#gemrb@irc.freenode.net logs for 13 Sep 2012 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[01:27:37] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[02:16:56] --> edheldil_ has joined #gemrb
[02:27:42] <-- edheldil_ has left IRC (Ping timeout: 252 seconds)
[05:24:41] --> edheldil_ has joined #gemrb
[05:31:54] <-- edheldil_ has left IRC (Ping timeout: 264 seconds)
[07:05:23] --> edheldil_ has joined #gemrb
[07:16:38] <-- edheldil_ has left IRC (Ping timeout: 272 seconds)
[07:17:26] --> lynxlynxlynx has joined #gemrb
[07:59:19] --> Husar1 has joined #gemrb
[08:03:08] --> edheldil_ has joined #gemrb
[09:21:56] <CIA-45> GemRB: 03lynxlupodian * rdafd510017f3 10gemrb/CMakeLists.txt: cmake: incorporated part of netbsd's downstream patch
[09:21:56] <CIA-45> GemRB: 03lynxlupodian * r829453dfb752 10gemrb/CMakeLists.txt: cmake: converted rogue space indentation
[09:32:46] <CIA-45> GemRB: 03lynxlupodian * rf3d3ff6902f4 10gemrb/autogen.sh: autogen: sh compliance fix from Thomas Klausner
[09:42:19] <CIA-45> GemRB: 03lynxlupodian * r828b5f420fea 10gemrb/ (CMakeLists.txt gemrb/CMakeLists.txt):
[09:42:19] <CIA-45> GemRB: cmake: added a separate path var for the example configs
[09:42:19] <CIA-45> GemRB: netbsd likely to install them elsewhere
[09:48:00] --> rocket_hamster has joined #gemrb
[11:21:29] <-- rocket_hamster has left IRC ()
[11:35:22] --> lynxlynxlynx_ has joined #gemrb
[11:42:24] <-- lynxlynxlynx has left IRC (*.net *.split)
[11:54:39] <-- nutron has left IRC (Ping timeout: 240 seconds)
[11:56:02] --> nutron has joined #gemrb
[12:25:48] <-- edheldil_ has left IRC (Ping timeout: 253 seconds)
[12:25:48] <-- edheldil has left IRC (Ping timeout: 288 seconds)
[12:26:57] --> edheldil_ has joined #gemrb
[12:27:45] --> edheldil has joined #gemrb
[12:27:45] --- ChanServ gives channel operator status to edheldil
[12:35:31] --> rocket_hamster has joined #gemrb
[12:55:16] --> gembot_ has joined #gemrb
[12:56:11] --> lynxlynxlynx has joined #gemrb
[12:56:11] <-- lynxlynxlynx has left IRC (Changing host)
[12:56:11] --> lynxlynxlynx has joined #gemrb
[12:56:12] --- ChanServ gives channel operator status to lynxlynxlynx
[12:57:21] --> kida_ has joined #gemrb
[13:02:22] <-- lynxlynxlynx_ has left IRC (*.net *.split)
[13:02:23] <-- gembot has left IRC (*.net *.split)
[13:02:23] <-- kida_laptop has left IRC (*.net *.split)
[13:04:16] <lynxlynxlynx> cool, 0.7.1 still builds on netbsd (was on 0.6.5 before)
[13:17:35] <-- rocket_hamster has left IRC (Ping timeout: 248 seconds)
[13:24:30] --> rocket_hamster has joined #gemrb
[13:44:35] <-- rocket_hamster has left IRC (Ping timeout: 244 seconds)
[13:54:30] <-- fuzzie has left IRC (Ping timeout: 250 seconds)
[13:58:42] --> fuzzie has joined #gemrb
[15:55:28] <-- duckpunch has left IRC (Ping timeout: 252 seconds)
[15:56:12] --> duckpunch has joined #gemrb
[16:03:43] --> avenger has joined #gemrb
[16:03:43] --- ChanServ gives channel operator status to avenger
[16:04:26] <avenger> i think the buildbot is a bit over eager now. Is it possible to tune it down to weekly releases?
[16:11:31] <tomprince> Should just clean out the release.
