#exult@irc.freenode.net logs for 12 Apr 2016 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[00:12:10] <-- frob has left IRC (Ping timeout: 252 seconds)
[01:05:36] --> GitHub has joined #exult
[01:05:36] <GitHub> [exult] marzojr pushed 1 new commit to master: https://git.io/vVFY5
[01:05:36] <GitHub> exult/master f05119d Marzo Sette Torres Junior: * configure.ac, headers/common_types.h: Trying a better way of setting the...
[01:05:36] <-- GitHub has left #exult
[01:05:47] <Marzo> Dominus: when you can try, there it is ^
[01:15:06] --> DominusExult has joined #exult
[01:15:06] --- ChanServ gives channel operator status to DominusExult
[01:18:47] <-- Dominus has left IRC (Ping timeout: 250 seconds)
[01:18:47] --- DominusExult is now known as Dominus
[01:26:37] <Marzo> Hi Dominus
[03:27:51] --> GitHub has joined #exult
[03:27:51] <GitHub> [exult] marzojr pushed 2 new commits to master: https://git.io/vVF4E
[03:27:51] <GitHub> exult/master 6784697 Marzo Sette Torres Junior: * mapedit/*: Eliminating some -Wold-style-cast warnings that somehow escaped...
[03:27:51] <GitHub> exult/master c3ed0a0 Marzo Sette Torres Junior: * mapedit/*: Eliminating -Woverloaded-virtual warnings by making explicit...
[03:27:51] <-- GitHub has left #exult
[04:37:46] <Dominus> Marzo, now it's back to erroring with objserial :(
[04:37:52] <Dominus> begins the same as before http://pastebin.com/i44wF0xh
[05:20:43] --> frob has joined #exult
[05:25:40] <-- frob has left IRC (Ping timeout: 252 seconds)
[07:15:58] --> ttarrant has joined #exult
[07:38:30] <-- ttarrant has left IRC (Quit: I'm off)
[07:43:17] --> ttarrant has joined #exult
[07:48:00] --> frob has joined #exult
[07:53:00] <-- frob has left IRC (Ping timeout: 276 seconds)
[08:01:46] <Dominus> Marzo, wjp, as I'm cleaning out ifdef MACOS, I'm puzzled how something like the one in audio/Audio.cc:59 works
[08:02:09] <Dominus> it says #if defined(MACOS) include blabla
[08:02:49] <Dominus> and then before the #endif it says #if !defined(UNDER_CE) include blabla
[08:03:22] <Dominus> so if I get rid of the #if defined(MACOS), I can get rid of the whole block?
[08:03:52] <Dominus> I'm puzzled by this
[08:04:56] <Dominus> if that is only when defined MACOS why go to the length of add !defined (UNDER_CE)?
[08:08:05] <Dominus> or does that work like an "else if"? that when neither MACOS or UNDER_CE are defined those get included?
[08:09:09] <Dominus> That's how I understand files/utils.cc:34
[08:40:29] <Dominus> another purge question wjp, Marzo: in https://github.com/exult/exult/commit/62f7326bf207f0e4916225cbbe01fd855f118b87 7 years ago, Fingolfin started to get rid of ALPHA_LINUX_CXX. Is it save to just go further ahead with this? He only purged some few, not nearly everything...
[09:05:19] --> ttarrant_ has joined #exult
[09:05:38] <-- ttarrant has left IRC (Read error: Connection reset by peer)
[09:06:06] --- ttarrant_ is now known as ttarrant
[09:50:31] <-- Lightkey has left IRC (Ping timeout: 268 seconds)
[10:03:07] --> Lightkey has joined #exult
[10:15:17] --> frob has joined #exult
[10:20:07] <-- frob has left IRC (Ping timeout: 268 seconds)
[11:09:25] <-- ttarrant has left IRC (Ping timeout: 252 seconds)
[11:42:13] <Marzo> Dominus: what happens when you apply this patch: http://pastebin.com/FixPC5CX
[11:49:12] <Dominus> Marzo, that seems to work with my fast 64bit build. I'm now rebuilding the long one for 32/64 and ppc... will report back in a coupl eof minutes :)
[11:49:23] <Marzo> k
[11:49:42] <Marzo> I am thinking that the only issue will be in Windows
[11:50:05] <Dominus> ppc failed
[11:50:13] <Dominus> it's using a old gcc
[11:50:24] <Marzo> Can you post the error log?
