#exult@irc.freenode.net logs for 28 Dec 2012 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage

[00:00:22] <Malignant_Manor> We also have the range combat pathfinding issue. I noticed earlier that cyclopes in BG could hit each other.
[00:01:28] <Malignant_Manor> Too many pathfinding issues.
[00:06:11] <Malignant_Manor> Maybe cyclopes hitting each other will by fixed by the blocking fix.
[00:07:12] <Malignant_Manor> The cyclopes bug isn't the easiest to reproduce.
[00:18:20] <Marzo> Malignant_Manor: fix away
[00:18:33] <Marzo> Back to try to figure out attacking through walls
[00:22:44] --> Kirben has joined #exult
[00:22:44] --- ChanServ gives channel operator status to Kirben
[00:29:00] <rrva> Marzo: my mouse event recording seems not reliable :(
[00:29:08] <rrva> Marzo: Do you still want a bug report?
[00:29:28] <Marzo> It would be useful to prevent it from being forgotten
[00:33:14] <rrva> are containers, the inventory etc all called gumps?
[00:38:48] <Malignant_Manor> yes, popups are gumps
[00:43:58] <Malignant_Manor> Marzo: The cyclops problem seems alignment related since the eggs mix hostile and random. With them all the same alignment, I haven't reproduced it. I need to check the original. Too bad I don't have the improved teleport.
[00:46:59] <-- sh4rm4 has left IRC (Remote host closed the connection)
[00:48:37] --> sh4rm4 has joined #exult
[00:49:11] <Malignant_Manor> Marzo: The cyclopes hit each other in the original game engine too.
[00:49:23] <Marzo> So no problem then :-)
[00:49:53] <Malignant_Manor> Yeah, I wonder if it was intentional.
[00:50:30] <Marzo> There are several places in the original that have things like this
[00:50:31] <Malignant_Manor> It doesn't really make sense to be on purpose though.
[00:50:44] <Marzo> Usually, its with different types of creatures
[00:51:15] <Marzo> Maybe they were trying to show rival factions of cyclops? Who knows
[00:52:42] <sh4rm4> hmm i cannot reproduce any crashes dragging stuff around
[00:52:54] <sh4rm4> rvva: mind to share a savegame ?
[00:54:24] <Marzo> OK, managed to fix the attack through walls bug; now I have to fix the squeeze between door and wall bug that has been made apparent by it
[00:54:42] <rrva> sh4rm4: sure
[00:55:22] <sh4rm4> i use black gate, and drag around stuff inside my backpack. is that ok ?
[00:55:44] <rrva> open inventory and drag that window around
[00:55:59] <rrva> pretty fast and far
[00:57:12] <sh4rm4> do i need to release ?
[00:59:13] <rrva> yes i think so
[00:59:24] <rrva> that was in my xmacro recording
[00:59:35] <rrva> I have added savegames to https://sourceforge.net/tracker/index.php?func=detail&aid=3598675&group_id=2335&atid=102335#
[01:03:14] <Marzo> Damn... its Actor::move_aside that is causing the party members to walk through walls
[01:04:15] <rrva> sh4rm4: please look at xmacro.log, its actually human-readable
[01:04:35] <rrva> sh4rm4: the problem with that log is that is resolution-dependent with absolute coordinates
[01:04:59] <rrva> sh4rm4: even for me it stopped reproducing the bug when the exult window was placed a bit different
[01:05:22] <rrva> sh4rm4: but during 3 consecutive runs of that file it crashed exult
[01:05:55] <rrva> and i think serpent isle paperdolls should be off
[01:06:05] <rrva> for xmacro.log to work
[01:09:36] <rrva> well, thanks so far for looking into it. I will monitor the sourceforge bug report and maybe try some more debugging tomorrow
[01:12:57] <sh4rm4> thanks
[01:13:14] <sh4rm4> you should idle here until the issue is solved
[01:13:40] <sh4rm4> if i cant reproduce, i'll try to download ubuntu 13.04 beta
[01:14:06] <rrva> the only thing weird in the xmacro logs is button3 and button1 being pressed, interleaved
[01:16:21] <rrva> first i bring up the inventory, with "i", then I guess its middle mouse button (I have a laptop touchpad), and then before button3 is released button1 is pressed, then I moved the cursor fast
[01:16:51] <rrva> will idle
[01:19:14] <Malignant_Manor> I repeat that crash with those instructions.There is no need to move the cursor.
