#exult@irc.freenode.net logs for 21 Nov 2010 (GMT)

Archive Today Yesterday Tomorrow
Exult homepage


[00:31:54] --> jvlee has joined #exult
[01:24:57] <-- jvlee has left IRC (Quit: jvlee)
[02:36:27] --> jvlee has joined #exult
[02:37:41] <-- jvlee has left IRC (Client Quit)
[03:07:55] * dominus gives up for now... compiled all night long all the @&%§$ gtk+/glade stuff manually and when I finally got there I failed with building Exult because I had SDL built without X some time ago...
[03:08:04] <dominus> resumeing tomorrow,,, :)
[03:55:22] <-- Morde has left IRC (Read error: Connection reset by peer)
[06:29:18] <-- dominus has left IRC (Ping timeout: 245 seconds)
[07:30:47] --> dominus has joined #exult
[07:30:47] --- ChanServ gives channel operator status to dominus
[08:20:08] <-- dominus has left IRC (Ping timeout: 245 seconds)
[09:55:14] --> Malignant_Manor has joined #exult
[09:58:04] --> jvlee has joined #exult
[10:01:14] <Malignant_Manor> These two bug reports seem invalid due to being the same as the original game
[10:01:15] <Malignant_Manor> Respawning (includes file) https://sourceforge.net/tracker/?func=detail&aid=2989069&group_id=2335&atid=102335
[10:01:17] <Malignant_Manor> Chests explode, but no major damage https://sourceforge.net/support/tracker.php?aid=3053840
[10:01:19] <Malignant_Manor> These have bug fixes.
[10:01:20] <Malignant_Manor> Keyring Mod: companion death summons guards https://sourceforge.net/support/tracker.php?aid=3011711
[10:01:21] <Malignant_Manor> Saving/Resotring game causa Full Mana https://sourceforge.net/support/tracker.php?aid=3001287
[10:28:08] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
[12:08:44] --> Dominus has joined #exult
[12:08:45] --- ChanServ gives channel operator status to Dominus
[12:19:45] <Dominus> moved the respawn bug to feature requests
[12:37:08] <wjp> hm, I wonder about this "companion death summons guards" bug
[12:37:24] <wjp> this taking-damage stuff is too complicated... :-(
[12:38:23] <Dominus> sounds a bit risky to change without having a full overview of what it might do besides fixing that
[12:39:18] <wjp> in theory it should be pretty safe since it adds a _very_ specific check, but yes...
[12:41:25] <Dominus> wjp can you find a former mention of triple crossbow bolt or some variation in the logs? it's about https://sourceforge.net/tracker/index.php?func=detail&aid=3053282&group_id=2335&atid=102335 and I'm sooooo convinced we had this problem before...
[12:42:35] <Dominus> it's driving me nuts to be thinking that I knew of this before but can't remember clearly...
[12:44:59] <wjp> none of the mentions of 'bolt' in here seem relevant
[12:48:19] --> SiENcE has joined #exult
[12:49:18] <Dominus> strange... the mind ... a wonderful thing to waste...
[12:51:28] <Dominus> I'm worried about the random crashes report in the forum
[12:57:43] <Dominus> hmm, any idea what "NPC write" means in our console?
[12:58:20] <wjp> saving npcs to file
[12:58:26] <wjp> a pretty pointless message
[12:58:31] <wjp> (like many of them)
[12:59:40] <Dominus> :)
[13:08:01] <-- jvlee has left IRC (Quit: jvlee)
[13:09:02] <-- Kirben has left IRC ()
[13:09:50] <wjp> committed that "companion death summons guards" thing after all
[13:09:59] <Dominus> :)
[13:10:13] <wjp> also silenced two annoying stdout messages
[13:10:43] <Dominus> yeah, I saw that and had to smile :)
[13:11:16] <Dominus> approach nsteps is really annoying :)
[13:16:13] <Dominus> hmm, another annoying message is the "saved game #-2 successfully." This could be worded better, maybe add the filename or savegame name instead of #-2 (whatever #-2 is supposed to mean)
[13:18:35] <wjp> quicksave, I think
[13:22:45] <wjp> I wonder if that 'Error while listing files: The system cannot find the file specified.' is relevant
[13:23:40] <wjp> but probably not...
