.4 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/175
[21:24:10] <edheldil_> ok
[21:24:24] <Avenger> it cannot
[21:24:44] <Avenger> most likely, it is not separated from the other field it was hacked into
[21:26:52] <Avenger> lower part of x is unpressed, lower part of y is selected. I guess that is enough for most purposes
[21:27:18] <gembot> build #175[namespaces] of cmake g++-4.6 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.6/builds/175
[21:27:45] <Avenger> ok, i'm curious enough to test how anchor works
[21:27:47] <-- Avenger has left IRC (Quit: bye!)
[21:30:16] <edheldil_> heh
[21:30:55] <gembot> build #166[namespaces] of cmake g++-4.5 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.5/builds/166
[21:31:04] <gembot> build #181 of osx-test is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/osx-test/builds/181
[21:32:50] <pmjdebruijn> Win 10
[21:32:51] <pmjdebruijn> sorry
[21:35:53] <gembot> build #160 of osx-xcode-binary is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/160
[21:37:51] --> Avenger has joined #gemrb
[21:37:51] --- ChanServ gives channel operator status to Avenger
[21:38:34] <Avenger> ok, the anchor point could be any of the 5 (rectangle points, center) depending on the other text justification bits
[21:38:48] <Avenger> it works at least in tob
[21:39:58] <Avenger> hmm make this more than 5
[21:41:07] <Avenger> maybe this could be used for properly aligning the portrait icons in iwd2
[21:44:34] <Avenger> meh, there is no DDX_text for unsigned short, silly...
[21:45:53] <gembot> build #233[namespaces] of nmake-msvc++10 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B10/builds/233
[21:48:32] <gembot> build #161[namespaces] of osx-xcode-binary is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/161 blamelist: tomprince
[21:48:44] <gembot> build #563[namespaces] of cmake clang++ is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/cmake%20clang%2B%2B/builds/563 blamelist: tomprince
[21:49:10] <gembot> build #182[namespaces] of osx-test is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/osx-test/builds/182 blamelist: tomprince
[21:49:17] <gembot> build #175[namespaces] of cmake g++-4.2 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.2/builds/175 blamelist: tomprince
[21:49:49] <gembot> build #177[namespaces] of cmake g++-4.4 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/177 blamelist: tomprince
[21:50:14] <gembot> build #177[namespaces] of cmake g++-4.6 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.6/builds/177 blamelist: tomprince
[21:50:39] <gembot> build #168[namespaces] of cmake g++-4.5 is complete: Failure [4failed compile] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.5/builds/168 blamelist: tomprince
[21:53:09] <edheldil_> Avenger: so - how does it work: is it relative to control's corner chosen by other justification flags?
[21:53:19] <Avenger> yes
[21:53:40] <Avenger> relative to center seems fishy, though
[21:53:53] <Avenger> relative to top/left works perfectly as i expected
[21:54:23] <edheldil_> so BOTTOM | RIGHT | ANCHOR and -2, -2 would set button label more towards center?
[21:54:40] <Avenger> didn't check bottom right
[21:55:22] <edheldil_> ok
[21:56:46] <Avenger> a 0,0 anchor to center with anchor enabled moved it, i didn't expect that. It was like the bottom/right point of the text was aligned to the center.
[22:00:13] <Avenger> 0,0 bottom right anchor... made the whole text disappear
[22:00:19] <gembot> build #518[namespaces] of mingw32 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/518
[22:00:43] <Avenger> i have to find where it went
[22:02:59] <gembot> build #564[namespaces] of cmake clang++ is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20clang%2B%2B/builds/564
[22:03:22] <edheldil_> there goes my hope for usable interface :)
[22:05:20] <gembot> build #507[namespaces] of msvc++6 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/msvc%2B%2B6/builds/507
[22:05:41] <Avenger> i'm not sure the anchor could be negative
[22:05:54] <Avenger> any attempts of setting something like -1,-1 failed
[22:06:16] <Avenger> the text completely vanishes
[22:06:22] <edheldil_> than maybe it always points towards center
[22:06:38] <gembot> build #176[namespaces] of cmake g++-4.2 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.2/builds/176
[22:06:45] <Avenger> that wouldn't make it disappear, unless it is 65535 :D
[22:07:13] <Avenger> i used center alignment, a -1-1 wouldn't remove it totally
[22:07:47] <Avenger> so far i could master only the top left version, that is probably useful in some cases
[22:08:14] <gembot> build #183[namespaces] of osx-test is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/osx-test/builds/183
[22:08:30] <edheldil_> what about bottom right +2, +2?
