#gemrb@irc.freenode.net logs for 25 Mar 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[00:06:49] <fuzzie> oh, because multiplying things by zero is not smart.
[00:08:22] <fuzzie> ok, magic missiles all worky.
[00:28:03] --> Gekz_ has joined #GemRb
[00:29:39] <-- barra_away has left IRC (*.net *.split)
[00:29:39] <-- |Cable| has left IRC (*.net *.split)
[00:29:41] <-- otwieracz has left IRC (*.net *.split)
[00:29:41] <-- raevol has left IRC (*.net *.split)
[00:29:41] <-- Gekz has left IRC (*.net *.split)
[00:29:43] <-- tomprince has left IRC (*.net *.split)
[00:29:46] <-- anji has left IRC (*.net *.split)
[00:29:46] <fuzzie> it is going wrong because the magic missiles are happily rushing on *past* their target.
[00:31:57] <ratpor> I'd call that a good balance fix for Magic Missle.
[00:32:24] --> tomprince has joined #GemRb
[00:32:53] --> anji has joined #GemRb
[00:32:53] --> barra_away has joined #GemRb
[00:32:53] --> |Cable| has joined #GemRb
[00:32:53] --> otwieracz has joined #GemRb
[00:33:54] --> raevol has joined #GemRb
[00:34:25] <fuzzie> ok, much better now.
[00:38:59] <Genraznx> Multiply by zero ftw
[00:39:03] <Genraznx> infinity loop!
[00:41:10] <fuzzie> i first said i would fix this in June, certainly no-one can accuse me of rushing..
[00:41:13] <fuzzie> night all!
[00:41:15] <-- Enverex has left IRC (Quit: Enverex)
[00:41:47] <fuzzie> (am I still allowed to commit to svn?)
[00:44:58] <Genraznx> night fuzzle
[00:45:41] <Genraznx> http://art.penny-arcade.com/photos/817651900_UPD5c-L.jpg
[00:48:26] <Edheldil_> good night
[00:48:39] <-- Edheldil_ has left IRC (Quit: Really?)
[01:01:39] <-- barra_away has left IRC (Read error: Connection reset by peer)
[02:04:53] <-- anji has left IRC (*.net *.split)
[02:04:53] <-- |Cable| has left IRC (*.net *.split)
[02:04:54] <-- otwieracz has left IRC (*.net *.split)
[02:06:32] --> anji has joined #GemRb
[02:06:32] --> |Cable| has joined #GemRb
[02:06:32] --> otwieracz has joined #GemRb
[03:16:58] <Genraznx> I just had a idea
[03:17:12] <Genraznx> Anyone here familiar with Strategy RPG genre?
[03:17:14] <Genraznx> Tactic Ogres
[03:17:18] <Genraznx> Final Fantasy Tactics
[03:55:04] --> Brendan__ has joined #GemRb
[03:55:30] <-- Gekz_ has left IRC (Read error: Connection reset by peer)
[04:19:27] --- Brendan__ is now known as Gekz
[04:19:34] <-- Gekz has left IRC (Changing host)
[04:19:34] --> Gekz has joined #GemRb
[04:32:21] <-- ratpor has left IRC (Quit: Get used to disappointment.)
[04:48:39] <-- Gekz has left IRC (Read error: Connection reset by peer)
[04:49:38] --> Gekz has joined #GemRb
[05:21:45] <-- Gekz has left IRC (Read error: Connection reset by peer)
[05:22:12] --> Gekz has joined #GemRb
[05:33:57] <tomprince> svn import pushed to or.cz
[05:34:17] <Genraznx> Ah Tomeprince!
[05:34:17] <tomprince> One is the raw import, and one is with the commit messages cleaned up.
[05:34:22] <Genraznx> Just the person I wanted to speak to
[05:34:31] <Genraznx> Are you familiar with Strategy RPG's?
[05:34:43] <tomprince> No.
[05:34:43] <Genraznx> Tactic Ogre's Final Fantasy Tactics etc etc
[05:34:46] <Genraznx> Okay
[05:34:53] * Genraznx ponders how to explain it then.
[05:35:05] <Genraznx> In essence it works on a 'episodic' based design
[05:35:17] <Genraznx> PLot plot plot and then your put into map to deal with stuff.
[05:35:37] <Genraznx> I think for anything would be the best short term goal to design
[05:36:03] <Genraznx> Also... I was curious that if it was possible to design a switch and replace system so we could switch out 'placeholder' stuff
[05:36:12] <Genraznx> until we can find a artist and such for maps and shit
[05:38:19] <tomprince> Yes.
[05:39:02] <Genraznx> Well
[05:39:05] <tomprince> Really, all you would need to do to switch fake art for real art is to move all the stuff around.
[05:39:15] * Genraznx nods.
[05:39:19] <Genraznx> Well then, that helps quite a bit
[05:40:27] <tomprince> Have you had a look at dltcep?
[05:40:34] <Genraznx> A little bit
[05:40:40] <Genraznx> just skimmed it mostly though
[05:40:48] <Genraznx> Thats why i am so confidant that making a game is rather easy
[05:40:57] <Genraznx> Just make maps place them and whatnot
[05:41:03] <Genraznx> 'doesnt' seem too hard
[05:41:07] <Genraznx> but you never know heh
[05:41:49] <-- Gekz has left IRC (Read error: Connection reset by peer)
[05:41:53] <tomprince> Well, so the only thing you would need to do, when you got real art, is switch out the graphics, and adjust the position of everything to match the art.