[13:26:35] <Dominus> some kind of time stamp could come in handy in the error log :)
[13:27:58] <Dominus> about "saved game #-2..." it only prints when saving to a new save
[13:28:08] <Dominus> #-1 when overwriting a savegame
[13:30:15] --> Malignant_Manor has joined #exult
[13:30:38] <Malignant_Manor> Charmed is broken (not related to recent changes)
[13:43:16] <Dominus> did that ever work? :)
[13:43:38] <Dominus> thanks for looking into all of this. much appreciated, btw
[13:44:09] <Malignant_Manor> Well, I'm not sure if they always had a free for all after someone got charmed.
[13:45:30] <wjp> party members being charmed or charming monsters? who are 'they'?
[13:48:09] <Malignant_Manor> For sure, a party member being charmed and the whole party seems to have a free for all.
[13:48:29] <Malignant_Manor> I'll see about some charmed monster checking.
[13:48:55] <Malignant_Manor> 1.2 charm doesn't even highlight, let alone affect combat.
[13:51:43] <Dominus> hence my thinking it never really worked
[13:57:35] <-- Malignant_Manor has left IRC (Ping timeout: 260 seconds)
[13:59:10] <Dominus> wjp, my x11 problem is based on a problem with newer xorg stuff. I compiled all the necessary glade/gtk stuff myself against the system headers (and today compiled SDL with x11 support - I left that out for the Mac snapshot) it works fine
[13:59:23] --> Malignant_Manor has joined #exult
[14:00:19] <Malignant_Manor> https://sourceforge.net/tracker/index.php?func=detail&aid=2797068&group_id=2335&atid=102335
[14:01:34] <wjp> Dominus: ok, good :-)
[14:02:42] <Dominus> malignant, that's similar to the bug I encountered at https://sourceforge.net/tracker/?func=detail&aid=3053811&group_id=2335&atid=102335 (not the reported bug, but the one I encountered)
[14:04:58] <Dominus> wjp, now I only need to find out which xorg lib is the problem maker and report it :)
[14:05:01] <Dominus> *only*
[14:05:56] <Malignant_Manor> I'm still could never get party members to attack thorough means other than charm though unlike the report by the tracker creator.
[14:23:25] --> Morde has joined #exult
[14:28:19] <-- Malignant_Manor has left IRC (Ping timeout: 240 seconds)
[14:33:33] --> Malignant_Manor has joined #exult
[14:34:55] <Malignant_Manor> This machine compiles slow and I keep losing my connection so I have no idea when I can track down when any of this started happening.
[14:48:34] <Dominus> sorry, I probably can't be of much help... don't know whether it will compile really old revisions here...
[14:57:59] <-- Malignant_Manor has left IRC (Ping timeout: 255 seconds)
[14:59:37] --> Malignant_Manor has joined #exult
[15:00:38] <Malignant_Manor> Charm was working properly with party members at 5911 (BIG Commit (TM))
[15:02:22] <Dominus> June 4th 2008...
[15:03:29] <Dominus> cab´n you attach a savegame to the bug report so I can just "point and shoot, I have snapshots since april so I can at least trouble shoot whether it broke till then
[15:06:12] <-- SiENcE has left IRC (Quit: cya @all)
[15:07:39] <Malignant_Manor> just charm a party member with the cheat menu and then start combat
[15:08:32] <Malignant_Manor> Also, Exult at least improves on explosives and has them actually hurt the Avatar.
[15:13:52] <Malignant_Manor> 6030 might be a likely candidate for breaking (Make NPCs more aggressive about fighting back)
[15:15:55] <Dominus> at least it was broken in April already, that I can say...
[15:22:24] <Malignant_Manor> At least, I was right about it actually working at one point in time.
