#gemrb@irc.freenode.net logs for 30 Dec 2011 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:27:36] <pintin> fixing the select sounds a little bit
[00:27:58] <pintin> lynx: Oh no I missed the deadline....putting up the patches now
[00:28:18] <pintin> Yoshimo: btw I'm lloyd
[00:33:52] <pintin> patch 1: http://pastebin.com/zkLGzebS -- add rare select sounds sounds into bg1 vcremap
[00:36:44] <pintin> patch2: http://pastebin.com/dRMwcgkX -- use Actor::SelectActor to play select sounds
[00:38:32] <pintin> patch 3: http://pastebin.com/rXRhzWda -- change rare select change from 1% to 5%
[00:40:48] <pintin> patch 4: http://pastebin.com/GyUQeJ1F -- fix the verbalconstant and selectactor methods
[00:40:56] <pintin> Patches must be done in order!
[00:51:54] --> brad_a has joined #gemrb
[00:56:22] <brad_a> tomprince: any chance of setting up buildbot to upload the mac distribution for 0.7.0?
[00:58:03] <gembot> build #28 of osx-test is complete: Success [build successful] Build details are at http://buildbot.gemrb.org/builders/osx-test/builds/28
[02:25:40] <-- brad_a has left IRC (Quit: brad_a)
[03:29:47] <-- PixelScum has left IRC (Read error: Connection reset by peer)
[03:30:45] --> Drakkar has joined #gemrb
[04:40:16] --> Beholder has joined #gemrb
[04:57:35] <-- Beholder has left #gemrb
[05:18:47] --> spike411 has joined #gemrb
[06:04:06] <-- pintin has left IRC (Ping timeout: 244 seconds)
[07:47:38] --> lynxlynxlynx has joined #gemrb
[07:47:39] <-- lynxlynxlynx has left IRC (Changing host)
[07:47:39] --> lynxlynxlynx has joined #gemrb
[07:47:39] --- ChanServ gives channel operator status to lynxlynxlynx
[09:16:14] <-- spike411 has left #gemrb
[09:23:51] --> duckpunch has joined #gemrb
[09:31:02] <CIA-41> GemRB: 03npintin * r131a799fe8e1 10gemrb/gemrb/core/GUI/GameControl.cpp: Now uses Actor::SelectActor to play select sounds
[09:31:04] <CIA-41> GemRB: 03npintin * rab6afff1d6e0 10gemrb/gemrb/core/Scriptable/Actor.cpp: changed the change of triggering a rare select sound from 1% to 5%
[09:31:05] <CIA-41> GemRB: 03npintin * r178a96a6f5bc 10gemrb/gemrb/override/bg1/vcremap.2da: added rare selects to bg1 vcremap
[09:31:06] <CIA-41> GemRB: 03npintin * r1f910a3fcc86 10gemrb/gemrb/core/Scriptable/Actor.cpp:
[09:31:06] <CIA-41> GemRB: Fixes the Actor::SelectActor and Actor::VerbalConstant to
[09:31:06] <CIA-41> GemRB: work for main characters, and improves VerbalConstant's
[09:31:06] <CIA-41> GemRB: sound selection
[09:40:36] <gembot> build #361 of autotools g++-4.4.5 is complete: Success [build successful] Build details are at http://buildbot.gemrb.org/builders/autotools%20g%2B%2B-4.4.5/builds/361
[11:09:21] --> pintin has joined #gemrb
[11:29:07] <lynxlynxlynx> pintin: i've made some cosmetic changes to your last two patches, so you'll have to reset
[11:29:18] <lynxlynxlynx> they were committed while you were out
[11:45:21] <pintin> lynx: cool np thanks
[11:45:40] --> Avenger has joined #gemrb
[11:45:40] --- ChanServ gives channel operator status to Avenger
[11:46:20] <Avenger> pintin: use 4 rare sound slots, and use vcremap to reduce it to 2 if needed.
[11:46:36] <fuzzie> oh, was that not done?