[05:41:58] --> Gekz has joined #GemRb
[05:42:07] * Genraznx nods
[05:45:06] <Genraznx> http://www.youtube.com/watch?v=feP9LG_VgHc
[05:45:12] <Genraznx> Think of that battle system in terms of IE
[05:45:31] <-- wjp has left IRC (*.net *.split)
[05:45:33] <Genraznx> Where you just go around a map with a small band on 'mission's
[05:46:12] <Genraznx> http://www.youtube.com/watch?v=S-ZeZXR3CKk&feature=related
[05:46:17] <Genraznx> here is a better
[05:46:27] <Genraznx> representation
[05:52:40] --> wjp has joined #GemRb
[05:55:06] <tomprince> I don't think that battle system maps very well to gemrb. You could probably get the same tactical feel, but it would be vastly different, sine it is essentially realtime.
[05:55:26] <tomprince> wjp: svn import uploaded to or.cz, if you want to have a look.
[05:56:12] <Genraznx> I realize that of coarse
[05:56:14] <Genraznx> but in essence
[05:56:16] <Genraznx> you go around a map
[05:56:18] <Genraznx> kill shit
[05:56:20] <Genraznx> and advance the plot
[05:56:25] <Genraznx> from map to map
[05:56:26] <Genraznx> in essence
[05:56:31] <tomprince> That is certainly doable.
[06:16:32] <Genraznx> if anything a more plausible goal
[06:41:46] <tomprince> Good night all.
[06:42:36] <Genraznx> night tom
[06:42:51] <tomprince> And if anybody wants anything else done to the svn-import, I can do it. I can probably make git-filter-branch sit and beg, if I need to. :)
[07:33:47] <Genraznx> seems that there is competition for GemRB
[07:34:03] <Genraznx> The Broken Hourglass is a game made by polish? people or something in the style of Baldurs GAte
[07:38:25] <-- Gekz has left IRC (Read error: Connection reset by peer)
[07:38:46] --> Gekz has joined #GemRb
[08:25:22] --> lynxlynxlynx has joined #GemRb
[08:25:22] --- ChanServ gives channel operator status to lynxlynxlynx
[08:43:48] <Edheldil> tomprince: do you think it's better to leave git-svn IDs in as-is, rewrite them to st. like "SVN revision: 1234" or get rid of them completely? For the other modules I am going to get rid of them
[08:52:26] <Lightkey> why SVN revision?
[09:00:46] <-- Gekz has left IRC (Read error: Connection reset by peer)
[09:01:01] --> Gekz has joined #GemRb
[09:05:00] --> Gekz_ has joined #GemRb
[09:06:54] --> Brendan__ has joined #GemRb
[09:07:38] <-- Gekz has left IRC (Ping timeout: 246 seconds)
[09:08:50] --> Gekz has joined #GemRb
[09:08:54] <-- Brendan__ has left IRC (Read error: Connection reset by peer)
[09:09:40] <-- Gekz_ has left IRC (Read error: Connection reset by peer)
[09:31:27] <fuzzie> Genraznx: one of the The Broken Hourglass people comes in here to chat sometimes, I think.
[09:34:32] <-- |Cable| has left IRC (Remote host closed the connection)
[09:35:42] <Edheldil> Lightkey: because of the potential references in bugreports etc.
[09:36:10] <Edheldil> fuzzie: installers module can't be cloned from svn to git :(
[09:36:20] <Edheldil> could you try it?
[09:36:32] <fuzzie> What's the svn url?
[09:38:48] <lynxlynxlynx> gemrb/installers
[09:39:04] <lynxlynxlynx> same level as ielister and chitem
[09:40:03] <Edheldil> https://gemrb.svn.sourceforge.net/svnroot/gemrb/installers
[09:40:48] <Edheldil> Tried git svn clone https://gemrb.svn.sourceforge.net/svnroot/gemrb/installers --authors-file=users.txt --no-metadata -s installers , but it apparently gets lost. Could anyone try it?
[09:41:56] <fuzzie> .. interesting.
[09:43:28] <fuzzie> It is not worth cloning, I guess.
[09:43:35] <fuzzie> It clones fine for me, but with three revisions.
[09:44:22] <fuzzie> If it's really only the bg1/install-bg1.sh then we should just drop it, since http://git.njw.me.uk/cgit/cgit.cgi/gemrb-gameinstallers/ exists.
[09:44:30] <Edheldil> it clones for you? Then it's my version of git, I suppose
[09:45:01] <Edheldil> is it the same project?
[09:45:05] <fuzzie> It complains about missing history while cloning, but then "W: Do not be alarmed at the above message git-svn is just searching aggressively for old history.".
[09:46:46] <Edheldil> yes, but then it walks svn revisions and ends without any output. But I see that installers is not standard layout, guess that's the reason
[09:46:48] <fuzzie> Edheldil: I'm not sure it's the same project, but it certainly makes the old script irrelevant.
[09:46:54] <fuzzie> Oh, right, you were passing -s.
[09:47:14] <Edheldil> ok, I will drop it, then
[09:47:47] <lynxlynxlynx> i'm not sure if the two changes were synced
[09:48:04] <lynxlynxlynx> but this can be fixed without an import
[09:48:25] <Genraznx> im curious
[09:48:34] <Genraznx> is there any limitations graphic wise for maps and models?