[15:23:50] <Dominus> :)
[15:24:39] <Dominus> I'm trying to build 6029
[15:27:02] <-- Malignant_Manor has left IRC (Ping timeout: 264 seconds)
[15:30:16] * Dominus thinks that compiling old versions is annoying...
[15:31:30] --> Malignant_Manor has joined #exult
[15:32:25] <Malignant_Manor> 5980 works (just randomly picked one before looking at the revision log)
[15:33:12] <Dominus> #6029 doesn't
[15:36:07] <Dominus> checked out 6018 now
[15:36:40] <Dominus> and preparing compile (editing autogen.sh and coreaudiomidi)
[15:37:51] <Malignant_Manor> 6026 seems like another likely, trying 6025
[15:44:51] <Malignant_Manor> Does anyone here use Debian? I wonder if the compiling issue got fixed recently.
[15:46:38] <Dominus> malignant you can stop compiling 6025
[15:46:44] <Dominus> 6018 is also broken
[15:47:32] <Dominus> going to try 6008
[15:50:05] <Malignant_Manor> It's basically guessing time.
[15:50:45] <Dominus> yes, that's what I'm doing... going ten revisions back to pinpoint the breakage
[15:51:14] <Dominus> I just hope I don't run into any gcc issues
[15:56:17] <wjp> hm, no commit messages between 5980 and 6018 really stand out at a glance
[15:58:48] <Dominus> 6008 is also bad
[15:59:08] <-- Malignant_Manor has left IRC (Ping timeout: 276 seconds)
[16:04:28] --> Malignant_Manor has joined #exult
[16:06:08] <Dominus> 5988 is good
[16:08:30] <Dominus> trying 5994 now
[16:09:28] <wjp> hmm... having trouble seeing which of the commits in that interval it is :-(
[16:09:44] <wjp> but 5992 is huge... *scans*
[16:10:40] <Dominus> we'll get there....
[16:11:31] <Malignant_Manor> It would be nice if this problem couldn't be in a bunch of different files.
[16:14:14] <Malignant_Manor> I wish I had a super computer that could compile from scratch in less than 30 seconds.
[16:15:37] <Malignant_Manor> 5996 broken too
[16:15:40] <Dominus> he he
[16:16:16] <Dominus> 5994 isn't
[16:16:20] <Dominus> so...
[16:16:56] <Dominus> it's either 95 or 96 that broke it
[16:17:31] <Dominus> I'll have a go at 5995
[16:19:38] * wjp is now really confused
[16:20:29] <Malignant_Manor> My compiler stopped working
[16:21:23] <Dominus> Malignant_Manor: kick it
[16:22:03] <Malignant_Manor> I wonder if a process isn't deleting itself when I exited the program.
[16:22:52] <Malignant_Manor> Make runs, eats up all my cpu, but not a single line of input.
[16:25:02] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
[16:26:28] <Dominus> damnit
[16:26:37] <Dominus> 5994 IS broken
[16:27:08] <Dominus> I only set the wrong flag accidently (cursed insted of charmed)
[16:28:02] <Dominus> and 5988 is also broken
[16:28:12] * Dominus is A noob
[16:29:14] * Dominus is annoyed
[16:29:21] <Dominus> trying 5983
[16:33:29] --> Malignant_Manor has joined #exult
[16:35:13] <Dominus> Malignant_Manor: I made a mistake 5988 was already broken
[16:35:30] <Malignant_Manor> I'm compiling 5980 again to make sure it didn't just appear to be working. I have a better save to test with.
[16:36:28] <Malignant_Manor> It would really suck if I screwed up back them.
[16:36:43] <Dominus> true :)
[16:36:59] <Dominus> if not it would be really cool, because 5983 is broken
[16:37:07] <Malignant_Manor> It just needs 15 or so minutes.
[16:37:21] <Malignant_Manor> Oh, I hate ai.
[16:37:45] <Dominus> oh well, in that time, I can have 5981 compiled :)
[16:38:19] <wjp> hm, interesting
[16:38:36] <wjp> 5988 claims to revert previous commits, but misses a single line
[16:39:34] <Malignant_Manor> Is it fixed later?
