[02:59:49] <-- Dominus has left IRC (Ping timeout: 246 seconds)
[03:00:21] --> Dominus has joined #exult
[03:00:21] <-- Dominus has left IRC (Changing host)
[03:00:21] --> Dominus has joined #exult
[03:00:21] --- ChanServ gives channel operator status to Dominus
[04:28:03] --> azeem_ has joined #exult
[04:31:22] <-- azeem has left IRC (Ping timeout: 240 seconds)
[08:32:22] <-- Lightkey has left IRC (Ping timeout: 240 seconds)
[08:33:02] <-- Sevalecan has left IRC (Ping timeout: 244 seconds)
[08:33:34] --> Sevalecan has joined #exult
[08:45:21] --> Lightkey has joined #exult
[10:08:55] <Dominus> aaarrggghhh, getting a working macports install for Exult Studio takes ages...
[10:10:32] <Dominus> Marzo: how does travis set the OS X VM up? Does it keep it around for each project or does it set up a new VM each time a test is run? I guess the former because.... it takes ages to get the dependencies up...
[10:11:33] <Marzo> They set up clean VMs; when you run the test, it fetches dependencies, runs the test, then discards the VM state
[10:11:53] <Dominus> wow...
[10:12:03] <Marzo> It does have dependency caching, so that the dependencies install faster, but they don't explain how it works
[10:14:47] <Dominus> I've "patched" the osx script for travis a tiny bit https://www.dropbox.com/s/0ioot7p66ywnobu/travispatch.diff?dl=0
[10:15:37] <Dominus> just getting newest macports and the port install can be shorter as most is pulled in by install libsdl and all should work in one line
[10:16:01] <Dominus> I didn't want to commit as I thought you'd like to test it out first
[10:16:29] <Dominus> yeah and forgot the "| cat"
[10:16:35] <Dominus> in the last line
[10:17:45] <Marzo> The way I "test" is by sending it up and hoping it works :-)
[10:19:04] <Marzo> (I don't know if there is a way to test other than this, and I didn't see it in their documentation)
[10:19:15] <Dominus> ha he
[10:20:10] <Marzo> The good thing is that it really allows us to narrow down Exult's dependencies
[10:20:34] <Marzo> In a way we can't on our own computers
[10:20:53] <Marzo> Because we already have everything installe
[10:20:55] <Marzo> d
[10:21:17] <Dominus> true... reminds me that I need to update the README:MacOSX
[10:23:10] <Marzo> By the way, I have been meaning to ask: maybe we should have a "readme.md" file in our repositories with a short version of the full readme and with links to the documentation page
[10:23:44] <Marzo> This would also give us a place to put the Travis and Coverity badges
[10:23:57] <Dominus> yes, our full readme is a bit much on the github page
[10:29:23] <Dominus> Seriously, my macports installation is building dependencies for about an hour now
[10:29:23] <Dominus> libglade!!!
[10:36:49] <Dominus> marzo, you can scratch libpng from the port install line as well - it is pulled in by libglade2 (because that needs fontconfig, which needs freetype, which needs libpng)
[10:38:29] <-- Marzo has left IRC (*.net *.split)
[10:38:59] --> ParuNexus has joined #exult
[10:39:39] --> Marzo has joined #exult
[10:39:58] <Marzo> Net split
[10:40:04] <Marzo> Let me see what I lost
[10:40:30] <-- ParuCodex has left IRC (Ping timeout: 250 seconds)
[10:41:37] <Marzo> Not much
[10:42:08] <Marzo> Hm, maybe I should go about checking exactly what is needed in Ubuntu too
[10:42:34] <Marzo> I could set up a local VM myself to test
[10:44:47] <Dominus> yes, our "INSTALL" file is hopelessly outdated - I didn't touch it much because that is more for the Unix side of things..