[09:48:48] <Edheldil> lynxlynxlynx: what changes?
[09:49:03] <lynxlynxlynx> 3 revisions == 2 changes
[09:50:19] <Edheldil> what about the launcher? it looks stale to me as well and I vaguely remember some rm -rf ~ nastiness, but I am not sure if there's a better source for it somewhere
[09:52:35] <Genraznx> Also, any idea if its possible to put in your own font?
[09:53:38] <Edheldil> it is
[09:54:26] <Genraznx> cool
[09:54:56] <Genraznx> Just making sure you can put any form of image in ditcep and then make it into a map?
[09:55:04] <Genraznx> or is there a limit on polygons or detail or some shit
[09:55:46] --> Gekz_ has joined #GemRb
[09:55:47] <-- Gekz has left IRC (Read error: Connection reset by peer)
[09:59:47] <-- Gekz_ has left IRC (Read error: Connection reset by peer)
[10:00:50] --> Gekz_ has joined #GemRb
[10:09:18] <Genraznx> Nice, I go to a irc chatroom based on animation
[10:09:20] <Genraznx> all I get are bots
[10:11:53] <fuzzie> Edheldil: No idea about the launcher.
[10:12:15] <Edheldil> ok
[10:12:20] <fuzzie> but http://bitbucket.org/nickdaly/gemrb-launcher/ was updated 10 days ago.
[10:13:16] <fuzzie> I don't know if that is the same thing.
[10:25:08] <Edheldil> I think it is
[10:25:51] <Genraznx> Mmm...
[10:26:00] <Genraznx> Trying to think
[10:28:38] <Edheldil> fuzzie: the author is the same, but the version we have here seems to be obsolete
[10:29:08] <fuzzie> I would assume it's not worth importing, unless lynxlynxlynx thinks so.
[10:29:33] <lynxlynxlynx> it's in a separate repo now, so no need to bother
[10:29:46] <fuzzie> oops, late for a meeting.. bbl
[10:31:44] <Edheldil> hehe
[11:13:57] <Lightkey> Edheldil: does that mean it would still be available through Subversion or what does that have to do with Subversion?
[11:16:30] <lynxlynxlynx> subversion is getting subverted
[11:16:59] <Edheldil> yep
[11:17:49] <Edheldil> we want to switch SVN off. I am not sure if SF will keep it read only or not at all
[11:38:05] <Lightkey> Edheldil: I was referring to your proposal for the *SVN* revision to remain in the commits
[11:38:21] <Lightkey> does not make any sense to me
[11:44:20] <Genraznx> Hms... deviant art isnt a bad place to look for artists.
[11:44:27] <Genraznx> Just really need a animator I suppose ;(
[11:46:36] <wjp> we can't really turn SVN off, only hide it
[11:47:01] <wjp> (but that shouldn't be a problem)
[11:51:00] <Lightkey> soka
[12:11:20] <-- Gekz_ has left IRC (Quit: Leaving)
[12:24:43] <fuzzie> so can I commit to svn, or some git repository? :)
[12:27:04] --> barra_library has joined #GemRb
[12:28:57] --> Gekz has joined #GemRb
[12:39:01] <Edheldil> Lightkey: I was talking about leaving SVN revision number in Git commit messages. Clear now?
[12:40:18] <Edheldil> i.e. if in addition to 'Fixed PST rendering problem' commit message there should be also written 'SVN rev: #1234'
[12:41:03] <Edheldil> to facilitate references from old tbugreports
[12:41:45] <Edheldil> wjp: I know, but what hits me as strange that we can completely destroy Git repo :)
[12:41:58] <Edheldil> at least I think so, did not try :)
[12:42:07] <wjp> well, we can destroy the SVN repo too, by re-importing an empty one
[12:42:30] <fuzzie> i think a sf staffer has to ok any destructive re-imports, no?
[12:42:48] <Edheldil> fuzzie: no idea, depends on tomprince's progress withit
[12:42:50] <fuzzie> but git repositories are so easily entirely cloned that it's not such a big problem
[12:43:33] <Edheldil> regardless, hiding it to prevent confusion is enough
[13:02:02] <fuzzie> I need a reliable way to work out the ID of a missile, I guess.
[13:02:13] <fuzzie> Don't suppose anyone knows how the projectiles work?
[13:02:45] <fuzzie> My curved missiles patch is just doing '70 - type', but that is presumably only going to work for the specific spell I'm using.
[13:06:28] <-- Genraznx has left IRC ()
[13:13:07] <lynxlynxlynx> can't you get the 70 from the projectile name?
[13:13:35] <fuzzie> I don't have access to any of the data..
[13:13:48] <lynxlynxlynx> these are in our override
[13:14:12] <fuzzie> I mean, when magic missile is fired, all the projectiles from the relevant spell level downward are made.
[13:14:40] <lynxlynxlynx> oh, it's not in the name
[13:14:44] <fuzzie> So when Imoen casts type 70, I get a type 70, a type 69 and a type 68, without context.
[13:15:33] <lynxlynxlynx> that's too much for her level?
[13:15:36] <fuzzie> So I guess I need to add some context..
[13:15:53] <fuzzie> no, but I need to translate those to something which tells me how they should arc :)
[13:16:17] <fuzzie> I can simply hard-code it to start at 68, but then it only works for magic missiles.
[13:16:22] <lynxlynxlynx> the id shouldn't matter, just the number of them
[13:17:23] <fuzzie> Unfortunately there's no context at the moment, they just exist individually..