[01:20:09] <rrva> Malignant_Manor: you can repeat it? Great :) :)
[01:28:10] <Marzo> That is easy to reproduce: hold right, press left, let go of right, let go of left
[01:28:52] <Malignant_Manor> It deletes dragging.
[01:29:13] <rrva> ah, I was not so observant when I was doing the testing on what I was doing. Its the touchpad which makes it weird also, it has no proper buttons
[01:39:00] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.89 [Firefox 17.0.1/20121128204232])
[01:43:19] <Marzo> Fixed and committed
[01:44:11] <Marzo> (far easier than trying to figure out why move_aside's can chain up and cause an actor to walk through walls)
[01:44:23] <rrva> Marzo: Wow! Thanks :) :)
[01:44:53] <rrva> exult is really impressive stuff, thanks again
[01:56:26] <rrva> Marzo: hrm, now all windows I try to drag disappear instantly. No crashes, but no dragging either
[01:56:49] <Marzo> OOps
[01:58:01] <Marzo> I am doing a full recompile; when it is done, I will see what I did wrong
[01:58:09] <Marzo> (but I think I already know)
[02:04:19] <Marzo> Fixed and committed; the mistake I just made proves that I should go to bed and sleep, so good night
[02:06:21] <rrva> good night
[06:50:45] --> DominusMobile has joined #exult
[06:52:01] <DominusMobile> Marzo: I had noticed the stepping through the seams when I reported the attack through walls. But then I wasn't able to reproduce it on demand;)
[06:53:38] <DominusMobile> When I moved the party through teleport cheat behind thewall and started combat it almost always happenedbut when I made a save directly before and after and reloaded I couldn't reproduce it all...
[06:54:35] <DominusMobile> These bugscan drive one crazy ;)
[06:54:46] <-- DominusMobile has left IRC (Client Quit)
[11:17:12] --> Kirben_ has joined #exult
[11:17:13] --- ChanServ gives channel operator status to Kirben_
[11:18:06] <-- Kirben has left IRC (Ping timeout: 264 seconds)
[12:39:02] --> i30817 has joined #exult
[12:40:10] <i30817> Seems like there is actually a way to specify a daily built library on launchpad
[12:40:12] <i30817> Just copy it to there or make the respective daily build recipe output to the same ppa as the program's using it. Doh!
[12:42:34] <i30817> So if i package mt32emu lib on a daily build by itself and ouput it to the exult daily build ppa, i could probably use it there... maybe.
[12:42:36] <i30817> Actually, there probably is a copying action i could do from another ppa that does that without duplication, since multiple projects will use it.
[12:43:37] <i30817> It probably loses some optimization from now compiling togheter thou
[12:44:10] <i30817> now -> not
[13:18:47] <-- i30817 has left IRC (Quit: ChatZilla [Firefox 17.0.1/20121129151842])
[13:44:00] <-- Kirben_ has left IRC ()
[14:32:24] --> TheCycoONE has joined #exult
[15:32:31] <-- TheCycoONE has left IRC (Read error: Connection reset by peer)
[15:32:41] --> TheCycoONE has joined #exult
[17:26:48] --> Dominus has joined #exult
[17:26:48] --- ChanServ gives channel operator status to Dominus
[17:38:12] --> Malignant_Manor has joined #exult
[17:39:06] <Malignant_Manor> Marzo: How did the original handle path eggs on restore? Currently, Exult screws it up.
[17:39:31] <Marzo> The original saves schedule information which we don't
[17:39:48] <Marzo> All schedules resume from where they left off in the original
[17:39:50] <Malignant_Manor> You should have an SI saved game called Move objects through door that shows the automaton open the door he closed and walk away.