[16:11:51] <tomprince> This way, if somebody reports a problem, the buildbots will build binaries to test the fix.
[16:14:30] --> Earnestly has joined #gemrb
[16:15:10] <lynxlynxlynx> welcome back
[16:15:17] <lynxlynxlynx> the beta finally over?
[16:18:27] <edheldil> Hi, Avenger
[16:18:50] <avenger> nah
[16:19:10] <avenger> still working on it, and then bg2ee :D
[16:19:51] <avenger> but last week i commited something already :D
[16:20:54] <lynxlynxlynx> :)
[16:21:26] <lynxlynxlynx> fuzzie has a prehistoric question for you
[16:21:41] <lynxlynxlynx> as usual
[16:21:52] --> brada has joined #gemrb
[16:22:06] <brada> lynx: thanks for getting that merged
[16:22:07] <lynxlynxlynx> now the party is complete :)
[16:22:14] <brada> me?
[16:22:25] <lynxlynxlynx> np, it was on my todo anyway, you just saved me some work
[16:22:29] <lynxlynxlynx> yep
[16:22:56] <wjp> lynxlynxlynx: so now we can venture forth? ;-)
[16:23:02] <brada> ha ha
[16:23:12] <lynxlynxlynx> :)
[16:23:31] <avenger> fuzzie seems to be afar
[16:23:52] <fuzzie> oh, right
[16:24:02] <lynxlynxlynx> oh right
[16:24:16] <lynxlynxlynx> fuzzie: can you fix all those leavearea problems for the next release?
[16:24:39] <avenger> the question is about the movie player flag?
[16:24:46] <fuzzie> I was wondering when the GUIScript function PlayMovie needs to be called, when a movie should only be played once.
[16:24:53] <Earnestly> I was hoping someone might be able to help me with an error I can't seem to find on google (or I'm failing at google-fu). Everything compiles successfully (arch linux) but aftr creating a basic gemrb.cfg for pst I get the following error: "gemrb: symbol lookup error: /usr/lib/gemrb/libgemrb_core.so: undefined symbol: dlopen"
[16:25:23] <avenger> i wanted to play the starting movies only once
[16:25:48] <lynxlynxlynx> Earnestly: did you use the build from abs?
[16:26:35] <avenger> by the way, it seems bg2 doesn't let you break a movie when it was played first, or is that just my feeling?
[16:26:49] <fuzzie> avenger: Only once ever though? (since it checks .ini entry)
[16:26:56] <lynxlynxlynx> brada: i forgot what you were working on before sidetracking to guiopt :s
[16:27:07] <brada> there were a couple things
[16:27:13] <brada> 1 was fixing the status icons
[16:27:18] <fuzzie> I forget what original bg2 behaviour was there.
[16:27:21] <brada> and the other big one was chapter text
[16:27:24] <lynxlynxlynx> oh right, the alignment stuff
[16:27:35] <fuzzie> I am bured in university work right now, but the leavearea stuff looked simple to fix.
[16:27:36] <avenger> fuzzie: someone might have implemented that independently :)
[16:27:36] <brada> both kinda rely on us figuring out something with the fonts first
[16:28:02] <fuzzie> avenger: you first added it with comment "if it was already played once (set in .ini)" :)
[16:28:05] <lynxlynxlynx> fuzzie: that'd be great
[16:28:23] <lynxlynxlynx> i'll just make the problem worse with >6 char parties
[16:28:25] <fuzzie> Right now I am trying to remember how differential equations work, unfortunately.
[16:28:43] <fuzzie> So telling me the release schedule might be a good idea if it's really soon.
[16:28:44] <avenger> fuzzie that is not in the guiscript, is it?