[22:10:30] <gembot> build #178[namespaces] of cmake g++-4.4 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.4/builds/178
[22:11:59] <Avenger> nothing
[22:12:17] <-- SiENcE has left IRC (Quit: cya)
[22:14:00] <gembot> build #162[namespaces] of osx-xcode-binary is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/162
[22:14:07] <gembot> build #178[namespaces] of cmake g++-4.6 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.6/builds/178
[22:15:37] --> SiENcE has joined #gemrb
[22:17:37] <gembot> build #169[namespaces] of cmake g++-4.5 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/cmake%20g%2B%2B-4.5/builds/169
[22:18:37] <gembot> build #280[namespaces] of nmake-msvc++6 is complete: Success [3build successful] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B6/builds/280
[22:20:04] <edheldil_> too bad :)
[22:27:10] <Avenger> hah, found some 'useful numbers'
[22:29:06] <Avenger> so, it seems, with bottom right, it subtracts the anchor values from the bottom right coordinate. So a 0,0 would totally be invisible. To have it rendered like normal bottom/right, you need to use the actual width/height.
[22:29:16] <Avenger> of the text...
[22:29:48] <Avenger> this is entirely counter intuitive, for me at least
[22:31:46] <edheldil_> no, it makes sense
[22:32:02] <Avenger> this anchor stuff is useless, because it requires you to know the text size, which you won't
[22:33:23] <Avenger> it would be better if bottom right would align the bottom right side of the text to the bottom right side of the control.
[22:33:43] <Avenger> so, a 0,0 would be the same with/without anchor
[22:33:43] <edheldil_> it does not make text right aligned?
[22:34:00] <Avenger> why not?
[22:34:42] <Avenger> 0,0 - no effect. 10,0 moved 10 pixels to the left, 10,10 would move it 10 pixels to upper left
[22:34:57] <Avenger> that would be actually useful
[22:35:26] <Avenger> but this one aligns the top left point of the text, so 0,0 makes it vanish
[22:36:25] <edheldil_> I mean: bottom | right , anchor +2,+2 would make bottom-right text label, right aligned, moved 2 pixels towards center - exactly what's needed for items' amounts in inventory
[22:36:58] <Avenger> i think i said the same
[22:37:05] <Avenger> but it isn't working like that
[22:37:27] <Avenger> actually, i'm tempted to say, we should implement it correctly
[22:37:42] <Avenger> in the current form it is useless and unused
[22:37:45] <edheldil_> I was tempted to do it yesterday :)
[22:39:00] <Avenger> basically, i say, the offset should always move it towards center. and you always align the closest points of text and control. So the modder doesn't have to know the actual text dimensions
[22:40:08] <edheldil_> given that number bams have the anchor somewhere in the middle of a digit, it has
[22:40:17] <edheldil_> he has
[22:41:32] <Avenger> i don't understand that
[22:42:18] <Avenger> all my number bam frames got 0,0 positions, if you meant that
[22:45:47] <Avenger> and in the example you wrote, you use constant +2, +2. regardless of the number. You can do that only if you implement it like i said. It wouldn't work in the original engine
[22:46:28] <-- Avenger has left IRC (Quit: bye!)
[22:59:31] <edheldil_> Avenger: NUMBER.BAM in ps:t has most of them ~ 2:3 (bitmap center), NUMBER2.BAM has it 1:1
[23:01:57] <-- nutron|w has left IRC (Quit: I must go eat my cheese!)
[23:02:05] <-- nutron has left IRC (Quit: I must go eat my cheese!)
[23:05:48] <lynxlynxlynx> but now we have ttf fonts too
[23:06:05] <lynxlynxlynx> just something to keep in mind
[23:12:43] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:32:28] <-- SiENcE has left IRC (Quit: cya)