[17:40:08] <Marzo> Yeah, I noticed it too
[17:41:55] <Malignant_Manor> I think range combat is working better than it was. I'm not sure if this bug is fixed. https://sourceforge.net/tracker/?func=detail&aid=3294929&group_id=2335&atid=102335
[17:42:35] <Malignant_Manor> I think this one is fixed. https://sourceforge.net/tracker/?func=detail&aid=3598584&group_id=2335&atid=102335
[17:53:34] <Marzo> I have some ideas for improving ranged combat further, but they will take some time
[17:54:47] <Malignant_Manor> I've had some issues with screen fading.
[17:55:28] <Malignant_Manor> It doesn't fade out all the time.
[17:56:17] <Dominus> marzo, he comes, he sees, he shoots
[17:56:33] <Dominus> thanks for another bug killing spree
[17:56:44] <Dominus> finally returned home again
[17:57:18] <Dominus> I'll see about combing through some of the bugs again to see whether they are still valid.
[17:57:33] <Malignant_Manor> Just walk through the gate. Restore if fade clears until you can get the screen to stay faded out.
[17:59:44] <Marzo> I managed to trigger it once -- in 40 attempts so far
[18:00:17] <Malignant_Manor> It happens more often for me but usually takes half a dozen tries.
[18:00:54] <Malignant_Manor> It's something pretty bad for users to encounter.
[18:21:48] <Malignant_Manor> Und sched. opcode 0x00 comes up 3 times in a row when it happens.
[18:22:43] <Dominus> what are the steps to reproduce it?
[18:22:46] <Malignant_Manor> This is after Returning from usecode 06a3
[18:23:09] <Malignant_Manor> just keep going into the gate
[18:23:22] <Dominus> what gate? what savegame?
[18:23:47] <Malignant_Manor> Restore if it doesn't happen. I tired to send the file.
[18:24:10] <Dominus> ah, my irc client isn't acdepting files, I think
[18:24:14] <Malignant_Manor> I just used the teleport by the FoV dragon.
[18:24:53] <Malignant_Manor> Unfortunately, after being really frequent, it started to be less so.
[18:29:40] <rrva> Marzo: new, less frequent, dragging crash: http://hastebin.com/fagafamafi.vbs (no reproduce yet, but suspect something touchpad-related, button1/button3)
[18:48:37] <-- TheCycoONE has left IRC (Read error: Connection reset by peer)
[18:53:40] <-- Matt_O has left IRC (Quit: Leaving)
[19:04:03] <Marzo> rrva: what where you doing when the crash happened?
[19:04:14] <Marzo> I mean, what where you dragging and into what
[19:04:55] <rrva> Marzo: I don't remember :(
[19:05:01] <rrva> I will update here when it crashes again
[19:05:20] <rrva> with your fix from yesterday, play is much less crashy
[19:12:03] --> Matt_O has joined #exult
[19:28:47] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.89 [Firefox 17.0.1/20121128204232])
[20:26:32] <Dominus> marzo, I think you broke ships
[20:26:45] <Marzo> Please explain
[20:26:52] <Dominus> you cannot double click the sail anymore
[20:26:57] <Dominus> it says "blocked"
[20:27:56] <Dominus> also the gangplank is saying "blocked"
[20:28:43] <Dominus> chairs are also blocked
[20:29:25] <Dominus> beds as well
[20:31:47] <Marzo> Can you give some saves for the beds and chairs? My attempts with them worked perfectly
[20:32:12] <Dominus> hmm, do you still have the save for the ship pathfinding?
[20:32:20] <Dominus> with a full party
[20:33:19] <Marzo> Ah, yes, I see it
[20:33:46] <Marzo> With the chairs, not so much the bed I am trying
[20:34:05] <Marzo> Ah, got it
[20:34:12] <Dominus> with starting the game again, no more problems with chairs and beds
[20:34:14] <Dominus> curios
[20:34:18] <Dominus> same savegame
[20:35:37] <Dominus> and seems if I hit the sweet spot the sail uncursl as well
[20:35:48] <Dominus> but can't uncurl it
[20:37:41] <Dominus> https://sourceforge.net/tracker/?func=detail&aid=3158257&group_id=2335&atid=102335 this should be the save
[20:41:21] <Marzo> WFT? How is it that "fromvol.x = t.tx - fromvol.w + 1;" with t.tx = 2168 and fromvol.w = 1 ends up as 2164?