[11:47:00] <Avenger> in latest commit i see something like: +#define NUM_RARE_SELECT_SOUNDS 2 //in bg and pst it is actually 4 TODO: check
[11:47:02] <fuzzie> i had an hour of discussion in here about how vcremap should be a superset so that we can use maximum sets of sound slots, rather than matching bg2
[11:47:20] <Avenger> hehe, good, i totally agree
[11:47:33] <pintin> mmm yeah I ran into problems with that
[11:47:39] <pintin> (I think)
[11:48:02] <Avenger> yes, there could be problems, i didn't really think this over
[11:48:15] <fuzzie> i thought it over and i think it should be fine
[11:48:21] <Avenger> if we use all maximum slots, there would be overlaps
[11:48:24] <fuzzie> i didn't try it though :)
[11:48:29] <Avenger> unless we use a 'gemrb layout'
[11:48:33] <fuzzie> yes, i propose a gemrb layout
[11:48:44] <Avenger> then all will be fine
[11:48:45] <fuzzie> so our slots don't match *any* game, and we always have a vcremap
[11:49:19] <pintin> It was I couldn't map the extra slots for the rares in bg2 because they extra slots were used for other things in bg2
[11:49:28] <pintin> fuzzie: yeah that's what should be done
[11:50:12] <fuzzie> \o/
[11:50:35] <pintin> a 2da with something like NAME INDEX COUNT
[11:51:11] <pintin> and then when they are imported its used to put them into lists of different types of sounds
[11:51:28] <pintin> also do the same for main character
[11:51:28] <fuzzie> well i was thinking we can just hard-code gemrb slots
[11:51:42] <fuzzie> because that can actually be quickly done in just half an hour or so
[11:51:44] <pintin> then you don't have to do checking for wavs or whatver
[11:51:57] <Avenger> you always need to do that check
[11:52:03] <Avenger> that's how the original works
[11:52:10] <Avenger> they can leave slots empty
[11:52:44] <fuzzie> and you can change the soundsets etc
[11:52:51] <pintin> yeah but we could just do the check at creature import time and abstract away from the fact that main characters are handled differently in the original?
[11:52:55] <fuzzie> so it's not worth trying to integrate this
[11:53:07] <fuzzie> maybe as a seperate patch, yes
[11:53:26] <fuzzie> but you will waste way too much time if you try and do it at the same time as fixing the slots :)
[11:53:35] <pintin> it would obviously take quite a bit of code but it would simplify stuff so much
[11:53:48] <pintin> yeah probably
[11:55:26] <Avenger> pintin: i thought about that name/index/count 2da too, but doing the remap seemed more useful. It can do more, because it contains more information, and it is simply done
[11:56:59] <Avenger> you can make weighted slots, or select slots from different ranges.
[11:59:41] <Avenger> a universal internal layout is the only thing we really need. I just didn't do that because i was lazy creating complete maps for 5 games
[11:59:44] <pintin> it is simple so long as bg2 has the most of each type of sound...or at least I couldn't figure out a simple way to do it wiht vcremap
[11:59:57] <pintin> even though fuzzie convinced be it could be done :P
[12:00:29] <Avenger> yes, with an internal layout, you can even keep everything basically bg2, and add the extra slots in the end
[12:01:30] <pintin> ya
[12:01:42] <fuzzie> well it seems simpler to just have an internal layout where all the slots are grouped together
[12:01:55] <fuzzie> but then you do have to remap bg2 which is annoying
[12:02:46] <Avenger> i'm not against doing some of the conversion during load/save time
[12:03:08] <Avenger> so the internal format is like you want, all grouped together
[12:03:46] <pintin> ^^ I think thats the best solution in the long term
[12:05:32] <Avenger> but you still need to do that with maps, and a static internal format, it would be unnecessary to add another customisation inside for the global layout, no?
[12:06:10] <Avenger> mostly because missing slots need to be ignored, and the rest to be used
[12:07:25] <pintin> I mean each game has a map of its sounds -- then at load time you load refernces to the sounds into different lists in the actor class (ie actor->AttackSoundLIst)
[12:07:43] <pintin> then when you want to play a sound you call actor->playAttackSound
[12:07:57] <pintin> and all you have to do is play the next one on the list
[12:08:23] <pintin> use the maps at load time and then forget about them basically
[12:09:26] <Avenger> well, you can do that too
[12:09:59] <Avenger> just make sure you can regenerate the save game
[12:10:34] <pintin> ugh I don't know how the saves work but I wouldn't have thought it would matter....
[12:10:43] <Avenger> hehe
[12:10:45] <pintin> are the sounds a particular actor has used in the save game?
[12:10:51] <pintin> that would be annoying
[12:11:11] <Avenger> only the slots are in the save file
[12:11:20] <pintin> the strrefs?
