#exult@irc.freenode.net logs for 8 Mar 2013 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[00:30:18] <-- wjp has left IRC (*.net *.split)
[00:39:11] --> wjp has joined #exult
[03:50:09] <Conf> Ow, I just noticed a small difference, I was going on a massacre just for fun. And I didn't get the guardians voice saying You had best not do that Avatar (voice seems to be working fine otherwise, intro is fine)
[03:51:42] <Conf> I tested it again by attacking the guard to the right, right at the start in black gate, guardian wanrning in dosbox, but not in exult :). I know it's minor but hey thought I'd mention differences I found
[03:56:45] <Conf> oh thats on black gate + forge of virtue with the 1.5.x (unstable) snapshot + digital sound pack installed + default midi + gm + speech on
[04:45:07] <sh4rm4> Conf, your report would be much more useful if you'd use the latest svn version
[05:05:29] <Conf> Ow, now I'm confused I'looking through the files on sourceforge but can only find up to 1.4.9rc1 is 1.5.x not higher?
[05:07:32] <-- Kirben has left IRC ()
[05:07:34] <sh4rm4> how old is that snapshot ?
[05:08:16] <sh4rm4> anyway: the best thing is to compile the newest version directly from the source control versioning system
[05:08:25] <sh4rm4> in other words, SVN
[05:10:36] <Conf> Ow that's above my head, but thanks for the information
[05:11:06] <sh4rm4> oh, actually the snapshot is from 25.2.2013
[05:11:19] <sh4rm4> so it even has the latest commit included
[05:11:40] <sh4rm4> so it's as recent as it can get. nevermind.
[05:11:51] <Conf> Ah good to know, my timing to play this again is good I guess
[06:12:09] --> Kirben has joined #exult
[06:12:10] --- ChanServ gives channel operator status to Kirben
[09:54:02] --> i30817 has joined #exult
[09:55:30] <i30817> Hey. If a hypothetical 'only ambient' music setting was written, that disabled combat music, what would be better: keep playing the previous music without interruption even if combat started; or 'play' silence?
[09:57:15] <i30817> What annoys me particularly is the interruptions, so i guess the first would be logical, but the second has it's justification of not just ignoring the situation. BTW, does U7 music have optimal 'endpoints' where it can be interrupted so it seems natural?
[10:17:08] <Dominus> i30817: If you go to the trouble of turning off combat music, then it would be logical to give a choice. Combat music, turn off previous song, keep on playing
[10:22:03] <Dominus> I don't think there are such break points defined.
[10:22:55] <Dominus> To start looking into it, I'd look through the changelog whether we implemented combat music once ;)
[11:33:55] <i30817> It's probably a bad idea to overload the 'looping' boolean for that...
[11:33:56] <i30817> So how about names:
[11:33:58] <i30817> 'Enabled', 'Ambient', 'Ambient no pause' , 'Disabled' ?
[11:33:59] <i30817> Obviously 'Ambient no pause' is a very bad thing to put in a button. Any suggestion?
[11:36:03] <i30817> Maybe it should refer to the function directly (ie: turn off combat music).
[11:36:27] <i30817> Music: No combat ?
[11:36:28] <i30817> Silly
[11:36:50] <i30817> Music: No combat no pause?
[11:36:51] <i30817> Sillier
[11:36:55] <wjp> what would those 4 options do exactly?
[11:38:31] <i30817> Enabled : all music (as usual)
[11:38:32] <i30817> Ambient: all music except combat music (and the runaway and victory music), pause when combat happens
[11:38:34] <i30817> Ambient no pause: as previous, but just keep playing
[11:38:35] <i30817> Disabled: no music (as usual)
[11:40:25] <wjp> how about two options? One to enable/disable music entirely, and one controlling combat music behaviour?
[11:41:04] <wjp> and then an option Combat Music: Combat (or Normal?) / Silence / Ambient
[11:41:21] <i30817> Makes sense, i was just being lazy by thinking of reusing the music button
[11:41:39] <i30817> I'm not sure there's space otherwise thou
[11:42:33] <i30817> That 'Speech options' title label with only one option underneath is a bit silly though.
