#exult@irc.freenode.net logs for 3 Mar 2010 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[02:04:42] --> wepy has joined #exult
[02:04:51] <wepy> hey
[02:05:00] <wepy> svn is crashing a bit for me..
[02:46:15] <Marzo> wepy: please submit a bug report here: http://sourceforge.net/tracker/?group_id=2335&atid=102335
[02:52:19] --> Malignant_Manor has joined #exult
[02:57:30] <-- Marzo has left IRC (Read error: Connection reset by peer)
[02:57:59] --> Marzo has joined #exult
[02:58:53] <-- Marzo has left IRC (Client Quit)
[03:01:55] --> Marzo has joined #exult
[03:58:04] <-- Marzo has left IRC (Ping timeout: 276 seconds)
[04:03:35] <-- julien- has left IRC (Ping timeout: 246 seconds)
[04:04:01] --> julien has joined #exult
[05:51:48] <Sevalecpp> I use git <3
[05:51:57] <Sevalecpp> oh I just realised thats not what you meant
[05:51:57] <Sevalecpp> :P
[06:09:42] <Malignant_Manor> Are you typing in the wrong IRC Sevalecpp?
[06:10:08] <Sevalecpp> no, no I am not.
[06:10:13] <Sevalecpp> --> wepy (~wepy@ip70-179-127-73.dc.dc.cox.net) has joined #exult
[06:10:13] <Sevalecpp> <wepy> hey
[06:10:13] <Sevalecpp> <wepy> svn is crashing a bit for me..
[06:10:17] <Sevalecpp> then I realised
[06:10:20] <Sevalecpp> svn wasn't crashing
[06:10:28] <Sevalecpp> it was presumably the svn head version of exult
[06:10:29] <Sevalecpp> :P
[06:10:55] <Malignant_Manor> Ah, I had no idea what you were replying to.
[06:12:27] <Malignant_Manor> wepy are you going to post a bug report about it like Marzo asked stating what conditions you are crashing under?
[08:17:22] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
[11:39:43] --> Dominus has joined #exult
[11:39:43] --- ChanServ gives channel operator status to Dominus
[13:02:06] <-- Kirben has left IRC ()
[13:22:55] <-- Colourless has left IRC (Quit: casts improved invisibility)
[14:32:00] <-- ettin has left IRC (*.net *.split)
[14:36:37] --> ettin has joined #exult
[14:55:09] --> shazza has joined #exult
[15:26:08] --> andromada369 has joined #exult
[15:26:18] <andromada369> Hello
[15:28:42] <wjp> hi
[15:35:27] <Dominus> hi
[15:35:37] <andromada369> Just a quick question. Can I install Exult on a pendrive
[15:39:20] <wjp> for which operating system?
[16:06:49] <andromada369> windows
[16:08:26] <Dominus> yes you can
[16:08:36] <andromada369> alright thank you
[16:24:00] <-- shazza has left IRC ()
[16:32:43] <-- andromada369 has left #exult
[16:51:44] --> Malignant_Manor has joined #exult
[18:11:03] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
[18:18:06] <Dominus> I really give up now on building a static exult studio. I just can't make it happen. I even downloaded a gtk+ framework but this it's not happening as well
[19:02:16] --> Malignant_Manor has joined #exult
[19:25:32] <Dominus> strange lately I'm getting "Glob error −3" in the console. and the only google result for this is our log http://log.usecode.org/exultlog.php?log=2Oct2006
[19:34:53] --> Marzo has joined #exult
[19:35:22] <Malignant_Manor> Do you know what is causing it?
[19:36:06] <Dominus> nope, no idea
[19:36:23] <Malignant_Manor> Marzo, there is a translation thread that could use you input. http://exult.sourceforge.net/forum/read.php?f=1&i=332698&t=332698
[19:36:37] <Marzo> I am reading it right now
[19:42:52] <Malignant_Manor> U7ListFiles in listfiles.cc seems to be your problem. Case 3 should be no matches.
[19:44:55] <Malignant_Manor> Is it on loading a saved game?