[16:39:53] <wjp> when testing, you should skip *85, *86, *87, by the way
[16:40:12] <Dominus> wjp, I already tested 83 and it was bad
[16:40:54] <wjp> ah, oops
[16:47:21] <-- Malignant_Manor has left IRC (Ping timeout: 240 seconds)
[16:49:22] --> Malignant_Manor has joined #exult
[16:51:32] <Malignant_Manor> 5980 does seem to work. I'll transfer over the SI file to see if training is also working.
[16:53:53] <Dominus> hmm, that's strange 5981 does NOT work - test case a new game in BG, changed NPC flags of iolo to charmed, attack mode
[16:54:30] <Dominus> and the only files changed in 5980 are studio.cc/h
[16:54:33] <wjp> 5981 is a mapedit commit...
[16:55:42] <Malignant_Manor> Is training also broken on your end?
[16:56:12] <Dominus> I'm not testing training, just whether charmed works without iolo and avatar fighting
[17:00:21] <-- Malignant_Manor has left IRC (Ping timeout: 240 seconds)
[17:03:32] --> Malignant_Manor has joined #exult
[17:03:44] <Malignant_Manor> grrr, I changed weapons around and it was buggy
[17:04:36] <Malignant_Manor> 5911 better work or it's the crazy bug for me
[17:07:27] <Malignant_Manor> 5924, 5925, and 5927 look like good places for screwups
[17:08:30] <Malignant_Manor> That's assuming 5911 actually works.
[17:11:30] <Malignant_Manor> The training bug only seemed to happen with party members and not the Avatar.
[17:19:00] <-- Malignant_Manor has left IRC (Ping timeout: 265 seconds)
[17:19:51] --> Malignant_Manor has joined #exult
[17:20:13] <Dominus> I'm checking out 5911 now, 5953 was still buggy
[17:20:34] <Dominus> keeping all those revisions around for further testing :)
[17:21:25] <Malignant_Manor> 5911 worked with SI training. It seems to work with charm except it doesn't choose new targets after charm wears off so you need to toggle combat.
[17:22:25] <Dominus> can't compile 5911, configure error
[17:22:39] <Malignant_Manor> Try 5927
[17:23:10] <Dominus> k
[17:23:52] <Dominus> just so we are on the same page, the companions *shouldn't* attack the Avatar once they are charmed, right?
[17:24:08] <Dominus> because right now it seems charming just reverses allignement
[17:25:06] <Malignant_Manor> The charmed person should be attacked by the rest of the party.
[17:25:17] <Dominus> anyone being hostile should be friendly and friendly should be hostile - it doesn't seem to consider *who* charms
[17:25:23] <Malignant_Manor> The charmed person should attack a random party member.
[17:25:51] <Dominus> ok, and what is it doing wrong?
[17:25:56] <Malignant_Manor> Yeah, charm is neutral
[17:26:05] <Dominus> because that's what happens in all the builds for me
[17:26:34] <Malignant_Manor> I forget what charm does to neutral targets.
[17:27:29] <Dominus> so what is it actually doing wrong?
[17:27:44] <Dominus> or have we been going on a wild goose chase?
[17:28:15] <Dominus> please give me a real test case :)
[17:28:33] <Dominus> I'm off to do some other work :)
[17:34:05] <-- Malignant_Manor has left IRC (Ping timeout: 250 seconds)
[17:34:30] --> Malignant_Manor has joined #exult
[17:36:20] <Malignant_Manor> trying 5927 from a clean compile. For some reason, make keeps getting stuck. Maybe I need a new checkout.
[17:40:40] <Dominus> I still don't know what to look for anymore :)
[17:42:11] <Malignant_Manor> SI training, party members aren't attacked afterward when toggling combat.
[17:42:41] <Malignant_Manor> Charmed party member attacks random party member. Everyone else attacks the charmed person.