[12:11:25] <Avenger> yes
[12:11:37] <pintin> omg...now I see why you did it like this :P
[12:12:13] <Avenger> well, take a look at creimporter :)
[12:12:52] <Avenger> atm, it is just doing:
[12:12:55] <Avenger> for (i=0;i<100;i++) {
[12:12:57] <Avenger> str->ReadDword( &act->StrRefs[i] );
[12:12:58] <Avenger> }
[12:13:08] <Avenger> that will be significantly more complicated with the lists :)
[12:13:36] <pintin> yep but it would make all the actor methods simpler to compensate
[12:13:38] <Avenger> and then, just stuff all back in the box
[12:13:55] <pintin> but if actor slots need to be saved that would make it not worth it probably
[12:14:17] <Avenger> theoretically, i prefer a common, simple and fast internal format, and delegate most of the repeated conversion into save/load
[12:14:36] <Avenger> but the save/load must stay binary compatible with the original
[12:14:42] <Avenger> that's hard sometimes
[12:16:40] <pintin> you mean the save load format just has to be the same?
[12:16:51] <pintin> I mean the save file format
[12:17:32] <Avenger> the same as in the original, yes
[12:18:00] <Avenger> or rather: as in the originalS, we have 5-6 different save formats to support
[12:18:42] <Avenger> we have a 'save as original' flag, which need to be honored
[12:19:01] <Avenger> if it is not set, then it basically a sixth (or seventh) format, that is the gemrb native forma
[12:19:03] <Avenger> t
[12:19:10] <pintin> right yep its a worthwhile feature!
[12:19:44] <Avenger> it is mostly working, it isn't entirely perfect, but i need it mostly for debugging
[12:20:18] <Avenger> and users sometimes also try to load a gemrb save back into the original
[12:21:19] <pintin> I would expect them to be compatible unless told otherwise :P
[12:22:32] <Avenger> yeah, but sometimes the original engine is quite a jerk
[12:22:40] <Avenger> especially pst with its variables
[12:22:54] <Avenger> that's totally silly
[12:23:35] <Avenger> if the save game contains a variable that isn't listed in var.var, the load crashes
[12:25:29] <pintin> it must be really complicated to handle all these games :P
[13:28:10] --> Yoshimo has joined #gemrb
[13:32:32] <-- Avenger has left IRC (Quit: ChatZilla 0.9.88 [Firefox 9.0.1/20111220165912])
[14:09:15] <CIA-41> GemRB: 03lynxlupodian * r412a4864dd16 10gemrb/ (configure.in gemrb/includes/globals.h): set the version number for 0.7.0
[14:27:05] <lynxlynxlynx> all tests went fine, time for uploading
[14:28:38] <CIA-41> GemRB: 03avenger_teambg * rfa55c90d2371 10gemrb/gemrb/core/ (Scriptable/Actor.cpp damages.h): hit sounds for all games, not just PST
[14:35:19] <CIA-41> GemRB: 03lynxlupodian * r8b1f3facc7e7 10gemrb/ (NEWS configure.in gemrb/includes/globals.h): restarted the development cycle
[14:35:20] <lynxlynxlynx> on mirrors, just announcements left
[14:45:42] <lynxlynxlynx> just the forums left
[14:45:52] <lynxlynxlynx> fire up the build machines :)
[14:45:59] --- lynxlynxlynx has changed the topic to: GemRB 0.7.0 | http://gemrb.org | Be wary of your words for there are Modron sensors in this channel: http://log.usecode.org/gemrblog.php | Hey <CHARNAME>, we need some awesome screenshots! | import pdb; pdb.set_trace()
[14:56:30] <lynxlynxlynx> ok, i'll wait for the buildbot to build with that last revision
[14:56:43] <lynxlynxlynx> windows users rejoice
[14:56:50] <lynxlynxlynx> if it is not buggy, of course :)
[14:59:00] <gembot> build #356 of mingw32 is complete: Exception [exception upload] Build details are at http://buildbot.gemrb.org/builders/mingw32/builds/356 blamelist: avenger_teambg@sourceforge.net, lynxlupodian@users.sourceforge.net
[15:01:50] <lynxlynxlynx> bah
[15:03:32] <lynxlynxlynx> gembot: force build mingw32 trouble
[15:03:32] <gembot> builder control is not enabled
[15:08:37] <lynxlynxlynx> tomprince: can you help?
[15:15:19] <Yoshimo> why should a windows user be happy?
[15:20:03] <lynxlynxlynx> rfa55c90d2371 gemrb/gemrb/core/ (Scriptable/Actor.cpp damages.h): hit sounds for all games, not just PST <-- this revision will be uploaded, not the previous one
[15:28:28] --> Maighstir has joined #gemrb
[15:31:48] <lynxlynxlynx> cool, new droid builds already
[15:40:50] <-- Maighstir has left IRC (Quit: .)