[13:17:32] <lynxlynxlynx> but yeah, you don't know where the limit is
[13:17:45] <fuzzie> I guess CreateIteration() needs to increment a coun ter.
[13:17:54] <lynxlynxlynx> there is also a force missile spell, but i don't remember how it was animated
[13:19:37] <fuzzie> do you have gemrb right now?
[13:20:35] <lynxlynxlynx> right now
[13:21:11] <fuzzie> http://fuzzie.org/nfs/gemrb/missile_arc.txt is what I wrote - would like to know if it works with higher levels.
[13:21:34] <fuzzie> I've forgotten how it's meant to look, but it can be tweaked, I guess.
[13:25:40] <lynxlynxlynx> i didn't see any missile
[13:25:54] <lynxlynxlynx> and for some reason gemrb is awefully slow
[13:26:03] <lynxlynxlynx> eating all of my cpu
[13:28:18] <fuzzie> any idea if that's caused by the patch?
[13:28:27] <lynxlynxlynx> it worked before in that save
[13:28:38] <lynxlynxlynx> could be it's a magic resistance thing though
[13:28:44] <lynxlynxlynx> it works with imoen
[13:28:51] <lynxlynxlynx> 4 or 5 orbs
[13:29:23] <lynxlynxlynx> i doubt it has an effect on the nomnomnom
[13:30:06] <lynxlynxlynx> the orbs look ok, but in the original the animation was slower
[13:30:48] <lynxlynxlynx> oh, i thought of something - how do you turn off the advanced fog?
[13:30:55] <lynxlynxlynx> maybe it got reset for me
[13:32:57] <fuzzie> i don't know.
[13:37:18] <lynxlynxlynx> ah, 3d accel needs to be off
[13:39:21] <lynxlynxlynx> doesn't help much :(
[13:40:18] <lynxlynxlynx> but let's focus on the missiles
[13:40:25] <fuzzie> nothing else showing up in top?
[13:40:44] <lynxlynxlynx> oh come on :P
[13:40:59] <fuzzie> well, it is not as crazy a question as it seems
[13:41:22] <fuzzie> when i run gemrb w/pulseaudio running, pulseaudio jumps to use all available cpu
[13:42:10] <fuzzie> otherwise svn is as fast as ever, for me.
[13:42:14] <lynxlynxlynx> i guide my own boat, so i don't have pulseaudio
[13:53:59] <fuzzie> lucky :)
[13:57:25] <lynxlynxlynx> it's really counterproductive to have major distros be early adopters
[13:58:07] <fuzzie> well, this is ubuntu, who messed it up quite a bit anyway.
[14:04:38] <-- raevol has left IRC (Quit: Leaving.)
[15:31:57] --> Maighstir_laptop has joined #GemRb
[15:38:54] <tomprince> I don't have any problem running pulse under gentoo.
[16:02:41] <-- Maighstir_laptop has left IRC (Quit: Maighstir_laptop)
[16:06:46] <fuzzie> you are presumably running a modern version, though :)
[16:10:10] <tomprince> Yes.
[16:11:44] <tomprince> Brach posted with git-svn-id cleaned up.
[16:12:41] <fuzzie> as long as you have a modern version and any ALSA apps are happy running under emulation, it's really nice now.
[16:13:44] <tomprince> It is, yes. The version that first shipped with ubuntu was horrible though, and I have that disable on the computers I maintain.
[16:18:30] <fuzzie> The svn-import-svn-rev branch?
[16:20:47] <tomprince> Yes.
[16:21:07] <fuzzie> Looks good.
[16:21:51] <lynxlynxlynx> did you fix your authorship?
[16:22:25] <fuzzie> Seems so.
[16:23:52] <tomprince> Yes. There is also the raw branch, from before I did that.
[16:24:17] <lynxlynxlynx> then it's good to go?
[16:24:34] <wjp> tomprince: do you have a log of the exact commands you ran?
[16:24:38] <lynxlynxlynx> i'll start updating the docs on the wiki
[16:25:33] <tomprince> I could probably extract them. I did a number of things and didn't keep a record of exactly what.
[16:25:37] <tomprince> Why do you ask?
[16:26:34] <lynxlynxlynx> why don't you just pull his branch?
[16:27:37] <wjp> for future reference
[16:27:49] <wjp> and educational purposes :-)
[16:28:18] <tomprince> git svn clone -s -A $PWD/gemrb/svn.authors --use-log-author https://gemrb.svn.sourceforge.net/svnroot/gemrb/gemrb/ gemrb-svn
[16:28:53] <tomprince> git filter-branch --env-filter 'grep -q $GIT_COMMIT /tmp/ids && export GIT_AUTHOR_NAME="Tom Prince" GIT_AUTHOR_EMAIL=tom.prince@ualberta.net || true ' --msg-filter 'grep -q $GIT_COMMIT /tmp/ids && vipe || cat' db3dff9e24c014dedbbedfe6f5eff8c4ab561a9e..master
[16:29:16] <tomprince> where /tmp/ids contained the commits to edit, and db3dff9e24 was the commit before the first one.
[16:29:32] <tomprince> git filter-branch --msg-filter 'cat -s'
[16:29:47] <tomprince> to get rid of the extra blank line before the git-svn-id
[16:30:16] <tomprince> git filter-branch --msg-filter "sed '/git-svn-id/s/.*@\([0-9]*\).*/svn-id: r\\1/'"
[16:30:24] <lynxlynxlynx> what's vipe?