[17:43:01] <Dominus> and that is wrong?
[17:43:10] <Dominus> because if that is right, then i
[17:43:19] <Dominus> that works for me with rc1
[17:43:58] <Dominus> the training I will have to try much later or tomorrow
[17:44:39] <Malignant_Manor> RC 1 party members that aren't charmed attack each other.
[17:48:29] --> Malignant_Manor_ has joined #exult
[17:48:37] <Malignant_Manor_> It's broken in 5927
[17:49:40] <-- Malignant_Manor has left IRC (Ping timeout: 245 seconds)
[17:49:41] --- Malignant_Manor_ is now known as Malignant_Manor
[17:50:41] <Malignant_Manor> Also, other training people seem to attack the party member who is training.
[18:03:46] <-- Malignant_Manor has left IRC (Ping timeout: 265 seconds)
[18:04:32] --> Malignant_Manor has joined #exult
[18:04:49] <Malignant_Manor> 5924 seems to be working fine
[18:09:39] <wjp> it's starting to sound like *27 then
[18:14:23] <Malignant_Manor> Well, unfortunately, I cannot compile minor changes without cleaning out all the files.
[18:14:59] <Dominus> why don't you do a checkout for each revision? That's what I have been doing and thus am keeping all the compiled files
[18:16:49] <Dominus> sorry don't know how good your internet connection is...
[18:18:41] <-- Malignant_Manor has left IRC (Ping timeout: 245 seconds)
[18:19:07] --> Malignant_Manor has joined #exult
[18:19:32] <Malignant_Manor> 5927 is Fixed monsters attacking each other
[18:20:10] <Malignant_Manor> I don't have a whole lot of hard drive space to be keeping tons of extra stuff.
[18:21:14] <Malignant_Manor> 5925 is Fixed bug that could cause large creatures to attack themselves with projectiles. This should also prevent other creatures from attacking allies.
[18:23:53] <wjp> nah
[18:24:06] <wjp> 5925 just changed the tile a missile is being launched from
[18:24:22] <wjp> (to prevent it from directly hitting the attacker, presumably)
[18:25:50] <Malignant_Manor> 5925 works
[18:28:08] <Malignant_Manor> So hopefully you can find what is wrong with 5927. At least, this is 2 bugs in one.
[18:34:17] <-- Malignant_Manor has left IRC (Ping timeout: 255 seconds)
[18:35:50] --> Malignant_Manor has joined #exult
[18:37:04] <wjp> ok, 5927 is definitely weird
[18:40:11] <wjp> can you summarize what _exactly_ is going wrong? The backlog is too confusing
[18:42:39] <Malignant_Manor> In SI training at the fields, party members other than the Avatar are attacked when toggling combat after finishing training.
[18:43:20] <Malignant_Manor> Other trainers seem to attack the training member sometimes when if in range. (I guess)
[18:44:09] <wjp> what do you mean by "are attacked"? by who?
[18:44:27] <Malignant_Manor> When a party member is charmed, every party member seems to try to kill everyone else. (haven't tested with more than 3 members for a long time)
[18:44:47] <Malignant_Manor> The other party members attack them.
[18:45:40] <Malignant_Manor> Party members all kill each other.
[18:46:35] <wjp> 5927 changed things to this logic: if X is attacking a party member, then X becomes a target for _any_ party member
[18:46:38] <wjp> I think
[18:49:55] <-- Malignant_Manor has left IRC (Ping timeout: 240 seconds)
[18:51:51] --> Malignant_Manor has joined #exult
[18:53:18] <wjp> do you read the logs or should I repeat?
[18:54:17] <Malignant_Manor> I read the statement about logic change.
[18:54:43] <Malignant_Manor> List field save should be here https://sourceforge.net/tracker/?func=detail&aid=3053811&group_id=2335&atid=102335
[19:00:06] <Malignant_Manor> I have no idea what to do though.