[11:43:53] <i30817> Same for 'SFX options'
[11:57:59] <i30817> Oh wow, there's actually a reason for that (at least for 'SFX options').
[11:58:13] <i30817> The ingame menu is different for some reason.
[11:58:21] <i30817> Strange
[12:00:50] <i30817> Well how about leaving that one alone and modify the 'Speech options:' to 'Speech: Enabled' (like 'Music: Enabled') to free up a row and putting those options as part of the music subrows.
[12:01:25] <Dominus> Yes that should work.
[12:02:00] <Dominus> Sxf is different in game because you chose the sfx pack per game
[12:02:59] <Dominus> And this way it is easier to expand it for custom games (if there will ever be some ;))
[12:03:22] <Dominus> Marzo: how is the university mess?
[12:03:49] <i30817> Could do the same for SFX, as in 'SFX: enabled'. Then ingame it would have the pack subrow.
[12:04:29] <i30817> Would be more consistent if they all looked the same.
[12:04:39] <Marzo> Dominus: a mess
[12:05:22] <Marzo> The new system has consistently not enrolled people in all classes they requested for no reason; probably an off by 1 error by the "programmers" involved
[12:05:23] <Dominus> i30817: Yes, I was thinking this as well
[12:05:39] <Marzo> So there have been rather huge lines to fix up the enrollment
[12:05:50] <Dominus> Marzo: This sounds really really bad
[12:06:23] <Dominus> I'd hate to be the one having to handle those enraged students
[12:08:04] * Dominus is happily running a znc on his old mac mini ppc which will hopefully fix my constant time outs
[12:22:32] <-- Kirben has left IRC ()
[13:04:57] --> TheCycoONE has joined #exult
[13:21:06] <i30817> hey, if you disable 'digital music' on the exult menu, wouldn't it be best if that also hidden the 'midi driver' 'device type' and 'effects' buttons? I mean, they do nothing in that case right?
[13:25:05] <i30817> Actually, the other way around; enable it (with the music packs i guess), hide those other options.
[13:27:00] <i30817> What happens without the music packs with the 'digital music' button?
[13:46:57] <-- Marzo has left IRC (Read error: Connection reset by peer)
[13:47:03] --> Marzo1 has joined #exult
[13:47:03] --- Marzo1 is now known as Marzo
[15:41:58] --> Matt_O has joined #exult
[17:30:18] <i30817> SYN_SENT
[17:36:16] <i30817> So, can the 'digital music' button when enabled hide the 'midi driver', 'device type' and 'effects' buttons? Or is it more complicated that that and there are actually multiple 'digital musics' one for each device listed? (doubt it)
[17:50:42] <Dominus> There might be a reason why the midi stuff is not hidden when digital music is enabled
[17:51:16] <Dominus> If there is, malignant_manor or Colourless might know and speak up later
[17:51:21] <i30817> Probably button clicking not yet ready for that
[17:51:39] <Dominus> But in theory the midi stuff could be hidden
[17:52:25] <Colourless> digital music and midi are not mutually exclusive
[17:52:37] <i30817> No?
[17:53:10] <Colourless> exult is supposed to fallback to midi if it cant find the ogg for the selected track
[17:53:53] <i30817> git reset
[18:15:31] --> DominusExult has joined #exult
[18:15:31] --- ChanServ gives channel operator status to DominusExult
[18:16:41] <-- Dominus has left IRC (Ping timeout: 245 seconds)
[18:16:41] --- DominusExult is now known as Dominus
[18:56:02] <i30817> Do you mind me changing the order of the SFX block with the Speech block? Currently it looks a bit weird by default (with empty space instead of the 'pack' or 'conversion' buttons).