[10:45:25] <Marzo> I will also see about making a more modern version for MinGW in Windows
[10:46:00] <Marzo> There are newer versions of GCC, and it seems that MinGW will be updating to GCC 5.3 (or maybe even newer)
[10:46:26] <Dominus> I think the mingw side is still maintained by Kirben to keep it compatible with Windows 9
[10:46:29] <Marzo> And at some point, we may consider updating the Visual Studio projects and finally flip the switch to C++11
[10:46:44] <Dominus> Windows 9x
[10:46:54] <Marzo> It is, yes; GCC itself still allows building for it
[10:47:08] <Marzo> The major issue is GTK+ stuff, not GCC
[10:47:38] <Marzo> There is a really old version of GTK+ which works for Win9x, and all newer ones drop support for it
[10:48:22] <Marzo> At some point we may consider dropping Win9x support, but we should have at least one more release for it
[10:48:22] <Dominus> ah, yes, I thought there was some other issues with those older libs and newer gcc but that should hold true for gcc 4.6x, too
[10:48:35] * Dominus coughs
[10:48:56] <Marzo> Well, GTK+ is C
[10:49:00] <Dominus> let's see about getting one release done at some point :)
[10:49:05] <Marzo> The C ABI has been stable for a long time
[10:49:44] <Marzo> So we shouldn't see any problems by that route
[10:50:17] <Dominus> Wasn't there another issue "lately" that made Kirben switch back from newer gcc to this one?
[10:50:32] <Marzo> (by the way, one of the goals of the MinGW folks is to ensure that the new GCC still supports Win9x)
[10:50:47] <Dominus> and yeah, my libglade2 prefix is finally done after 80 minutes)
[10:50:57] <Marzo> You would have to ask him; this is news to me
[10:52:08] <Marzo> Hm: https://sourceforge.net/projects/mingw-w64/files/Toolchains%20targetting%20Win32/Personal%20Builds/mingw-builds/
[10:52:22] <Marzo> GCC 6.2.0 O>o
[10:55:31] <Dominus> ah, in the Readme.mingw the note that the developer packages were done with gcc 4.6.2 and might break with newer gcc
[10:57:14] <Marzo> Seems like GCC 5.3.0 is now the default with mingw-get
[10:57:27] <Marzo> Well, I can update that
[10:57:51] <Marzo> Still have a XP VM
[11:00:55] <Dominus> Even though mingw is great for what it does - it's another reason I'm glad I left Windows behind - it proved to be easy to mess up...
[11:02:54] <Marzo> The issue is that we want to give an easy-to-use pre-made developer package
[11:03:27] <Marzo> Instead of either telling how to make one or where to download the required versions
[11:04:46] <Dominus> yes, for lock stock Exult it would be doable, SDL, ogg... But the Exult Studio variant takes too many hoops to jump through to make it work from scratch
[11:15:35] <Dominus> ok, something is not right with my Exult built with x11
[11:15:51] <Dominus> it won't start...
[11:15:53] <Dominus> Couldn't create window: No OpenGL support in video driver
[11:15:53] <Dominus> Couldn't create renderer: Invalid window
[11:15:53] <Dominus> Couldn't create texture: Invalid renderer
[11:18:08] <Dominus> ah...
[11:18:11] <Dominus> SDL2
[11:19:00] <azeem_> Marzo: in principle you could look at the Build-Depends for the Debian/Ubuntu exult package
[11:19:05] <azeem_> but I think that one is tracking 1.2
[11:23:51] <Dominus> bah, on OS X it is no fun to use Exult with x11. You need SDL1.2x *and* that is slow on newer OS X. With x11 it gets downright jerky...