[16:31:13] <wjp> editor in the middle of a pipe, I believe
[16:31:14] <tomprince> vipe is from moreutils, it runs $EDITOR on stdin, and the dumps it to stdout.
[16:31:22] <tomprince> wjp: exactly.
[16:31:34] <tomprince> git diff =(git log --pretty=fuller -p svn-import-raw) =(git log -p --pretty=fuller HEAD)
[16:31:45] <tomprince> to check for sanity for the changes.
[16:31:57] <tomprince> That last only works on zsh,
[16:32:09] <tomprince> That is more or less what I did.
[16:33:12] <wjp> is that <(...) in bash?
[16:33:16] <fuzzie> If the change is kept quiet for a couple of days, we could just rewrite the repository for any problems.
[16:33:29] <fuzzie> And so it seems a bit silly to worry about it too much.
[16:33:41] <wjp> (run command with output to some /dev/fd and substitute the name of the fd in its place)
[16:34:21] <wjp> what kind of edits did you do manually, by the way? (the /tmp/ids)
[16:35:03] <fuzzie> Removing my attributions to him from commit messages, I expect.
[16:35:16] <tomprince> Almost, it creates temporary files instead, as git treats the fds as special files.
[16:35:17] <fuzzie> In a couple of places I didn't think very much about how I did that.
[16:35:51] <tomprince> Yes, basically putting the original commit message I had in place of the truncated one that got into svn.
[16:36:06] <wjp> right
[16:36:36] <wjp> sounds good
[16:37:04] <lynxlynxlynx> "process substitution"
[16:37:42] <lynxlynxlynx> the wiki docs are subversion free
[16:40:25] <fuzzie> the wrong page seems to not be :)
[16:40:29] <fuzzie> erm, the front page.
[16:40:49] <tomprince> fuzzie: I'm fine with what is there. It isn't up to me at this point.
[16:41:06] --> Maighstir_laptop has joined #GemRb
[16:41:46] <wjp> any objections to doing the move this weekend?
[16:42:08] <wjp> I'd like to stare at the repo for a bit to see if there's anything else we want to rewrite while we have the chance
[16:42:44] <tomprince> Not from me.
[16:45:24] <lynxlynxlynx> we can work with git-svn as before and then just apply the patches once you're comfortable
[16:47:26] <tomprince> Or could add people to the or.cz repo.
[16:48:09] <tomprince> OR, as fuzzie suggested, just push it to sf, and then we can just change if we need to. So long as it isn't public, it shouldn't be a problem.
[16:49:00] <tomprince> As I said, it doesn't make a difference to me at this point.
[16:49:17] <lynxlynxlynx> we don't get much attention, but i doubt you can make private repos on sf
[16:52:20] <tomprince> I meant, simply don't publicize it. Anyway, git is fairly good about rebased branches nowadays. It still doesn't like merging, or pulling of them, but other than that it isn't a problem.
[17:38:20] <Edheldil> if you commit to SVN, you will have to update the git repo somehow afterwards
[17:38:53] * Edheldil has just finished writing threaded ping in perl. eek!
[17:40:03] <Edheldil> later, I will try the repo at home
[18:00:29] --> Enverex has joined #GemRb
[18:42:51] <lynxlynxlynx> fuzzie: need any more feedback on the missile patch?
[18:43:56] <fuzzie> I guess Avenger should probably look at it :)
[18:54:23] <fuzzie> i would commit it now if i knew the call lightning was instant too
[18:54:54] --> |Cable| has joined #GemRb
[18:55:16] <fuzzie> but that is how, right? so i know nothing about it
[18:55:39] <lynxlynxlynx> bg2 also has call lightning, but they're different
[18:56:19] <lynxlynxlynx> iirc it didn't work as intended before either; in bg2 it is activated once each turn of the duration
[18:56:34] <fuzzie> i think this code only handles the how case
[18:57:35] <fuzzie> (the SKYBOLT2 projectile)
[18:57:40] <tomprince> fuzzie: I have pushed a patch that contains all the autoconf changes.
[18:57:53] <tomprince> Well, all the ones that aren't in some other commit.
[18:58:14] * tomprince goest to seminar.
[18:59:37] <fuzzie> that dlsym thing seems .. odd
[19:02:48] <fuzzie> reinterpret_cast is no fix, though -- is the __extension__ the magic here?
[19:03:25] <fuzzie> i guess it must be
[19:03:38] <cfchris6> I assume the bug with overwriting slots while saving is known? ( "gemrb's buggy save code is trying to delete slot 8\nthat is not the slot 1 we were trying to save to, erroring out!")
[19:04:05] <fuzzie> cfchris6: yes.
[19:04:29] <fuzzie> i added the check after gemrb deleted all of my savegames one day.
[19:04:48] <fuzzie> it seemed better to have the check than have it eat people's saves.
[19:05:44] <cfchris6> yea, now that I saw it in the console it is not a real problem. I just wondered "hadn't I saved tha game that other day?" once.
[19:07:41] <fuzzie> the solution is to create autosave (slot 0) and quicksave (slot 1) saves first.
[19:08:14] <fuzzie> those are made almost immediately in the original engine, and most people use an existing install, so i guess it went unnoticed for a long time.
[19:08:37] <cfchris6> autosave is created on travel iirc. what was the default combination for quicksave again?
