[01:09:42] <Marzo> Dominus: out of curiosity, would you say you always prefer to use clang to compile for OS X?
[01:12:51] --- DominusExult is now known as Dominus
[01:19:25] <Marzo> Dominus: ^
[02:52:13] <GitHub> [exult] marzojr pushed 2 new commits to master: https://git.io/vVx1i
[02:52:13] <GitHub> exult/master a6b7eb8 Marzo Sette Torres Junior: * *.*: Eliminating most -Wuseless-cast warnings.
[02:52:13] <GitHub> exult/master 9213bf9 Marzo Sette Torres Junior: * configure.ac, hash_utils.h: Fixing unordered_{map,set} detection. Moved...
[04:52:49] <Dominus> Marzo: Clang is the only compiler that Apple supports since three or four years now. So, yes, definitely prefer clang ;)
[04:54:29] <Dominus> The development stuff Apple provides is the easiest to use. And for other package managers on OS X you need the Apple stuff first anyway.
[04:55:58] <Dominus> The package managers do provide gcc but why bother when you can just use clang ;)
[04:57:33] <Dominus> For PPC I use the Apple flavoured gcc that still compiles ppc - 6 or 7 years old.
[06:03:37] <Dominus> Marzo: when I time out and reconnect around 1am (gmt) it's the 24h disconnect by my isp (or rather my router that does it in the night)
[06:28:59] <Dominus> Malignant, can you test the Windows snapshot? Because of that recent forum post. Exult runs fine for me...
[08:22:04] <Dominus> Marzo: oddly my 32bit build fails at finding SDL after your last commit. I'm looking through the config log now...
[08:27:11] <Dominus> the problem is that while I'm building against the latest SDK, I'm targeting older OS X versions (in this case the first intel one, 10.5) and your latest change adds -stdlib=libc++ which is only supported from 10.7 upwards
[08:28:15] <Dominus> If I target OS X 10.7, it works
[08:52:30] <Dominus> Malignant_Manor, can you split up your commits? For example https://github.com/exult/exult/commit/fb41264815b59b47204e6cafafe37b3e025f5fe6
[08:53:25] <Dominus> it is *easier* for me to apply patches when you split shortcut_bar stuff from other stuff. You can still push it together but it makes my life much easier if you keep that seperate
[11:18:54] <Marzo> <Dominus> the problem is that while I'm building against the latest SDK, I'm targeting older OS X versions (in this case the first intel one, 10.5) and your latest change adds -stdlib=libc++ which is only supported from 10.7 upwards
[11:19:07] <Marzo> Weird, I just moved that check to a different spot
[11:23:42] <Marzo> Hm
[11:24:00] <Marzo> I think I know how to fix it
[11:24:41] <Marzo> Or rather, why it worked before
[11:27:30] <Dominus> Great!
[11:27:54] <Dominus> I wish git's patching were a lot fuzzier ;(
[11:28:25] <Dominus> Will have to try out a script I found that uses both patch and git am
[11:28:46] <Dominus> Otherwise I'll have fun times ahead of me ;)
[11:31:26] <Marzo> Dominus: out of curiosity, does any of the compiles fail if you remove lines 214-241?
[11:32:06] <Marzo> (configure.ac, sorry)
[11:33:26] <Dominus> Marzo, not at home right now :(
[11:33:46] <Marzo> Well, the instructions will remain up there :-p
[11:44:27] <Marzo> Next up, removing zaurus support
[11:44:50] <Marzo> (only has GCC up to 2.96, in which Exult no longer compiles)
[11:45:34] <Marzo> Wait, I am forgetting alpha flag
[12:02:41] <Dominus> MorphOS is a contender for deletion as well. I think the last time anything hapened in that code was before the last full version...
[12:02:53] <Dominus> Spring cleanse...