[11:50:25] <Dominus> *BUT* we can drop PPC
[11:50:43] <Marzo> Well, I'd still like to see the error log anyway :-p
[11:51:54] <Dominus> http://pastebin.com/BFBYqyJJ
[11:53:59] <Dominus> it fails at
[11:54:00] <Dominus> checking whether we are using clang on Apple OS... configure: error: in `/Users/Dominus/code/snapshots/exult':
[11:54:15] <Dominus> becuase I'm cross compiling...
[11:54:29] <Dominus> maybe needs more cross compile magic
[11:56:26] --> ttarrant has joined #exult
[12:04:05] <Dominus> interestingly, it shouldm't have gone on from that point
[12:04:35] <Dominus> becuase it is not clang I'm using for ppc
[12:04:54] <Dominus> so if ${ax_cv_cxx_using_clang_on_apple+:} false; then : $as_echo_n "(cached) " >&6
[12:05:05] <Dominus> should have been the result
[12:05:38] <Dominus> i686-apple-darwin10-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5494)
[12:05:59] <Dominus> ^^^/opt/xcode3/usr/bin/gcc-4.0 --version
[12:10:11] <Dominus> can I define ax_cv_cxx_using_clang_on_apple to false before the test through some configure/export?
[12:15:05] <Dominus> Hmm an answer file should work. Will look later into it...
[12:17:32] <Marzo> The issue there is that I compile a test program to check that
[12:20:35] <Marzo> I suppose I could try to find out if the compiler is clang before that
[12:24:58] <Dominus> nah, I get further with a cache file where the answer no is given for that
[12:25:28] <Marzo> Dominus: does it work if you replace m4/ax_cxx_compile_stdcxx_11.m4 by this: http://pastebin.com/Ay888PeC
[12:26:31] <Marzo> (test also the iOS build, please)
[12:28:20] <Dominus> marzo, that m4 replacement works for ppc
[12:29:22] <Marzo> Out of curiosity, can you post one of the compile lines?
[12:29:27] <Dominus> ios will have to wait. or rather the way *we* use configure etc. for iOS is obsolete. I will have to do it for Xcode
[12:30:10] <Marzo> Just to see if it is not sneaking anything stupid from the cache into the ppc build
[12:30:40] <Dominus> ok: /bin/sh ../../libtool --tag=CXX --silent --mode=compile /opt/xcode3/usr/bin/g++-4.0 -arch ppc -DHAVE_CONFIG_H -I. -I../.. -I./../../headers -I./../.. -I./../../files -I./../../shapes -DMACOSX -O2 -Wno-long-long -I/opt/ppc/include -isysroot /opt/SDKs/MacOSX10.4u.sdk -mmacosx-version-min=10.4 -I/opt/ppc/include -isysroot /opt/SDKs/MacOSX10.4u.sdk -mmacosx-version-min=10.4 -I/opt/ppc/include -isysroot /opt/SDKs/MacOSX10.4u.sdk -mmacosx-ve
[12:30:40] <Dominus> rsion-min=10.4 -arch ppc -m32 -O2 -MT frnameinf.lo -MD -MP -MF .deps/frnameinf.Tpo -c -o frnameinf.lo frnameinf.cc
[12:31:05] <Marzo> Looks fine
[12:32:11] <Marzo> Anyways, so things seem to work fine with the modified objserial and ppc works with this .m4 change
[12:33:36] <Dominus> let me see if it still compiles non ppc :)
[12:34:08] <Dominus> hmm, odd
[12:34:26] <Dominus> back on the ppc branch,I get some thing odd in the linker stage
[12:35:11] <Dominus> http://pastebin.com/nhffPXBr
[12:35:35] <Dominus> but could be cause by myself...
[12:37:05] <Dominus> marzo, the change in the m4 file again messes the clang compile
[12:37:07] * Dominus sighs
[12:37:23] <Marzo> Can you post the result of clang -v?
[12:37:35] <Dominus> ../hash_utils.h:30:14: fatal error: 'unordered_map' file not found
[12:37:37] <Dominus> one moment
[12:38:07] <Dominus> Apple LLVM version 7.3.0 (clang-703.0.29)
[12:38:08] <Dominus> Target: x86_64-apple-darwin15.4.0
[12:38:09] <Dominus> Thread model: posix
[12:38:41] <Marzo> In m4/ax_cxx_compile_stdcxx_11.m4, find 'clang version' and change it to 'clang'
[12:38:47] <Marzo> Then try again
[12:41:51] <Dominus> success
[12:44:10] <Dominus> I think the other problem I had with PPC was my own doing, mix-match of SDKs that SDL and Exult were built with...
[12:44:22] <Dominus> but it's still building my full test...
[12:44:35] <Dominus> yes, success as well
[12:45:05] <Dominus> now 32bit build...