[19:08:37] <fuzzie> definitely on the list of things to fix..
[19:11:21] <fuzzie> tomprince: s/Simply/Simplify/, remove the LT_INIT, fix the plugins-prepare.sh, change the DelayPlugin name in the default config, and warn in the commit message that it has to be re-run?
[19:12:21] <fuzzie> i also don't know whether removing @LIBPTHREAD@ works for everyone.
[19:15:43] <fuzzie> that is a frustrating number of thoughts for such a simple patch.
[19:25:02] <lynxlynxlynx> cfchris6: q
[19:26:18] <fuzzie> it's not very helpful in gemrb, though.
[19:26:22] <fuzzie> due to the slot problem.
[19:28:17] <cfchris6> hm, seems the widescreen mod wasn't such a good idea. now the gorion+sarevok cutscene does not work anymore
[19:28:26] <fuzzie> Oh?
[19:29:01] <cfchris6> gorion and sarevok fight for about 5min. then gorion dies (I thought this was shorter than 5min)
[19:29:16] <cfchris6> and now sarevog and one of his party members are standing there and nothing happens
[19:29:26] <cfchris6> s/sarevog/sarevok/
[19:29:26] <fuzzie> Yes, if they go on fighting for more than a few seconds then the script has failed, I guess.
[19:29:32] <fuzzie> Which version of gemrb is this?
[19:29:38] <cfchris6> 0.6.0
[19:30:11] <cfchris6> what maybe has to be said: the widescreen mod asks weather IE or GemRB. I chose IE as GemRB would fail to install
[19:30:28] <fuzzie> That is okay, I think.
[19:30:32] <lynxlynxlynx> you need a development version
[19:30:48] <cfchris6> will try
[19:31:00] <fuzzie> Damn, I thought zekflop and I had fixed that cutscene..
[19:31:05] <fuzzie> But it was fixed before 0.6.0.
[19:31:12] <cfchris6> the funny thing is, the unpatched BG Tosc and not widescreen mod worked when I tried
[19:31:14] <lynxlynxlynx> i thought it was fixed too
[19:31:16] <cfchris6> so I think it is the mod
[19:31:30] <fuzzie> cfchris6: It is more likely to be the savegame you used, if it is the same kind of problem.
[19:31:54] <cfchris6> I started a new game as the widescreen mod's readme states to to so on tots
[19:31:56] <cfchris6> +c
[19:32:19] <fuzzie> that is not a problem for gemrb, but okay, good that it is not a savegame load bug :)
[19:32:30] <fuzzie> I don't see how it can be the widescreen mod, though..
[19:33:12] <cfchris6> well, I'll build gemrb from svn (or is it git right now?) and check again
[19:33:22] <lynxlynxlynx> i meant the mod
[19:33:26] <fuzzie> oh right, I added a hack for this, and it was completely the wrong hack.
[19:33:28] <cfchris6> lynxlynxlynx: ah
[19:33:53] <fuzzie> This function has a worrying number of "fuzzie changed this" comments.
[19:34:43] <fuzzie> "fuzzie changed this", "fuzzie added this", "fuzzie moved this here", "i did it this way for now"..
[19:34:59] <lynxlynxlynx> and another helm of balduran bug -.-
[19:35:26] <lynxlynxlynx> with reequipping you can get any ac you want :)
[19:35:28] <Lightkey> it's so fuzzy!
[19:35:30] <cfchris6> lynxlynxlynx: do you know where I can find the dev version of widescreen mod?
[19:35:41] <lynxlynxlynx> github
[19:36:04] <fuzzie> The function also has such notes from wjp and Avenger, a popular place to be.
[19:36:13] <fuzzie> Must replace them all with the right code sometime..
[19:36:31] <lynxlynxlynx> ah, it's just the timing
[19:36:54] --> kettuz has joined #GemRb
[19:55:28] <cfchris6> the dev wsm installed fine in gemrb mode. old savegame didn't work with cutscene, new game did.
[19:57:09] <fuzzie> Looks like the savegame slot bug dates all the way back to the first code in November 2003.
[20:09:23] <fuzzie> I don't want to break our support for multiple savegames with the same slot number, it's quite helpful when people email saves for testing.
[20:09:46] <fuzzie> But I'm not sure how to cope with overwriting those slots, since we can't work out which game is being referenced.
[20:14:02] <-- |Cable| has left IRC (Ping timeout: 246 seconds)
[20:26:20] --> |Cable| has joined #GemRb
[20:33:58] --> barra_away has joined #GemRb
[20:36:58] <-- barra_library has left IRC (Ping timeout: 264 seconds)
[20:46:25] <cfchris6> funny, I slept one night in the friendly arm and now I have a savegame with 605 days playtime
[20:48:32] <-- otwieracz has left #GemRb
[20:48:38] <fuzzie> hehe. i wonder what causes that?
[20:52:48] <fuzzie> (is the gained HP really RentIndex+1?)
[20:53:11] <lynxlynxlynx> it was likely a high value before that too
[20:54:42] <cfchris6> hm, there are some enemies which I can't attack (I get a "forbidden" sign on mouseover)
[20:54:44] <fuzzie> The core also seems to set the "HOUR" token in Game::RestParty, as well as the caller GUIScript doing it.. bit silly?
[20:55:02] <cfchris6> before it was at two days btw
[20:56:28] <fuzzie> But it advances time by 2400 in-game seconds, seems reasonable enough. Someone needs to confirm by saving immediately before+after the rest, I guess.