[12:10:59] <Dominus> Ttarrant: about Marzo's comment yesterday to have github recognize your contributions: you will need to toggle privat/public email in github settings as well for a couple of hours to have it register for some reason
[12:11:22] <ttarrant> Dominus, you're making me do all this work ;)
[12:11:29] <Marzo> It has already registered for ttarrant
[12:11:46] <Dominus> Ah, great, was about to look
[12:12:03] <Dominus> Should have done it before speaking up ;)
[12:12:20] <Dominus> Now we need to force Kirben to do it ;)
[12:12:23] <Marzo> There he is: https://github.com/exult/exult/graphs/contributors
[12:13:19] <Marzo> It is funny, I added more lines, and removed slightly less lines, than wjp
[12:13:19] <Dominus> (Not to mention Jeff and Colourless)
[12:13:28] <Marzo> I wonder how it is possible
[12:13:41] <Marzo> They need github accounts for it to show
[12:14:08] <Dominus> I know ;)
[12:14:11] <Marzo> And then, they have to link the emails used for the commits to their github account
[12:14:47] <Dominus> Darke and maybe Simon Quinn as well, then we have covered almost everyone ;)
[12:15:24] <Marzo> I find it interesting how few people actually bothered to create github accounts and ask for access
[12:15:58] <Marzo> Most of the people added already had github accounts using the same emails, and github found them
[12:16:04] <Marzo> I just had to invite them
[12:16:31] <Dominus> Actually no one that didn't already have an account, I think...
[12:18:54] <Dominus> Funny how much it irks me that not everybody is accounted for in the contributers list
[12:19:19] <Dominus> Some weir sense of order/tidyness, probably...
[12:21:32] <Marzo> I feel the same way, actually
[12:46:53] <Dominus> Hmm, got to try that NES/8bit soundfont with Exult...
[12:56:29] <Marzo> Dominus: removing zaurus, will leave docs up to you
[13:11:46] <Marzo> Hm, now, where is GitHubBot?
[13:13:49] <GitHub> [exult] marzojr deleted shared-ptr at f8de5d5: https://git.io/vVjvn
[13:15:30] <GitHub> [exult] marzojr created shared-ptr (+2 new commits): https://git.io/vVjvC
[13:15:30] <GitHub> exult/shared-ptr caf169f Marzo Sette Torres Junior: Creating new branch to evaluate std::shared_ptr and std::weak_ptr in the context of Game_object and derived classes. The aim is...
[13:15:30] <GitHub> exult/shared-ptr f8de5d5 Marzo Sette Torres Junior: Initialized merge tracking via "svnmerge" with revisions "1-7095" from ...
[13:16:45] <GitHub> [exult] marzojr deleted shared-ptr at f8de5d5: https://git.io/vVjvn
[14:02:20] <Dominus> That's line 183-210 in configure.ac, now. Without those configure runs, copiling atm
[14:03:06] <Dominus> and compiles fine
[14:07:21] <Marzo> Ah, so I *was* right
[14:07:39] <Marzo> The new C++11 check also sets c++ preprocessor flags, which the old one did not
[14:08:04] <Marzo> So clang automatically uses libc++ or stdlibc++, as necessary
[14:08:15] <Marzo> You can remove the lines and commit then
[14:09:17] <Marzo> (I am off to work now)
[15:20:01] <GitHub> [exult] marzojr pushed 1 new commit to master: https://git.io/vVjav
[15:20:01] <GitHub> exult/master d092a75 Marzo Sette Torres Junior: * exult.cc, exultmenu.cc, usecode/compiler/uclex.ll: Eliminating all...
[16:38:18] <Malignant_Manor> Dominus: I can't reproduce the crash. I wonder if "teleport storm didn't work properly" means being able to go too far south in Exult.
[16:41:21] <Malignant_Manor> I tried 32 XP and 64 Win10 with digital music and gm music in each version. I played the intro in SS with and without the Keyring Mod. I had no and SB sfx on 64 bit. I had no and roland on 32.
[16:41:45] <Malignant_Manor> I could not test without SS.
[16:42:17] <Malignant_Manor> I have no idea why I missed removing that function Marzo.