[12:46:19] <Dominus> 32bit built worked as well
[12:48:09] --> GitHub has joined #exult
[12:48:09] <GitHub> [exult] marzojr pushed 2 new commits to master: https://git.io/vVbDm
[12:48:09] <GitHub> exult/master b57fc1d Marzo Sette Torres Junior: * server/objserial.h: Fixing compilation on iOS. Hopefully does not cause...
[12:48:09] <GitHub> exult/master 429ab3e Marzo Sette Torres Junior: * m4/ax_cxx_compile_stdcxx_11.m4: Fixing cross-compilation (PPC).
[12:48:09] <-- GitHub has left #exult
[12:49:06] <Marzo> While there are still a lot of warnings, I think I will kill alpha and beos next
[12:49:15] <Marzo> Then handle the remaining warnings
[12:50:59] <Dominus> Marzo: about my earlier question today about ifdef MACOS and in there ifndef UNDER_CE - is http://pastebin.com/C6Cz6Hfb the correct way?
[12:51:11] <Dominus> that's from Audio.cc
[12:51:47] <Marzo> Note that it is a #elif in "#elif !defined(UNDER_CE)", not a #if
[12:52:04] <Marzo> #elif is an 'else if' in preprocessor-speak
[12:52:37] <Marzo> So delete the MacOS part, keep the UNDER_CE part, and change the #elif to a #if
[12:52:53] <Dominus> yes, when I wrote that I must have already edited the el away from elif and got confused
[12:53:21] <Marzo> While we ought to drop CE support too, I think we should drop one obsolete target at a time
[12:53:49] <Dominus> yes
[12:54:30] <Marzo> Hm
[12:54:38] <Marzo> CE seems to still exist
[12:54:56] <Marzo> The latest release was 2-3 years ago
[12:55:18] <Marzo> Now the question is if Exult even works on CE still
[12:57:26] <Dominus> Probably not, but we can keep it around some more...
[12:58:00] <Dominus> And while it may still exist, I think no one is using it anymore ;)
[13:01:05] <Marzo> Fair point
[13:12:19] <Dominus> Marzo: while last ce release is three years old, the last Ce based Windows mobile was in 2011
[13:25:34] --> GitHub has joined #exult
[13:25:34] <GitHub> [exult] marzojr pushed 1 new commit to master: https://git.io/vVbdn
[13:25:34] <GitHub> exult/master 260d016 Marzo Sette Torres Junior: * *.*: Getting rid of BeOS port.
[13:25:34] <-- GitHub has left #exult
[13:25:35] <Marzo> Dominus: got rid of BeOS port; left the docs to you
[13:36:34] <Marzo> Well, will get rid of alpha linux tonight
[13:36:48] <Marzo> Although I am unsure what to do about alpha_kludges.h
[14:00:28] <-- ttarrant has left IRC (Ping timeout: 252 seconds)
[14:01:33] --> ttarrant has joined #exult
[14:18:43] --> GitHub has joined #exult
[14:18:43] <GitHub> [exult] marzojr pushed 1 new commit to master: https://git.io/vVNkG
[14:18:43] <GitHub> exult/master cced24a Marzo Sette Torres Junior: * m4/ax_cxx_compile_stdcxx_11.m4: Making cross-compile work with clang.
[14:18:43] <-- GitHub has left #exult
[14:57:28] --> GitHub has joined #exult
[14:57:28] <GitHub> [exult] DominusExult pushed 1 new commit to master: https://git.io/vVNsM
[14:57:28] <GitHub> exult/master 1897508 DominusExult: Removed lingering MACOS defines. Fingolfin removed support seven years ago in 2b2abd1.
[14:57:28] <-- GitHub has left #exult
[16:02:58] <-- ttarrant has left IRC (Read error: Connection reset by peer)
[16:03:16] --> ttarrant has joined #exult
[16:04:40] --> Marzo_android has joined #exult
[16:10:15] <Marzo_android> ttarrant: by the way: if you set your SF email as an alternate on GitHub and verify, it will eventually recognize your contributions to Exult as being yours
[16:11:05] <ttarrant> Marzo_android, thanks
[16:16:17] <Marzo_android> I *think* you also need to verify it, but this should be a simple matter
[16:36:39] <-- ttarrant has left IRC (Ping timeout: 248 seconds)
[16:49:57] <-- Marzo_android has left IRC (Quit: Bye)
[17:14:26] --> frob has joined #exult
[17:20:53] --> Malignant_Manor has joined #exult
[17:22:05] <Malignant_Manor> Marzo: Dominus: should we get rid of resolution_increase and decrease functions/keybindings? They haven't worked since colourles changed the video options.