[20:58:46] <cfchris6> hm, I can't reproduce it when trying again :/
[20:58:58] <cfchris6> fact is there is exactly one savegame with 605 days playtime
[20:59:07] <cfchris6> all the others are reasonable
[21:00:39] <fuzzie> lynxlynxlynx: Do you think all the projectiles are too fast, or just magic missiles?
[21:01:45] <lynxlynxlynx> i think the former, but i haven't tried recently
[21:01:57] <lynxlynxlynx> i do remember the aoe ones were too fast
[21:05:45] <fuzzie> The bg2 gemprjtl.ids file refrences SPMAGMIS, but that seems to only exist as a bam. I am confused.
[21:06:29] <lynxlynxlynx> maybe it's the bg1 version
[21:08:14] <fuzzie> Oh, there's a 'shared' folder in override. :)
[21:13:14] --> cfchris6_ has joined #GemRb
[21:15:05] --> Edheldil_ has joined #GemRb
[21:17:19] <-- cfchris6 has left IRC (Ping timeout: 276 seconds)
[21:27:39] <tomprince> fuzzie: I update the patch with those changes.
[21:28:16] <tomprince> I can't comment on the -lpthread changes, but none of the plugins care about threading, or reference any thread local variables or anything.
[21:34:46] <fuzzie> If you changed the makefiles, we presumably need a new line for NullSound.dll, too..
[21:35:31] <fuzzie> And that "-sahred" is still in there :)
[21:36:41] <Lightkey> sounds arabian
[21:39:23] <fuzzie> The pthread thing seems reasonable to drop, since it'd be trivial to revert it back in.
[21:40:49] <tomprince> I was wondering about MSVC for the dll?
[21:42:27] <fuzzie> Yes, so we'll have two dll filenames on Windows for a bit.
[21:49:55] <tomprince> Stupid recursive make. :(
[21:50:13] <tomprince> And libtool that recompiles on install.
[21:54:04] <tomprince> Pushed.
[22:06:05] --> Gekz_ has joined #GemRb
[22:07:41] <fuzzie> I still think it makes sense to put the vorbis bit in there, but mostly because I'm finding it hard to see which bits of "Resource support for sound" are about resource support as it is (I see a bugfix adding alDeleteBuffers in there), and I can understand why you'd find it silly to tweak makefiles which just get destroyed in that commit.
[22:07:41] <-- kettuz has left IRC (Quit: Leaving)
[22:12:50] <fuzzie> It needs a clean make, too, but I guess people will work that out.
[22:12:51] <tomprince> I could move it. The reason that I didn't originally, was because the first commit there just changes the Makefiles in a uniform way.
[22:13:43] <tomprince> It comes from watch the git mailing list for a couple of years. Where most every patch is tiny, changing the least amount possible, and major changes go in as a long patch series.
[22:14:34] <fuzzie> We usually try to keep our patches minimal, too.
[22:15:31] <tomprince> I can split out the vorbis configure bit and the bug fix if you want.
[22:17:18] <fuzzie> well, it doesn't really matter :) i just find it so difficult to review the patches, because I've made so many minor changes before and broken things, rearchitectural work makes me nervous.
[22:18:36] <tomprince> I can understand that. I wouldn't want any of the stuff I posted commited without someone familiar with the code base to review them for that reason.
[22:18:48] <fuzzie> I had thought the @LIBPTHREAD@ thing was more than just linking in the library, but now I've looked acinclude.m4, it's obviously just that.
[22:19:21] <fuzzie> So that seems completely harmless.
[22:19:39] <tomprince> I would like to think that even if the changes do break things, the new code is easier to understand and fix than the old code.
[22:19:50] <tomprince> That is my goal, anyway.
[22:20:44] <fuzzie> Yes, my worries are more along the lines of assumptions in the design .. for instance, I still didn't confirm whether it's fine for the Bitmap class to discard all the data it does.
[22:21:43] <fuzzie> The trouble being that we don't *use* the colour/high-bit data right now, but I never finished looking into exactly how the original engine behaves with the few irregular BMP files used.
[22:24:41] <fuzzie> ( http://forums.gibberlings3.net/index.php?showtopic=17717 )
[22:26:55] <tomprince> Well, if we do figure out what to do with those, we can fix the algorithm. And if Bitmap needs to store more precision, then we can change it.
[22:28:31] <tomprince> Unrelated question: is there any reason for the silent paramater the gets passed into ResourceMgr?
[22:29:04] <fuzzie> At a guess, it would mean "expected to fail, don't print an error".
[22:29:13] <lynxlynxlynx> console spam
[22:40:51] <fuzzie> If you want to re-architect something, you could fix SaveGameIterator. :-p
[22:44:18] <fuzzie> Refactoring the console output out of KEYImporter doesn't look like much fun.
[22:46:11] <lynxlynxlynx> a queue system for all the messages would be nice, now they can get in the middle of each other
[22:46:38] <lynxlynxlynx> but this is not really important
[22:47:03] <fuzzie> Well, there's a conflicting problem where it is nice to have the message output before the segfault happens.
[22:47:55] <lynxlynxlynx> you make it sound like we get that alot :P
[22:48:56] <fuzzie> well, it is nice to have an obvious thing to look at when someone provides a log :)
[22:51:38] <fuzzie> Presumably for debugging you could just have a nice obvious function to prod.