[11:24:09] <Dominus> *but* we have no other form of communication than x11 :(
[11:25:48] <azeem_> maybe switching to SDL2.0/OpenGL would make things better, but I guess that's a rather large project
[11:26:09] <Dominus> azeem_: we *do* have switched to SDL2 :
[11:26:11] <Dominus> :)
[11:27:02] <Dominus> *but* it seems that either SDL2+opengl don't work with OS X' X11 *or* Exult needs a better handling of this case
[11:27:40] <Dominus> on OS X it would be better if there were a better form of communication between Exult and Exult Studio than X11
[11:28:12] <Dominus> You can build both without X11 but then there is no communication between them
[11:29:05] <azeem_> I assumed you could bypass X11 if you use SDL2 and OpenGL, sorry
[11:29:25] <azeem_> ah, I missed the part about Exult and Studio communicating
[11:30:20] <azeem_> Dominus: so regular end-users who just care about playing Exult on Macos will typically get a no-X11 install and developers the X11 package of both Exult and Studio?
[11:39:33] <Dominus> Azeem_: normal users can download the prebuilt snapshot without x11 (but still SDL 1.2x because of some strange corruptin in 1x scaling)
[11:40:29] <Dominus> People who want Exult Studio will need to build it themselves because bundling Studio and Exult is beyond my abilities
[11:40:47] <Dominus> The dependencies are hellish for bundling
[11:44:33] <azeem_> ack
[13:14:29] <Dominus> odd. Exult with SDL2 +x11 does not run here... at all
[13:30:15] <Dominus> in imagewin.cc bool Image_window::create_scale_surfaces if (screen_window == NULL) triggers...
[16:31:14] --> i30817 has joined #exult
[17:24:04] <i30817> my exult ppa is building again.
[17:32:53] <azeem_> btw, if there was a stable release, I'd upload it for Debian and Ubuntu
[17:54:47] <i30817> i really have no idea if there was a stable release or not. With the speed the sifixes and bg keyring mods move at and usecode fixes all over the place i prefer using the bleeding edge. It's why i made the ppa after all.
[17:56:13] <i30817> Though i think it's a bit sad that there are still rendering errors, or the music not behaving the same or the schedules still not being perfect after all these years, that is vastly overtaken by usecode fixes and the keyring mod
[17:57:49] <i30817> which is why i really like that we can build the mods at the same time as exult, so that the ppa installs updates for both.
[17:59:25] <i30817> i wonder if there is a solution to the 'mountains on the ocean' bug on higher resolutions during the SI intro on the boat. It gets fixed immediately after.
[18:00:41] <i30817> or rather it gets fixed as soon as the boat teleports in, now that i rechecked.
[18:02:15] <i30817> since it's the first thing a user sees if they increased resolution, seems worthwhile
[18:03:18] <-- i30817 has left IRC (Quit: ChatZilla 0.9.93 [Firefox 50.0/20161114144739])
[18:59:05] --> i30817 has joined #exult
[19:01:56] <i30817> when you have :
[19:01:58] <i30817> --enable-exult-studio \
[19:01:59] <i30817> --enable-exult-studio-support \
[19:02:01] <i30817> on the configure, you're supposed to be able to call exult-studio from ingame during exult yeah? Only, i checked it now, and apparently the ctrl+alt+M shortcut to open it is gated by the cheats being on.
[19:02:21] <i30817> i thought it was not working but just now enabled cheats to see
[19:03:30] <i30817> Oh well, ppa is 100% successful, even runs exult studio, how nice.
[19:05:45] <i30817> check your previous discussion, my dependencies for a ubuntu build on the launchpad build bot are currently:
[19:05:47] <i30817> "debhelper (>= 9), dh-autoreconf, autotools-dev, autoconf-archive, po-debconf, libtool, libsdl2-dev, byacc, flex, bison, libgimp2.0-dev, libglade2-dev, libfluidsynth-dev, libogg-dev, libvorbis-dev, libmt32emu-dev"
[19:05:48] <i30817> Very probably not minimal (even considering i added autoreconf and libmt32emu-dev)
[19:07:12] <-- Rottingbeef has left IRC ()
[19:07:22] <i30817> i also have runtime deps for libfluidsynth1, rsync, mt32romdata. rsync is just for the install time script copying the games and mt32 roms to system dirs, mt32romdata is a dependency to reuse the mt32 rom directory for dosbox too and only uninstall them when both are uninstalled.