[20:42:31] <Dominus> that makes no sense
[20:43:06] <Marzo> This is happening right before my eyes when trying to get in a bed in the save I am testing
[20:44:43] <Dominus> but where is that equation? found only one fromvol.x in paths.cc
[20:44:57] <Marzo> Line 502
[20:45:05] <Marzo> It is supposed to be an assignment
[20:46:34] <Dominus> (grr, didn't search on, don't ask me why not)
[20:47:05] <wjp> optimizations causing out-of-order execution confusing the debugger?
[20:47:43] <Marzo> Even after the ensuing function call?
[20:48:02] <wjp> did you check t.tx before or after?
[20:48:25] <wjp> (since Get_closest_edge may modify that)
[20:48:37] <Marzo> I did before and after, yes
[20:54:11] <Marzo> Hrm. Dumping to stderr shows that it was indeed GDB being confused
[20:58:56] <rrva> anyone have working sound ultima7 linux dosbox?
[21:02:12] <wjp> sure
[21:02:35] <wjp> oh, if this is about pulseaudio I have no idea about that
[21:02:52] <Dominus> it is :)
[21:03:07] <wjp> yeah, I suddenly remembed the question in #dosbox from earlier :-)
[21:03:15] <rrva> so I should remove pulseaudio? I can do that
[21:03:35] <rrva> I tried "pasuspender dosbox".
[21:04:50] <wjp> who said anything about removing pulseaudio?
[21:06:54] <rrva> i was thinking if dosbox accessing the device through alsa directly would improve things?
[21:07:17] <wjp> I honestly have no idea about anything pulseaudio-related
[21:07:27] <wjp> never even used it
[21:08:40] <wjp> but I suppose the first things to check are if audio in other SDL applications/games works, and if audio in other games in dosbox works
[21:08:53] <wjp> dosbox does not access the device through alsa directly. It uses SDL
[21:09:29] <wjp> (which may in turn use alsa, but I'm not sure in which order that would try things)
[21:11:10] <wjp> oh, and as far as I know mididevice is irrelevant if you set up u7 to use soundblaster audio
[21:17:20] <-- Dominus has left IRC (Read error: Connection reset by peer)
[21:17:25] --> DominusExult has joined #exult
[21:17:26] --- ChanServ gives channel operator status to DominusExult
[21:31:03] --> Malignant_Manor has joined #exult
[21:32:08] <Malignant_Manor> Is the Avatar supposed to walk around at 0 hp? I noticed that control with the mouse is wrong. You also cannot move with the keyboard.
[21:33:08] <Malignant_Manor> Stat gumps aren't closeable with the checkmark when hp is 0 or lower. It should be closeable that way.
[21:38:59] <rrva> wjp: I confirmed dosbox soundblaster works in another app, but it does not in u7
[21:44:29] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.89 [Firefox 17.0.1/20121128204232])
[21:44:50] <rrva> sorry, found the cause
[21:46:59] <DominusExult> marzo, the initial commit for optimizing pathfinding rev 7224 is what broke the ships
[21:47:27] <DominusExult> on the plus side the pathfinding to the seats now works without teleporting :)
[21:47:48] <DominusExult> (even though still not as fluid as the original)
[21:50:07] <DominusExult> rrva: so what was it?
[21:56:21] <rrva> ehrm, embarrasingly it was u7 GAMEDAT/OPTIONS.CFG
[22:24:03] <-- Marzo has left IRC (Ping timeout: 260 seconds)
[22:44:00] --> Marzo has joined #exult
[22:50:44] --- DominusExult is now known as Dominus
[23:07:31] --> Kirben has joined #exult
[23:07:31] --- ChanServ gives channel operator status to Kirben
[23:14:11] --> Marzo1 has joined #exult
[23:14:12] <-- Marzo has left IRC (Disconnected by services)
[23:48:49] --> DominusExult has joined #exult
[23:48:49] --- ChanServ gives channel operator status to DominusExult
[23:51:41] <-- Dominus has left IRC (Ping timeout: 245 seconds)
[23:51:42] --- DominusExult is now known as Dominus
[23:57:50] <-- Dominus has left IRC (Read error: Connection reset by peer)
[23:58:07] --> Dominus has joined #exult
[23:58:07] --- ChanServ gives channel operator status to Dominus