[16:28:56] <avenger> i tried to implement it in the core
[16:29:00] <lynxlynxlynx> definitely not before the end of the month
[16:29:11] <fuzzie> avenger: well, it is in GUIScript.cpp
[16:29:16] <avenger> yes
[16:29:21] <avenger> that was me
[16:29:32] <fuzzie> but all the GUIScript callers got changed to pass the "play always" flag
[16:29:34] <lynxlynxlynx> i have some more config stuff to do: cleanup, load/save, storing in the ini
[16:29:41] <fuzzie> except for the credits video and iwd2's movies screen
[16:30:20] <avenger> yes, but i think that wasn't me :D
[16:30:23] <fuzzie> and neither of those make sense
[16:30:33] <fuzzie> so now it's just a mess :P
[16:30:34] <Earnestly> lynxlynxlynx: ABS? Arch Build System? Unfortunately there is not official gemrb package, but it does exist in the AUR which is what I used, the non-git version
[16:30:37] <lynxlynxlynx> wasn't there a problem that you couldn't play movies due to one of these changes?
[16:30:47] <Earnestly> In fact the both resulted in the same error
[16:30:48] <avenger> someone implemented if not skip_videos and not GemRB.GetVar ("SeenIntroVideos"):
[16:31:07] <fuzzie> well, of course that works differently, because it doesn't check .ini (which would disable them forever) but only in a single session
[16:31:11] <lynxlynxlynx> Earnestly: ok; i'll take a look what it does. This machine is also on arch, but i build manually
[16:31:17] <Earnestly> /not official/noofficial/
[16:31:26] <fuzzie> so, I don't really care, but it would be nice to have a decision on what we should do
[16:31:33] <Earnestly> Thanks, I'll be back in a moment
[16:31:34] <avenger> my intention was to disable them forever after played once
[16:31:45] <avenger> but that is admittedly not the same as the original
[16:31:50] <avenger> i think someone didn't like that
[16:31:54] <avenger> and added the flags
[16:32:34] <avenger> i didn't want to do a commit war over this, so i didn't change it anymore, it is not too important to me
[16:33:18] <lynxlynxlynx> there are at least 3 different cases here
[16:33:24] <lynxlynxlynx> not counting combinations
[16:34:10] <lynxlynxlynx> we have the skipintro setting to skip intros AND we also make sure that if it wasn't set, you only see it the first time when returning to Start.py (eg. from options or other menus)
[16:34:41] <lynxlynxlynx> that's what that quoted IF should be about
[16:34:54] <avenger> yes, playing the intro videos in the same session is definitely unwanted
[16:35:01] <lynxlynxlynx> but that's all just about intro videos and has nothing to do with flags
[16:35:04] <avenger> i mean twice
[16:35:46] <lynxlynxlynx> (archers also hit the py 2/3 bug we just fixed)
[16:36:32] <fuzzie> well, the one we just worked around, I'm not sure it's our bug
[16:37:07] <lynxlynxlynx> sure
[16:37:47] <Earnestly> lynxlynxlynx: Relevant: https://mailman.archlinux.org/pipermail/arch-dev-public/2012-September/023558.html
[16:38:07] <lynxlynxlynx> Earnestly: nothing suspicious in the recipe
[16:38:25] <Earnestly> lynxlynxlynx: Does it work on your machine?
[16:38:39] <avenger> what bug are you talking about?
[16:38:47] <fuzzie> avenger: cmake python bug
[16:38:58] <lynxlynxlynx> nothing to get upset about
[16:39:00] <avenger> ah ok, i thought some original IE bug :)
[16:39:34] <lynxlynxlynx> Earnestly: i doubt it, it can't find symbols from glibc in your error
[16:39:55] <Earnestly> avenger: http://ix.io/2ZH
[16:40:01] <avenger> its funny how many original IE bugs we found while testing bgee :D
[16:40:12] <lynxlynxlynx> http://www.gemrb.org/wiki/doku.php?id=installation#compile_it_yourself
[16:40:18] <lynxlynxlynx> i suggest you do that
[16:40:33] <lynxlynxlynx> get the git version for some extra fixes
[16:40:49] <Earnestly> the git version caused the same error
[16:41:07] <lynxlynxlynx> sure, i meant with manual compilation
[16:41:07] <Earnestly> But I'll do so manually anyway
[16:41:12] <Earnestly> Yup
[16:41:26] <lynxlynxlynx> pacman probably does extra stripping and whoknowswhat
[16:41:47] <Earnestly> pacman doesn't, makepkg can strip debug symbols
[16:42:02] <Earnestly> (which can be disabled)
[16:42:30] <lynxlynxlynx> dlopen is not just a debug symbol :)
[16:43:41] <Earnestly> That's true
[16:50:26] <Earnestly> Now I get "[PluginMgr]: Loading Plugins from /usr/local/lib/gemrb/plugins/"
[16:50:33] <Earnestly> Oh wait, nevermind
[16:55:02] --> Yoshimo has joined #gemrb
[16:57:05] <Earnestly> At any rate, the manually built version seems to work. I'll fix up my own PKGBUILD later this evening using the same steps.