[19:46:07] <Malignant_Manor> Newfile_gump.cc
[19:46:09] <Malignant_Manor> // Now read save game details
[19:46:10] <Malignant_Manor> char mask[256];
[19:46:13] <Malignant_Manor> snprintf(mask, 256, SAVENAME2, Game::get_game_type() == BLACK_GATE ? "bg" : "si");
[19:46:15] <Malignant_Manor> FileList filenames;
[19:46:18] <Malignant_Manor> U7ListFiles (mask, filenames);
[19:46:20] <Malignant_Manor> num_games = filenames.size();
[19:47:59] <Malignant_Manor> It's also in modmgr.cc. http://pastebin.com/4SkhW64J
[19:52:59] <Malignant_Manor> If the system is defined MACOS, it shouldn't use glob for u7ListFiles. Is MACOS not for MACOSX too?
[20:06:00] <Marzo> Malignant_Manor: posted an answer already
[20:06:12] <Marzo> (to the Phorum thingy)
[20:20:10] <Dominus> Malignant_Manor: it's not on reading the savegames, but I have an idea.. let me test :)
[20:22:43] <Dominus> it's giving me this when I have both blackgate and fov set up in the cfg
[20:29:02] <Dominus> Marzo: the cfg that exult makes when none is found could use some more defaults on OSX. It's missing all the <disk> entries, save for <save_compression_level>, can this be changed so it sets up the data and games stuff as well?
[20:29:52] <Marzo> It could, yes
[20:30:17] <Marzo> If you could enumerate which ones you would like to see written, I can go over it
[20:32:46] <Malignant_Manor> Marzo, I may try and separate the Keyring Dialog to a separate header for easy translating. Would you be upset if I tried this?
[20:33:05] <Dominus> <data_path><game><blackgate><path><waves> (for all four games)
[20:33:25] <Marzo> Dominus: when you have both BG and FoV in the cfg, it is likely that glob is failing to find any mods and is giving that error
[20:33:33] <Marzo> (for one of the games)
[20:34:02] <Marzo> It is probably not being able to read the 'mods' dir
[20:34:25] <Marzo> Malignant_Manor: go ahead
[20:34:58] <Dominus> when no exult.cfg is given and and therefore no path to any game, it gives the glob error −3, two times
[20:35:03] <Marzo> I would suggest using separate includes defining the strings and then referring them by name
[20:36:10] <Marzo> With no cfg, it is probably because it tries to find blackgate, forgeofvirtue, serpentisle and silverseed, finding maybe two in your case, and errors out as the dirs do not exist for the other two
[20:37:24] <Dominus> when I have the path set to bg and also have an EMPTY mods folder there it gives the error, with no mods folder no error :)
[20:37:39] <Dominus> (same for every game)
[20:37:51] <Marzo> The error it gives is that it found no mods
[20:38:09] <Dominus> ok, that explains it for me :)
[20:38:40] <Malignant_Manor> Can you please give an example so I know for sure what you are talking about?
[20:38:44] <Marzo> (the Windows version is even more annoying, but more explicit)
[20:39:13] <Dominus> Malignant_Manor: you mean the glob error?
[20:40:33] <Malignant_Manor> I was talking to Marzo. Sorry, Dominus.
[20:40:52] <Dominus> np :)
[20:42:06] <Marzo> Malignant_Manor: Sure: you can use lines like `string LAURIANNA_ASK_JOURNAL = "@Thou hast the journal in thy possession; wilt thou give it to me?@";` in a separate file and, in the adequate place, replace the text by LAURIANNA_ASK_JOURNAL
[20:42:10] <Marzo> Hm
[20:42:29] <Marzo> I guess I will do it myself; that way, I can write some scripts that do this automatically
[20:43:16] <Marzo> I also need to see if I ever added #include_once/#import to ucc
[20:44:01] <Malignant_Manor> That's what I was thinking about doing.
[20:44:42] <Malignant_Manor> I wouldn't know how to automate it though.