[18:57:10] <Dominus> actually, yes, the sfx should be before the speech, IMO
[19:01:06] <i30817> As long as i don't rearrange the enum i guess it's alright (i did insert the new button enum after 'music_looping' because the button has to be deleted when music is disabled, but that seems ok)
[19:01:55] <i30817> Now it's just writing the real feature!
[19:02:00] <i30817> And saving
[19:07:21] <i30817> BTW, these gump files would probably be much easier to work if the enum could be used a as a index to the y-row location.
[19:08:17] <i30817> Currently if you change one location you have to change the subsequent ones.
[19:09:47] <i30817> IDK if it's assured in C++ that enums are monotically increasing from 0... but the FORGET_OBJECT macro uses sure seem to think so.
[19:19:09] <-- sh4rm4 has left IRC (Ping timeout: 276 seconds)
[19:25:24] --> sh4rm4 has joined #exult
[19:26:37] <i30817> BTW, are you really going to kill the midi driver 'UnixSeqDevice'? (that one that crashed ubuntu because the cmd line tool it used was not installed).
[19:27:27] <i30817> Because then you can get rid of the weird button in the middle of this gump (the one that is misaligned and then overflows when it gets to UnixSeqDevice )
[19:30:58] <i30817> The 'device type' bellow remains misaligned thou (because of 'Fake MT32')
[19:31:44] <i30817> Going to leave it alone in this patch.
[19:32:15] <i30817> Both of them.
[20:07:09] <i30817> What's the accepted way to define a ternary result in c++? A int + 3 constants ?
[20:10:56] <i30817> enum probably
[20:11:01] <i30817> yes.
[20:11:35] --> exg has joined #exult
[20:12:58] <exg> hi, is the latest snapshot preferred over both 1.2 and 1.4.9rc1 ?
[20:36:15] <Dominus> exg: yes
[20:36:34] <Dominus> 1.2 is ancient and I wouldn't touch it if I had to :)
[20:36:59] <Dominus> and 1.4.9rc1 has some nasty bugs that are mostly fixed in the snapshot.
[20:37:44] <exg> Dominus: ok, thanks :)
[20:38:10] <Dominus> the snapshot only has a few regressions, one bad AFAWK, but Marzo will probably fix it as soon as his university problem is sorted out (which sounds bad and may need some time)
[20:38:38] <Dominus> i30817: sorry can't be much more help on the coding side.
[20:38:57] <Dominus> I'm not sure whether we kill unixseqdevice
[20:39:37] <Dominus> and I don't know how weird it looks, since I'd need a system with that :)
[20:39:53] <Dominus> maybe you can post a bug report with a screenshot of that
[20:40:52] <i30817> Then, to anyone who knows; how can you cast a enum to a int safely?
[20:41:06] <i30817> maybe i should google it
[20:41:32] <Dominus> bbl - need to do the dishes :)
[21:07:01] <i30817> The audio gump is a bit strange - _some_ options are saved to file there, but loaded from file in Audio.cc
[21:07:03] <Shozan> i30817, i believe enums in c++ are ints already and can be cast as such without much trouble
[21:07:31] <i30817> thanks
[21:32:10] <i30817> Annoying thing about the windows in exult is that when i press alt+f4 or press the window manager close button, it asks me if i want to quit.... for each window, instead of closing
[21:33:00] <Dominus> huh?
[21:33:02] <i30817> Maybe thats ok if i'm ingame and want to go to the main menu, but if i'm actually on the setup window.
[21:34:08] <Dominus> what are you talking about?
[21:34:16] <i30817> well, i just find it strange
[21:34:17] <i30817> 'are you sure you want to quit' and then only the 'main window' does quit (without asking)
[21:35:01] <Dominus> when do you see the "are you sure you want to quit"?
[21:35:41] <i30817> All the others 'ask' if i want to quit, when the behaviour should logically be the opposite (main window asks, others just quit, with the possible exception of the main game window, while ingame)
[21:36:01] <i30817> any gump on the setup menu for instance.