[19:04:49] --> Malignant_Manor_ has joined #exult
[19:05:05] <-- Malignant_Manor has left IRC (Ping timeout: 250 seconds)
[19:05:16] --- Malignant_Manor_ is now known as Malignant_Manor
[19:15:41] <Malignant_Manor> commenting out the whole 'else if (in_party)' seems to work fine unless the Avatar is charmed.
[19:19:50] --> Malignant_Manor_ has joined #exult
[19:20:26] <-- Malignant_Manor has left IRC (Ping timeout: 264 seconds)
[19:20:32] --- Malignant_Manor_ is now known as Malignant_Manor
[19:22:35] <Malignant_Manor> Actually, I does work if he is charmed.
[19:34:09] <-- Malignant_Manor has left IRC (Ping timeout: 265 seconds)
[19:35:42] --> Malignant_Manor has joined #exult
[19:41:07] <Malignant_Manor> I have no idea if that section is needed for anything. I'm really tired and need sleep.
[19:41:21] <-- Malignant_Manor has left IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])
[21:28:07] --> ParuCodex has joined #exult
[21:29:59] <-- ParuNexus has left IRC (Ping timeout: 276 seconds)
[21:44:15] --> ParuNexus has joined #exult
[21:46:36] <-- ParuCodex has left IRC (Ping timeout: 240 seconds)
[21:51:22] <Dominus> that part got a bit more ifs in current code
[21:51:50] <wjp> yes, but that particular bit survived
[21:52:35] <Dominus> any idea on it?
[21:59:02] --> Kirben has joined #exult
[21:59:03] --- ChanServ gives channel operator status to Kirben
[22:14:51] <wjp> hm
[22:15:15] <wjp> I'm not sure what it's intended to do
[22:15:17] <Dominus> hmm, I doubt that makes any sense but when I change line 388 in combat.cc to
[22:15:20] <Dominus> else if (!in_party)
[22:15:51] <Dominus> at least the attack after the training doesn't occur anymore and I think the charm thing malignant means doesn't happen
[22:16:02] <Dominus> but hard to tell who is attacking who
[22:16:07] <wjp> that makes monsters attack anyone attacking the party ;-)
[22:16:23] <Dominus> well... sounds great
[22:16:26] <Dominus> :)
[22:17:08] <wjp> there are already checks earlier checking if it's an enemy
[22:17:15] <wjp> so I'm not sure what this extra section should do
[22:18:07] <wjp> OTOH, this section predates the alignment check
[22:21:15] <Dominus> ah, at least I was now able to observe what malignant means with the chramed thing: have at least a party of four - hack move them apart on the screen. Charm one, then enter combat - you can see the party attack each other, not necessarily the charmed party member
[22:25:02] <Dominus> wjp, it's really strange but with that ! in line 388 it really seems to solve both issues
[22:25:15] <Dominus> the charmed members and the list field training attack
[22:29:20] <wjp> not so strange really :-)
[22:29:38] <wjp> it's probably better to just remove this entire block
[22:38:05] <Dominus> hmm, otoh the party is quite non-aggressive with the whole else if block removed
[22:40:59] <wjp> hm
[22:43:02] <wjp> I guess I don't know enough about how alignment works
[22:43:03] <Dominus> yes, both with "else if (!in_party)" or removing the whole block makes the party lambs
[22:45:17] <Dominus> we'd probably need to get Marzo to take a look, he's done a lot in combat.cc in the past years, so he probably knows more about it
[22:46:25] <wjp> yeah
[23:03:01] <Dominus> wjp, when I comment out 393 and 340, it seems to work
[23:03:02] <Dominus> if (t->get_flag(Obj_flags::in_party) || t == avatar)
[23:03:11] <Dominus> opponents.push_back(actor);
[23:03:27] <Dominus> so it's only that little part that poses problems
[23:04:42] <Dominus> hmm, except for the "charmed" part
[23:06:17] <Dominus> hmm, that is probably line 398 and following... we really need marzo for this, after all he added that in may :)
[23:07:05] <wjp> those push_backs are the lines actually marking actors as enemies
[23:39:57] --> jvlee has joined #exult