[17:43:44] --> GitHub has joined #exult
[17:43:44] <GitHub> [exult] malignantmanor pushed 1 new commit to master: https://git.io/vVNNc
[17:43:44] <GitHub> exult/master b7a87ae Malignant Manor: * keys.cc, keyactions.*, gumps/ShortcutBar_gump.cc : Add...
[17:43:44] <-- GitHub has left #exult
[17:49:39] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.92 [Firefox 45.0.2/20160407164938])
[18:04:45] <Epitrope> it's awesome to see so much activity :)
[18:45:42] --> GitHub has joined #exult
[18:45:42] <GitHub> [exult] marzojr pushed 1 new commit to master: https://git.io/vVALh
[18:45:42] <GitHub> exult/master fdd5aed Marzo Sette Torres Junior: Removing useless cast.
[18:45:42] <-- GitHub has left #exult
[19:43:54] <Dominus> Malignant_Manor: I completely forgot that we even had this function. Yes, get rid of it :)
[19:44:11] <Dominus> I'm starting tomorrow morning on merging the current changes into the iOS fork. Wish me luck :)
[20:59:37] <-- frob has left IRC (Ping timeout: 250 seconds)
[21:04:59] --> GitHub has joined #exult
[21:04:59] <GitHub> [exult] malignantmanor pushed 1 new commit to master: https://git.io/vVA6x
[21:04:59] <GitHub> exult/master 2b22df9 Malignant Manor: * defaultkeys.txt, exult.*, keyactions.*, keys.cc: Remove...
[21:04:59] <-- GitHub has left #exult
[21:05:23] --> Malignant_Manor has joined #exult
[21:07:14] <Malignant_Manor> Dominus: we really need to get Exult to include the needed iOS changes (and have the code be decent). You will need to update that fork after Marzo changes it again.
[21:09:58] <Dominus> Malignant_Manor: the problem is that the ios fork is also using a patched SDL 2.0.3
[21:10:44] <Malignant_Manor> Why does it needs the patch?
[21:11:56] <Dominus> some screwy things in SDL upstream, might be fixed or otherwise screwy im 2.0.4 - I'm no expert there...
[21:13:44] <Dominus> And the other thing is that xcode is a bit bad about generated data, so the fork has the flx files in revision control rather than having them autogenerated
[21:15:49] <Dominus> For the sdl stuff, let me look up the revision
[21:17:34] <Dominus> https://github.com/litchie/exult-ios/commit/f167ef113bba7ab06f1683b3961b5959b74582aa
[21:19:26] <Dominus> https://github.com/litchie/exult-ios/commit/77e1edeea1e70379d001e24ba86d7afc16ec2b2c
[21:21:23] <Dominus> https://github.com/litchie/exult-ios/commit/5084a227295a020d455b2247adaf5b03ba27a17b
[21:23:03] <Malignant_Manor> That looks like some terrible code. why if 0 and then latter add ifdef __IPHONEOS__ instead of ifndef __IPHONEOS__
[21:23:22] <Dominus> https://github.com/litchie/exult-ios/commit/29a32861977c851b71b632fcdf1399db55f18de3
[21:23:34] --> frob has joined #exult
[21:24:16] <Malignant_Manor> Of course, I don't know if Android has joystick support in SDL2
[21:28:59] <Dominus> I think joystick support only got better/saw a lot of fixes in sdl 2.0.4
[21:29:23] <Dominus> It probably interfered with stuff
[21:31:02] <Dominus> So, Litchie has done changes to SDL 2.0.3, perhaps to workaround irks in SDL or mane because he just doesn't know better
[21:31:40] <Malignant_Manor> You should probably build with the current repository and see what is broken.
[21:31:44] <Dominus> But it prevents easy changing to sdl 2.0.4
[21:32:12] <Malignant_Manor> Did he try and get upstream to do anything?
[21:32:15] <Dominus> Because he built up on his changes...
[21:32:28] <Dominus> I'm not sure, he cares much
[21:33:08] <Dominus> It seems he has a certain mindset that goes for easy hack instead of messing with upstream
[21:34:06] <Dominus> And right now he seems to be busy with other stuff
[22:04:19] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.92 [Firefox 45.0.2/20160407164938])
[23:03:36] --> GitHub has joined #exult
[23:03:36] <GitHub> [exult] Kirben pushed 1 new commit to master: https://git.io/vVxvV
[23:03:36] <GitHub> exult/master 3a21abf Kirben: Remove BeOS reference in Makefile.common.
[23:03:36] <-- GitHub has left #exult