[16:57:25] <lynxlynxlynx> it's already using the same stuff
[16:57:46] <lynxlynxlynx> btw, you can run it from the build dir, no need to install
[16:58:31] <Earnestly> It can't without hacking the paths and I much rather have my packages (or anything under /usr) to be managed
[16:59:08] <Earnestly> PKGBUILDs are also a cakewalk to write. The git build uses "cmake -DCMAKE_INSTALL_PREFIX=/usr -DINSOURCEBUILD=1 ." would that break it?
[16:59:19] <-- edheldil_ has left IRC (Ping timeout: 272 seconds)
[16:59:30] <Earnestly> (Sorry, I have very little experience with cmake)
[17:00:20] <Earnestly> Is it possible to not use the INSTALL_PREFIX and rather use `make --PREFIX=/usr ... install' instead?
[17:01:16] <lynxlynxlynx> if anything make PREFIX=/... but i don't know if cmake respects that
[17:01:22] <lynxlynxlynx> why complicate the matter?
[17:01:24] <Earnestly> :-/
[17:01:37] <Earnestly> Because arch uses /usr/bin
[17:01:52] <lynxlynxlynx> so does everyone else
[17:02:13] <Earnestly> Really? I thought the /usr merge was still lagging on other distros
[17:02:27] <lynxlynxlynx> we don't install anything to /lib or /bin
[17:02:52] <Earnestly> Oh good, so those two options probably aren't even needed.
[17:02:57] <lynxlynxlynx> for running from the build dir, you'd have to change your config, not the build
[17:03:14] <Earnestly> Yup
[17:03:22] <lynxlynxlynx> the first definitely not, since the fhs mode is default
[17:15:01] <Earnestly> I'll toy around with it, the maintainer of the AUR package is probably just doing it wrong.
[17:16:48] <-- avenger has left IRC (Quit: http://quassel-irc.org - Chat comfortably. Anywhere.)
[17:37:25] <lynxlynxlynx> that bit is fine, there isn't just one way of doing things
[17:44:35] <Earnestly> lynxlynxlynx: Just testing my build flags now, I usually use -flto which seems relevant, building now without it.
[17:45:25] <Earnestly> Since the only difference between the pkgbuild and a manual make is the CXXFLAGS
[17:45:56] <Earnestly> Well whadda' know.
[17:46:12] <Earnestly> Good ol' planescape up and running. :>
[17:46:28] <Earnestly> It was the -flto flag (link time optimisation)
[17:46:45] <Earnestly> Also, native planescape feels so good, thank you :-)
[17:47:08] --> edheldil_ has joined #gemrb
[17:50:11] <Earnestly> (http://gcc.gnu.org/wiki/LinkTimeOptimization)
[17:54:37] <lynxlynxlynx> for now
[17:56:09] <Earnestly> Oh?
[17:57:20] <Earnestly> RetroArch for arcade games and Gemrb (which is totally confusing if you code in ruby btw) for all of the decent PC games. The only thing that's left is FO1/2 and I'd be set. Thanks guys
[18:12:40] <Earnestly> Anyway, thanks for the assistence. Everything is working wonderfully after building without -flto. Ta
[18:12:54] <-- Earnestly has left #gemrb ("WeeChat 0.3.9-dev")
[18:13:10] <-- brada has left IRC (Quit: brada)
[18:29:48] <lynxlynxlynx> a bit too fast
[18:36:58] <wjp> I wonder why that breaks. Maybe core should be explicitly linked against libdl?