[15:41:24] --> Maighstir has joined #gemrb
[16:00:29] <lynxlynxlynx> ok, msvc is done; good enough for me
[16:06:37] <-- gembot has left IRC (Ping timeout: 252 seconds)
[16:17:22] * lynxlynxlynx summons avenger
[16:17:26] <lynxlynxlynx> http://pastebin.com/VPW41SeU
[17:01:21] <-- duckpunch has left IRC (Quit: leaving)
[17:20:10] --> brad_a has joined #gemrb
[17:38:57] --> gembot has joined #gemrb
[18:16:22] <tomprince> brad_a: The xcode command won't build the package?
[18:16:34] <brad_a> output?
[18:17:14] <tomprince> I haven't run it yet, I am just setting it up.
[18:17:39] <tomprince> I was just curious why xcode wasn't running the packagemanager command itself.
[18:17:42] <brad_a> oh no the xcode command doesnt but it could if i appended it to the buildscript
[18:18:00] <brad_a> i just dont want to because i dont want to build the package everytime i build while developing
[18:18:07] <brad_a> i guess i could make a buildbot target
[18:18:13] <brad_a> thats probably the best solution
[18:18:24] <brad_a> that way it can build a release build instead of debug build
[18:18:25] <tomprince> Well, not a buildbot target, but a packaging target.
[18:18:30] <brad_a> sure
[18:18:34] <brad_a> same thing :-P
[18:18:38] <tomprince> :)
[18:18:44] <tomprince> Names matter though. :)
[18:18:44] <brad_a> yes i will do that
[18:19:02] <brad_a> ok i will name it release package or something
[18:21:34] <tomprince> brad_a: Can you make the output include the version, or allow me to pass in the version?
[18:21:46] <brad_a> i think so
[18:21:53] <brad_a> is there someplace i can get that?
[18:22:27] <tomprince> I use ["git", "describe", "--tags", "--dirty=-patched"]
[18:22:37] <brad_a> neat
[18:22:45] <tomprince> I can also pass that to xcode.
[18:23:01] <brad_a> well id rather set it up automatically if i can
[18:23:16] <brad_a> but it should be doable one way or the other
[18:26:31] <brad_a> yup its pretty easy to do looks like
[18:30:18] <-- Maighstir has left #gemrb
[18:31:04] --> Maighstir has joined #gemrb
[18:33:07] <-- gembot has left IRC (Remote host closed the connection)
[18:33:18] --> gembot has joined #gemrb
[18:38:14] <brad_a> tomprince: how do i get just the base version number?
[18:38:37] <brad_a> i have 2 fields. 1 for something like 0.7.0 and one for a build number
[18:38:50] <-- gembot has left IRC (Quit: buildmaster reconfigured: bot disconnecting)
[18:39:04] --> gembot has joined #gemrb
[18:39:18] <-- gembot has left IRC (Remote host closed the connection)
[18:39:32] --> gembot has joined #gemrb
[18:40:39] <brad_a> meh it probably doesnt matter that its longer than expected
[18:46:27] <tomprince> brad_a: The command I gave you gives you the version number. It gives more than gust a 0.7.0, since it bulids for every commit, and we want to differentiate them.
[18:46:47] <brad_a> i know but that what the other field is for. it doesnt matter tho
[18:47:07] <-- gembot has left IRC (Quit: buildmaster reconfigured: bot disconnecting)
[18:47:17] --> gembot has joined #gemrb
[18:47:52] <brad_a> i dont know why finder is stll telling me gemrb 0.6.4 there is not such number in its plist...
[18:49:00] <-- gembot has left IRC (Remote host closed the connection)
[18:49:16] --> gembot has joined #gemrb
[18:49:34] <tomprince> I'd guess it is for when you build multiple times, from the same revision.
[18:52:09] <gembot> build #2 of osx-xcode-binary is complete: Success [build successful] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/2
[18:53:13] <brad_a> bah hopefully its just some stupid cache issue
[18:53:24] <brad_a> cuz i cant see anywhere in the project referencing it
[19:00:43] <brad_a> cool found it
[19:00:53] <brad_a> stupid locilization ;-)
[19:01:01] <brad_a> localization
[19:04:34] <CIA-41> GemRB: 03bradallred * r2c2856ddd9cb 10gemrb/apple/ (4 files in 4 dirs): Xcode: automatically update version number and build number using git.
[19:06:01] <tomprince> brad_a: Apprently xcode doesn't error out if it fails to build things. It looks like the above build failed to build SDLVideo.
[19:08:48] <-- brad_a has left IRC (*.net *.split)
[19:12:40] --> brad_a has joined #gemrb
[19:12:48] <-- brad_a has left #gemrb
[19:12:57] --> brad_a has joined #gemrb
[19:14:27] <brad_a> tomprince: it should at least throw a warning. where is the output?