[16:43:16] <Malignant_Manor> It would be nice to be able to run at least most warnings and run warnings as errors.
[17:27:50] <Dominus> Marzo: just deleting the configure.ac check for clang-on-Apple? Because it is no longer checked in the ac_cxx*.m4 files
[17:29:12] <Dominus> it seems to work fine without that check for all of my targets, though
[17:43:39] <Marzo> Dominus: yes
[17:49:27] <Marzo> Malignant_Manor: this is the reason why I have been removing warnings; the idea is to switch on -Werror at a high warning level by default
[17:53:06] <Marzo> Most warnings in audio code are in mt32emu; I will deal with those when updating to newer munt
[17:55:57] <Marzo> So the next steps are (in no order): (*) update mt32emu and get rid of any warnings in it; (*) update fmopl and get rid of any warnings in it; (*) make c++ libminizip; (*) get rid of all -Wsign-compare warnings
[17:56:47] <Marzo> Then, turn on -enable-paranoid-warnings by default, add -Werror to it, then start working on the bug queue
[18:18:08] <Dominus> quite a schedule :)
[18:41:42] <GitHub> [exult] DominusExult pushed 1 new commit to master: https://git.io/vwe8k
[18:41:42] <GitHub> exult/master f1bc656 DominusExult: The check for clang_on_apple was adding -stdlib=libc++ which is only available on OS X 10.7 and newer.
[18:41:58] <Dominus> arrrrghhhh
[18:42:14] <Dominus> I did chnge chaneglog, but I forgot to add it to the commit...
[18:42:24] <Dominus> Oh well, not a biggie in this case...
[18:44:48] <Dominus> now, should I struggle merging 30 commits into exult-iOS or remove SDL 1.3 cases from exult?
[21:58:26] <Dominus> Marzo, your changes do work on iOS
[22:02:44] <Dominus> apart from countless "Implicit conversion loses integer precision: 'unsigned long' to 'uint32' (aka 'unsigned int') warnings
[22:03:14] <Dominus> there is a "unused variable 'num_choices'" in sigame.cc
[22:05:20] <Dominus> in scale_xbr.h a couple of "Taking the absolute value of unsigned type 'unsigned int' has no effect" warnings
[22:06:34] <Dominus> in fmopl.cpp three "Shifting a negative signed value is undefined" warnings
[22:07:26] <Dominus> in Notebook_gump.cc "Unused variable 'usecode'"
[22:09:49] <Dominus> some unused variables 'num' x2 and 'npc_num' in cheat_screen.cc
[22:10:40] <Dominus> in bggame.cc also 'num_choices'
[22:12:08] <Dominus> in LowLevelMidiDriver. cpp unused variables 'sysex_max_data_size' 'rhythm_num_notes' 'patch_temp_size'
[22:14:34] <Dominus> usecode/ucinternal.cc:641:11: Absolute value function 'abs' given an argument of type 'long' but has parameter of type 'int' which may cause truncation of value - use std::abs instead?
[22:25:39] <Dominus> Marzo, a useful change to singletons (apart from some iOS stuff): https://github.com/eclectocrat/exult-ios/commit/8ce29884340db97bae2eee09db4d578d7218104c
[22:26:31] <Dominus> and this: https://github.com/eclectocrat/exult-ios/commit/764c396eecd6b0e46c0612ec1d7e29add3f54d2b ?
[23:27:31] <GitHub> [exult-web] DominusExult pushed 1 new commit to master: https://git.io/vwvGO
[23:27:31] <GitHub> exult-web/master 1523e01 DominusExult: Documentation updated with recent changes concerning Shortcutbar, BeOS and Zaurus
[23:29:23] <Marzo> Dominus: the second one definitely; the first one just moves the implementation around
[23:29:29] <Marzo> But it has its merits
[23:29:55] <Dominus> yeah, the location these are in right now seems odd when pointed at it like this :)
[23:45:31] <Dominus> good night, today was a good day, only nuked my local git repo once...
[23:59:16] <Marzo> Good night