[18:39:04] <lynxlynxlynx> it is
[18:40:06] <lynxlynxlynx> i was thinking of just passing -fno-lto where supported, but it would be nice to know more first
[18:44:37] <wjp> it is?
[18:45:04] <wjp> I don't see it in libgemrb_core.so
[18:45:22] <wjp> only libstdc++, libm, libgcc_s, libc
[18:50:18] <fuzzie> right, only the main binary gets CMAKE_DL_LIBS
[18:50:53] <fuzzie> oh, no, in the !STATIC_LINK case, gemrb_core should get it too .. ?
[18:51:38] <fuzzie> oh no, misreading it. was right in first place.
[18:51:41] <fuzzie> so I guess we should add that.
[18:51:53] <fuzzie> you'd think I'd learn to stop reading code in the evenings..
[19:11:57] --> brada has joined #gemrb
[19:11:58] <lynxlynxlynx> ah right, i was looking at the wrong code too
[19:16:21] <lynxlynxlynx> mine already links against it, but maybe it's from bloat
[19:16:47] <CIA-45> GemRB: 03lynxlupodian * r1430f82477bb 10gemrb/CMakeLists.txt: cmake: explicitly link to libdl, hopefully fixes -flto uses
[19:37:05] <-- kida_ has left IRC (Ping timeout: 260 seconds)
[19:41:55] <wjp> hm, my libgemrb_core.so still has no NEEDED libdl.so, but I don't know if that's expected or not
[19:43:44] <fuzzie> there's no libdl referenced from my libgemrb_core.so
[19:44:40] <wjp> I have *UND* dlopen/dlclose in there
[19:45:04] --> kida_ has joined #gemrb
[19:45:37] <fuzzie> me too
[19:45:55] <wjp> ugh, my brain doesn't seem to want to focus on linking issues tonight
[19:46:36] <fuzzie> of course all my plugins now link explicitly against it :-p
[19:46:40] <fuzzie> but core isn't one
[19:47:54] <wjp> heh, indeed
[19:48:01] <wjp> I don't remember if that was the case previously though
[19:48:37] <fuzzie> that is interestingly weird really
[19:49:05] <fuzzie> I get a NEEDED if I rebuild after having moved it to the right (core-only) place.
[19:49:32] <fuzzie> let me do a full rebuild
[19:49:57] <wjp> ah, but that last patch did add it to the ADD_GEMRB_PLUGIN macro of course
[19:50:49] <fuzzie> yes, I get it on reubild
[19:50:51] <fuzzie> let me commit this anyway
[19:51:50] <fuzzie> especially since I have no idea what I'm doing, it'll be fun
[19:52:01] <fuzzie> we can play "who makes the third commit"
[19:52:23] <CIA-45> GemRB: 03fuzzie * r74766730cd16 10gemrb/ (CMakeLists.txt gemrb/core/CMakeLists.txt): cmake: explicitly link core to libdl instead (see 1430f824)
[19:53:11] <wjp> that seems to have done the trick
[19:55:38] <fuzzie> well, I mean, I still get it for plugins
[19:57:25] <wjp> same here
[19:57:39] <wjp> but I suppose that doesn't hurt
[20:01:07] <wjp> it wasn't the case in master^^, by the way, so I guess it's taking a transitive closure when linking?
[20:02:26] <fuzzie> that is a rather nerdery way to put it, but that would be my guess, yes :-)
[20:05:39] <wjp> I spent half a minute trying to think of a simpler term without much luck :-)
[20:08:13] <-- Yoshimo has left IRC (Quit: Yoshimo)
[20:10:37] --> rocket_hamster has joined #gemrb
[20:40:07] <-- rocket_hamster has left IRC ()
[21:26:08] <-- |Cable| has left IRC (Read error: Operation timed out)
[21:39:55] --> |Cable| has joined #gemrb
[22:29:53] <-- CIA-45 has left IRC (Read error: Operation timed out)
[22:41:23] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[22:54:06] <-- brada has left IRC (Quit: brada)
[23:28:34] --> brada has joined #gemrb