[22:52:11] <tomprince> Well, or you could have a runtime option to queue messages until they are complete.
[22:52:24] <tomprince> Which would handle messages interrupting each other.
[22:52:32] <fuzzie> well, bug reports tend to come from people running release builds.
[22:52:35] <tomprince> Which you could turn off if debugging segfaults.
[22:52:44] <fuzzie> so, *flail*.
[22:53:04] <tomprince> Do we have release builds?
[22:53:24] <lynxlynxlynx> do we have debug builds? :)
[22:53:41] <fuzzie> i think the point where i start wondering about signal handlers is the point where i am obviously overthinking the problem, so you probably shouldn't listen.
[22:54:19] <lynxlynxlynx> it's a cosmetic thing anyway
[22:54:20] <fuzzie> but yes, most people are running "release" builds, in the sense that they're running a packaged released version.
[22:54:33] <lynxlynxlynx> optimised and stripped
[22:55:19] <tomprince> That is why you make it a runtime option. Flip a switch and all logging output is dumped as it is generated. And by default, it is nice and ordered.
[22:55:32] <fuzzie> and it's nice to be able to pinpoint problems from posted logs, because it's often a drive-by thing.
[22:56:32] <fuzzie> but it would definitely be a nice off-by-default option to turn on in the future. :)
[22:56:58] <fuzzie> and in any case i suppose it's not as if we are lacking in bugs.
[22:57:25] <fuzzie> so again, i am obviously overthinking this :)
[23:00:23] --> raevol has joined #GemRb
[23:06:39] <-- Gekz_ has left IRC (Quit: This computer has gone to sleep)
[23:13:52] <fuzzie> tomprince: maybe it's important to say: i just speak thoughts out loud here, please don't take anything i say as gospel -- in the end i am just a random contributor who wants to make sure anything committed is maintainable :)
[23:14:15] <tomprince> I assumed that. :)
[23:14:54] <fuzzie> ok. i just don't want you thinking that my random thoughts are any kind of requirements!
[23:15:10] <fuzzie> are you ok with waiting on the build system patch until we officially move to git?
[23:15:31] <tomprince> Yes.
[23:16:04] <fuzzie> it would be convenient from my point of view, because then all random users will make a clean checkout, and i don't have to worry about people becoming confused about the .so rename :)
[23:16:31] <tomprince> My repo seems to be the unofficial, offical repo until the switch is actually offical. :)
[23:16:50] <tomprince> Or something.
[23:17:14] <fuzzie> very nice to have had the commits tidied up :)
[23:17:52] <tomprince> Yes, my issue was more with the commit messages than the authorship, although the second is nice as well.
[23:19:49] <fuzzie> i will be particularly grateful for the decentralisation; i have had rearchitecting patches sitting on my web server as .txt files for weeks, just because in an svn world it was the easiest way to get it reviewed.
[23:20:11] <fuzzie> cherry-pick is still my favourite git toy.
[23:24:44] <tomprince> That is exactly why I setup the or.cz repo, since putting patches in to SF was getting to be a pain.
[23:25:10] <tomprince> I think I had like 20+ over a month, some being updated patches.
[23:25:34] <cfchris6_> having patch branches which get rebased on the master branch is my way to do that with git...
[23:26:23] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:27:56] <fuzzie> i would imagine that wouldn't work so well for rearchitecting patches which can interact with people's own changes?
[23:28:48] <-- Enverex has left IRC (Quit: Enverex)
[23:29:17] <fuzzie> not such a huge problem for anything i've ever done for gemrb, but working around svn tends to ensure people push changes to the master branch quickly..
[23:29:36] <fuzzie> but you can tell i am no version control expert :)
[23:30:24] <cfchris6_> that was at a "lawful good dictator" project where only the original author could push to the "main" tree
[23:30:32] <tomprince> No, probably doesn't work well if there are interactions between patches.
[23:31:04] <tomprince> The linux kernel?
[23:31:15] <tomprince> Except that uses merges rather than rebase mostly.
[23:31:27] <cfchris6_> nope, arora
[23:31:36] <cfchris6_> qt-webkit web browser
[23:32:16] <tomprince> Well, mostly I was just commenting that the kernel, and git itself both work that way, with a benevolent dictator.
[23:33:15] <tomprince> fuzzie: #define SHARED_OVERRIDE in KEYImporter.cpp?
[23:33:19] <fuzzie> It seems to work very well from an outside point of view.
[23:33:53] <fuzzie> tomprince: see gemrb/override/shared/ in the gemrb tree.
[23:34:03] <fuzzie> No idea why it's a define.
[23:34:19] <tomprince> It seems a silly define.
[23:36:37] <fuzzie> I think you're fine removing it, but lynx was the author, so he'd know why it's there.
[23:37:55] <fuzzie> There was no discussion about it, anyway, lynx just added the functionality after we'd talked on IRC about avoiding file duplication.
[23:39:08] <fuzzie> Very random :)
[23:42:08] <fuzzie> ok, ninight
[23:48:28] <tomprince> night.
[23:48:38] <-- Maighstir_laptop has left #GemRb
[23:49:57] --> Genrazn has joined #GemRb
[23:50:05] <Genrazn> hi
[23:54:54] --> Genraznx has joined #GemRb
[23:55:19] <tomprince> Hello.
[23:57:52] <-- Genrazn has left IRC (Ping timeout: 276 seconds)