[21:36:05] <Dominus> seriously, I don't understand
[21:36:15] <Dominus> I don't ever see that
[21:36:16] <i30817> alt+f4
[21:37:30] <i30817> the 'problem' is mostly that it doesn't quit, but just goes to the previous window
[21:37:46] <Dominus> then don't use alt+f4 :)
[21:37:48] <i30817> until it gets to the main exult window. Then it quits
[21:37:55] <i30817> (without asking lol)
[21:38:12] <Dominus> I understand what you mean now but can't see it because that is not present in OS X :)
[21:41:29] <Dominus> i30817: also doesn't happen on Windows
[21:42:43] <i30817> damn flaky window managers
[21:43:19] <i30817> well it's a bit bad (the crash with unixseqdevice on ubuntu is worse thou)
[21:46:49] <Dominus> again, it seems to be pretty limited to you :)
[21:46:55] <i30817> Asking about the audio api. There is a Audio::stop_music() method. Very well, but what if i want to 'pause' the music, can that be done easily? Or do i need to restart the playing music after? (to implement the 'silence').
[21:47:11] <i30817> Dominus: no way. Did you ask someone?
[21:47:25] <Dominus> he he. no.
[21:48:44] <-- TheCycoONE has left IRC (Quit: And then there were n-1)
[21:49:35] <i30817> There is a pause_audio() but no 'pause_music()' in particular. I wonder if it can be implemented easily.
[21:49:37] <Dominus> and i30817 it also does this when you click on the checkmark in the menu gump?
[21:50:12] <Dominus> wjp, marzo, does Exult behave that way for you? Asking whether you want to quit while you are in the menu and pressed alt+f4?
[21:51:14] <i30817> No it doesn't (assume you're asking 'do you really want to quit' popup)
[21:51:45] <Dominus> ah, I misunderstood the "windows manager close button" thing :)
[21:52:14] <Dominus> and yes, that does it too for me
[21:52:26] <i30817> It's just not the menus it's any gump or subgump inside the menu. It's strange it asks, but doesn't actually quit, but just goes to the parent gump
[21:52:54] <Dominus> yeah, it shoudl probably quit there
[21:53:13] <i30817> I'd expect alt+f4 to ask once and then quit the app. Maybe go to the initial menu.
[21:53:43] <i30817> And asking if you're not ingame is a bit of a waste too.
[21:53:55] <Dominus> yeah, please file a bug report. it will be forgotten otherwise
[21:54:09] <i30817> soon
[21:54:15] <Dominus> :)
[22:11:29] <i30817> Mmm it's not UnixSeqDevice driver that had that problem. It's forked_player.
[22:12:02] <i30817> And it looks to have already been disabled. Figures
[22:12:39] --> Kirben has joined #exult
[22:12:39] --- ChanServ gives channel operator status to Kirben
[22:13:31] <Dominus> :)
[22:14:28] <i30817> Looks like it might be possible to pause the music instead of restarting it
[22:14:45] <i30817> At least there is such a function on the mixer object
[22:50:44] <i30817> To someone who *knows*
[22:50:45] <i30817> Audio::start_music_combat sends a music into the midi driver to be played eventually, but has no return
[22:50:47] <i30817> Audio::pause_audio sets 'all' channels to 'paused'. I presume 'channel' also applies to sound effects and not only music. If i'm wrong warn me.
[22:50:48] <i30817> I suspect i have to 'record' the current channel of a music somewhere, but none of the start_music or start_music_combat returns on which channel the music is playing
[22:52:46] <i30817> And it looks that the root of this in in the drivers, which don't return the channel id, on Mixer::MyMidiPlayer::start_music
[22:53:18] <i30817> even though they're doing 'midi_driver->startSequence(SEQ_NUM_MUSIC, eventlist, repeat, 255);' in that function
[22:54:33] <i30817> Logically it makes sense that some drivers can't be paused, like the forked_player one (which was disabled)
[22:59:56] <i30817> Fuck it, it's going to start the music again.
[23:00:19] <i30817> No pause
[23:11:33] <-- Marzo has left IRC (Ping timeout: 276 seconds)