[19:08:36] <i30817> libfluidsynth1 is for a optional dependency for fluid-soundfont-gm (freepats is also optional but the required timidity files are already included on the package install).
[19:09:57] --> Rottingbeef has joined #exult
[19:10:01] <i30817> those are sound fonts, so huge and 'suggested', though exult-audiopack is 'Recommends'. I don't remember the difference.
[19:15:33] <i30817> Oh btw, the egg at the start of SI, probably for the thunderbolt because that's its symbol still shows even after disabling map mode after enabling it a few times and then disabling it (i don't remember if i had cheat mode on or not but it's not supposed to make a difference right?)
[19:23:04] <i30817> many of those deps above are just for debian style packaging building (debhelper (>= 9), dh-autoreconf, autotools-dev, po-debconf, autoconf-archive). autoconf-archive was needed because of a missing macro problem in exult source on the launchpad build bot for some reason, po-debconf is for translation of packaging script files (not that i translated anything), make is pulled by one of these.
[19:23:57] <-- i30817 has left IRC (Quit: ChatZilla 0.9.93 [Firefox 50.0/20161114144739])
[20:09:34] <Dominus> So here I thought I'm the only one ranting in here...
[20:10:59] <Dominus> For the record, if i30817 reads this, the starting scene was done intentionally like that in the original to prevent people storing their stuff in the boat before the storms take them off you
[20:11:49] <Dominus> There is nothing to be done about this and that's why it's not recommended to play at higher resolutions
[20:12:50] <Dominus> And working on mods and usecode is just that much more fun and rewarding than going through the schedules and do them properly...
[20:13:19] <Dominus> First rule of opensource...etc
[20:43:41] <azeem_> some schedules are being fixed though, AFAICT
[20:44:05] <azeem_> I had a look through the merged PRs once I realized exult went github
[21:25:56] <Dominus> They could still use a lot of work... besides the missing motions there is hardly any sfx coded in while there should be...
[21:35:01] <Marzo> <i30817> "debhelper (>= 9), dh-autoreconf, autotools-dev, autoconf-archive, po-debconf, libtool, libsdl2-dev, byacc, flex, bison, libgimp2.0-dev, libglade2-dev, libfluidsynth-dev, libogg-dev, libvorbis-dev, libmt32emu-dev" < byacc = Berkeley's yacc, a open source version of yacc. Bison is an open source and improved version of yacc. There is no need for both, and ucc most likely fails to build on byacc because it uses Bison extensions.
[21:53:07] --- azeem_ is now known as azeem
[22:24:23] --> KnightCaptain has joined #exult
[22:26:35] <KnightCaptain> Fixing the opening scene mountains in SI would require map edits, Usecode changes, or both. Since Exult uses the original game files, I don't believe there is a way to do that without including a mod of some sort as the default.
[22:26:50] <KnightCaptain> ^ @i30817
[22:27:27] <KnightCaptain> I'm hacking away at the next round of SIfixes updates, the same one I did a pull request for and then cancelled.
[22:28:05] <KnightCaptain> Is there any update on the improved UCXT program? I'd love to fix more things without having to re-do an NPC's entire conversation tree.
[22:35:29] <-- Rottingbeef has left IRC (Read error: Connection reset by peer)
[22:47:05] <Dominus> Marzo has been very busy fixing our code to be less error prone to "mysterious" bugs hidden in our code
[22:47:26] <Dominus> Which reminds me to look up how much is left now...
[22:48:27] <Marzo> A bit more than half
[22:49:07] <Marzo> I think a lot of it are false positives
[22:49:34] <Marzo> After fixing everything I will double-check the false positives and maybe find a way to fix them without just ignoring them
[22:49:47] <Marzo> Because maybe a compiler will also misinterpret the code
[22:56:05] <Marzo> Also, I will be reporting a lot more stuff to munt for mt32emu fixes