[20:45:27] <Dominus> where is it set where exult expects the data folder to be? I'd like to have the app bundle read the data from its bundle. In fact I'll probably have a feature request to write down. so exult first reads /data where it expects it to be and then also from the location given in the cfg :)
[20:45:37] <Marzo> In posix systems, with find/grep for dumping the text; then, I will generate names and use find/sed
[20:45:38] <Dominus> would be perfect for an appbundle :)
[20:46:33] <Marzo> Dominus: in exult.cc:564-593
[20:46:39] <Dominus> thanks
[20:47:35] <Marzo> EXULT_DATADIR is defined in fnames.h
[20:48:12] <Malignant_Manor> I still need to find what went wrong with my Debian unstable. I can't start x anymore.
[20:48:46] <Marzo> Malignant_Manor: "Debian *unstable*" <- there
[20:48:50] <Marzo> :-)
[20:49:26] <Malignant_Manor> Yeah, update broke a nice daemon and it wouldn't repair from reinstalling it.
[20:51:29] <Malignant_Manor> It think it was the only Squeeze available as an image at the time and I needed Squeeze to test a missing library issue that still exists and what turned out to be the poster forgetting to run autogen.sh. The poster denied failing to run autogen.sh though.
[21:01:31] <Malignant_Manor> Marzo can you ask Kirben not to add OpenGL to snapshots until after release?
[21:01:41] <Marzo> I already did
[21:01:58] <Marzo> He said that it is usually Colourless that does it, and he was in the room at the time too
[21:02:21] <Malignant_Manor> SVN snapshots not releases
[21:02:22] <Marzo> (not to mention that he is very "enthusiastic" about us not supporting OpenGL :-))
[21:02:29] <Marzo> Oh, snapshots
[21:02:33] <Marzo> Any reason why?
[21:03:45] <Malignant_Manor> I have two reproducible crashes. One of which doesn't crash in windib but does in Direct X.
[21:04:05] <Malignant_Manor> Also, fade to black doesn't work right for another bug.
[21:04:24] <Marzo> I tried to fix that latter one and failed miserably
[21:05:24] <Malignant_Manor> The one crashing Direct X and not windib is switching from point, scale2x, or interlaced in fullscreen to OpenGL.
[21:05:41] <Dominus> (I'm looking at exult.cc:564-593 and give up, that's all gibberish to me - not that important anyway :))
[21:05:44] <Malignant_Manor> It was something to do with changing to 8bit.
[21:06:01] <Dominus> oohh yeah, the 8bit scalers
[21:06:03] <Malignant_Manor> which made windib look horrible.
[21:06:09] <Dominus> my radeon hates them
[21:06:30] <Marzo> Dominus: It sets a default value for <DATA>, then tries to find exult.flx; it tries several other default values until giving up
[21:06:32] <Malignant_Manor> Have you tried my directx patch?
[21:06:49] <Dominus> my gf's nvidia chip on os x has no problems with the 8bit sclaers
[21:06:53] <Marzo> Nope, not yet
[21:07:29] <Marzo> Dominus: For OSX, another try could be made for the bundled data dir in Exult.app
[21:07:40] <Dominus> marzo, I can see THAT but I don't actually see how to hard code the /exult.app/contents/resources/data :)
[21:08:02] <Malignant_Manor> That was more at Dominus.
[21:08:50] <Malignant_Manor> I have the entire tracker feed filled at the moment so I'm not too worried about any particular patch.
[21:09:05] <Dominus> Malignant_Manor: not tried the directx patch as I don't have a windows development set up and for os x that patch possibly doesn't help :)
[21:09:23] <Malignant_Manor> I can send you a binary.
[21:09:45] <Marzo> Dominus: is that the actual path is is the / at the start not supposed to be there?
[21:13:47] <Marzo> Dominus: in any case, try this: replace lines 581-590 by this: http://pastebin.com/LPMj9ZrR
[21:14:03] <Marzo> (it should be clear where to try new paths until they work)
[21:14:04] <Dominus> I just tried and it doesn't matter whetehr the /is there or not
[21:14:07] <Malignant_Manor> Also, about OpenGL won't be in the release so problems with it may screw up testing.
[21:20:46] <Malignant_Manor> Dominus, cfg options are <directx> and <allow_8bitconv>. Both are yes or no.
[21:22:20] <Dominus> Marzo: without the / at the start is the correct one :)
[21:22:32] <Marzo> In any case, it works right?