[19:18:35] <brad_a> thats interesting
[19:19:19] <-- gembot has left IRC (*.net *.split)
[19:19:20] <-- DrMcCoy has left IRC (*.net *.split)
[19:21:52] --> gembot has joined #gemrb
[19:21:52] --> DrMcCoy has joined #gemrb
[19:22:14] <brad_a> looks like perhapps an environment diffrence betweeen building in xcode and on the command line
[19:23:15] <-- tomprince has left IRC (*.net *.split)
[19:23:16] <-- CIA-41 has left IRC (*.net *.split)
[19:23:17] <-- wjp has left IRC (*.net *.split)
[19:23:17] <-- brad_a has left IRC (*.net *.split)
[19:24:05] --> brad_a has joined #gemrb
[19:24:05] --> tomprince has joined #gemrb
[19:24:05] --> CIA-41 has joined #gemrb
[19:24:05] --> wjp has joined #gemrb
[19:34:21] <brad_a> tomprince: do i need to worry about having the project clean itself or does buildbot just nuke everything anyway?
[19:34:48] --> Maighstir_ has joined #gemrb
[19:35:24] <-- Maighstir_ has left #gemrb
[19:36:13] <brad_a> new build command assuming a clean isnt necessary: xcodebuild -project GemRB.xcodeproj -target ReleasePackage -configuration Release
[19:36:29] <brad_a> jus tneed to push it. and figure out why sdl plugin isnt building
[19:45:13] <CIA-41> GemRB: 03bradallred * r5fed4e904bb0 10gemrb/apple/GemRB.xcodeproj/project.pbxproj: Xcode: add a "ReleasePackage" target to build release configurations for both mac and iOS. Creates an iOS IPA and a mac pkg bundle.
[20:18:35] <tomprince> brad_a: What files does it produce?
[20:19:17] <brad_a> ios/GemRB.ipa
[20:19:28] <brad_a> osx/Install GemRB.pkg
[20:19:40] <tomprince> No version in the file names?
[20:19:48] <brad_a> i can make that happen
[20:20:56] <tomprince> If you use the string generated by the above command, or something passed in, I already have it as a variable in buildbot.
[20:22:43] <tomprince> I would pick GemRB-ios-%(version).ext' and 'GemRB-osx-%(version).ext'
[20:24:33] <brad_a> ok
[20:28:58] <brad_a> im not sure if changing the ipa name will affect the name in itunes or on the device tho
[20:29:52] <tomprince> Well, that probably wouldn't be a bad thing for auto-builds. Can always rename it for release versions.
[20:30:14] <brad_a> sounds good
[20:30:18] <-- gembot has left IRC (Quit: buildmaster reconfigured: bot disconnecting)
[20:30:28] --> gembot has joined #gemrb
[20:35:21] <brad_a> tomprince: it this target is going to be build by buildbot each time then we dont need the existing ios one
[20:35:35] <brad_a> the mac one is using cmake so that is still usefull
[20:45:44] <-- gembot has left IRC (Quit: buildmaster reconfigured: bot disconnecting)
[20:45:58] --> gembot has joined #gemrb
[20:52:07] <CIA-41> GemRB: 03bradallred * r03c14d416efa 10gemrb/apple/GemRB.xcodeproj/project.pbxproj: Xcode (Release Target): use the version string in the distribution file name and Configure the release configuration to not have a dSYM file.
[20:52:07] <CIA-41> GemRB: 03bradallred * r967c19184efd 10gemrb/apple/GemRB.xcodeproj/project.pbxproj: Xcode: trim some more useless fat. unneeded settings from the plugins.
[20:52:37] <brad_a> tomprince: i used the naming scheme you suggested
[21:08:21] <gembot> build #6 of osx-xcode-binary is complete: Failure [failed compile] Build details are at http://buildbot.gemrb.org/builders/osx-xcode-binary/builds/6 blamelist: bradallred@me.com
[21:25:49] <gembot> build #56 of nmake-msvc++10 is complete: Success [build successful] Build details are at http://buildbot.gemrb.org/builders/nmake-msvc%2B%2B10/builds/56
[21:43:31] --> SiENcE has joined #gemrb
[22:19:47] <-- Maighstir has left IRC (Quit: .)
[22:41:49] <-- SiENcE has left IRC (Quit: cya)
[23:07:37] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:19:59] --> brad_a_ has joined #gemrb
[23:19:59] <-- brad_a_ has left IRC (Client Quit)
[23:23:07] <-- brad_a has left IRC (Ping timeout: 240 seconds)
[23:31:57] <tomprince> lynx: Is there a way to automate uploads to sf, and do we have a limit on our space there?
[23:32:19] <tomprince> I was thinking of moving the build products off my server.