[21:22:52] <Marzo> You can commit it yourself if so
[21:23:05] <Dominus> yes it seems it works correct like that
[21:23:09] <Marzo> (by "works" I mean Exult is able to find and use it)
[21:23:17] <Dominus> thanks, I'll commit it then
[21:23:54] <Dominus> yes, it finds it boht when in the bundle and also when exult binary is started in my trunk folder where /data resides
[21:24:10] <Dominus> this comes handy
[21:24:12] <Malignant_Manor> allow_8bitconv works with 2xSaI, bilinear, BilinearPlus, SuperEagle, Super2xSaI, Hq3x, and Hq2x. It basically allows their use if hardware depth is set to 8. It also has a hackish create a window first if fullscreen to stop a crash.
[21:24:54] <Dominus> Malignant_Manor: is the 8bit conv tied to the directx patch?
[21:25:00] <Marzo> Dominus: I decided to put it as the last check more-or-less for that reason
[21:26:50] <Malignant_Manor> allow_8bitconv is just an option you can try. It might help or not. I believe it only triggered on my computer when switching from point. scale2x, or interlaced to one of the other non-OpenGL scalers in fullscreen to fullscreen. Well, I'll be back later. I hope it helps.
[21:27:25] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6/20100115144158])
[21:41:34] <Dominus> ooh, maybe I should de-hardcode it from the bundle name - ../Reources/data
[21:44:15] <Dominus> nope, that doesn't seem to work
[21:50:03] <Marzo> Try ./ instead of ../
[21:50:32] <Marzo> (./Contents/Resources/data)
[21:51:07] <Dominus> nope doesn't work, Exult needs to know that it must look in the folder exult.app
[21:51:09] <Marzo> I don't know if it will work, but Apple "should" have make it work like that to prevent the hard-coding of names if it doesn't
[21:51:24] <Dominus> strange how this app bundle works
[21:51:58] <Marzo> There is probably an API call to get the bundle name; poking Fingolfin may yield an answer
[21:52:07] <Dominus> yeah :)
[21:52:27] <Dominus> maybe I need to poke inside scummvm to find that :)
[22:05:17] <Dominus> bah, couldn't find it
[22:05:24] <Dominus> maybe fingolfin looks in here sometime :)
[22:20:19] --> Malignant_Manor has joined #exult
[22:29:05] <Malignant_Manor> I think my allow_8bitconv hack was stupid. What it really did while I was testing was change the hwdepth.
[22:29:44] <Malignant_Manor> So, force_hwdepth_w and force_hardwaredepth_f would be the better oprions.
[22:31:28] <Dominus> if you can make another patch with that I'll give it a try with my osx machine tomorrow or so. if this works, maybe this could be the default
[22:33:31] <Malignant_Manor> Forcing a hwdepth probably shouldn't be default but I imagine most machines could handle 16 bit.
[22:34:01] <Malignant_Manor> Just seems like yet another thing for you to document.
[22:34:10] <Dominus> noooooo
[22:35:11] <Dominus> I really need to get working on this, since I'm leaving for the usa for a month on the 15th and at least the windows paths stuff needs to be properly documented when a release candidate drops
[22:44:56] <Malignant_Manor> I'm just not sure whether to use the force_hwdepth settings for all video options.
[22:52:26] <Dominus> I have no idea :)
[22:52:52] <Dominus> Uploaded a new exult OS X snapshot with the data path changes
[23:05:41] <Dominus> good night
[23:05:46] <-- Dominus has left IRC (Quit: Leaving.)
[23:39:19] <wepy> hm
[23:39:19] <wepy> bug report eh?
[23:39:19] <wepy> well, I was thinking: I have built exult with --enable-debug
[23:39:19] <wepy> and I can run gdb..
[23:39:28] <wepy> maybe I could help debug the crash next time it happens? I might need some advice
[23:46:58] <-- julien has left IRC (Ping timeout: 265 seconds)
[23:47:32] --> julien has joined #exult
[23:50:52] <Malignant_Manor> If you can debug it, great. If you can reproduce the crash, then post how to reproduce it.
[23:51:54] <Malignant_Manor> The only current crashing problems I know of are scaler related.