#gemrb@irc.freenode.net logs for 24 Mar 2010 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:00:05] <fuzzie> I am trusting it to get the --add-author-from right..
[00:01:00] <CIA-43> gemrb: 03fuzzie * r7496 10/gemrb/trunk/gemrb/ (240 files in 38 dirs): (log message trimmed)
[00:01:00] <CIA-43> gemrb: Rework plugin interface to Core.
[00:01:00] <CIA-43> gemrb: - Add macros to simplify plugin interface.
[00:01:00] <CIA-43> gemrb: - Rename interface functions to be GemRB specific.
[00:01:00] <CIA-43> gemrb: - Get rid of ClassDesc.
[00:01:01] <CIA-43> gemrb: - Make plugins be explicit about registering themselves.
[00:01:01] <CIA-43> gemrb: - Add cleanup callbacks. (Used by opcodes and AREImporter)
[00:01:13] <fuzzie> it did.
[00:02:16] <Edheldil_> heh, read to the end of the log :)
[00:02:32] <tomprince_loki> Excellent.
[00:02:55] <Edheldil_> I like the design without cogs a bit more, the "e" is easier to read
[00:02:57] <fuzzie> go forth and all test that, please :-)
[00:03:04] <Lightkey> Edheldil_: read me too?
[00:03:18] <fuzzie> i am kind of hoping Avenger will notice and test it, I am wary about applying more changes on top without his OK.
[00:04:19] <Edheldil_> Lightkey, I played part of DH and it's working without an apparent glitch, but the game dialogs are rather annoying to sit through
[00:05:52] --> _pickle has joined #GemRb
[00:10:07] <Lightkey> ganbatte!
[00:21:43] <tomprince_loki> fuzzie: about the build system. Another thing that I would like to do at somepoint is make the make not be recursive, at least inside plugins.
[00:22:45] <fuzzie> I don't even know what that would change.
[00:23:23] <tomprince_loki> It would make rebuild faster on multicore systems.
[00:23:35] <fuzzie> The only changes I'd like would be to fix cross-compilation, and that is a huge pain, but not made any more difficult by such changes..
[00:24:06] <tomprince_loki> Since most plugins only have a single file, because of how recursive make is set up, the plugin building isn't parallel.
[00:24:16] <tomprince_loki> What breaks for cross-compilation?
[00:24:24] <fuzzie> The configure script.
[00:24:31] <tomprince_loki> I have spent some time fixing that for other stuff.
[00:25:10] <fuzzie> Somewhere in the IRC logs I walked someone through sabotaging it step-by-step.
[00:25:52] <fuzzie> But basically, anything which tries running binaries is obviously going to fail.
[00:26:27] <fuzzie> And this is further complicated for the python macros, which run python to get the contents of sys.prefix.
[00:26:43] <fuzzie> I'm sure there's another project who has already fixed it all, though.
[00:28:20] <fuzzie> At a glance, I don't see why anything other than python would be failing.
[00:28:40] <tomprince_loki> Well, I don't know anything about the python stuff. But I don't see anything else particularily tricky. I do have a couple patches in my tree that get rid of a lot of cruft in configure.in.
[00:29:21] <tomprince_loki> I just never got around to cleaning it up, since there was already a backlog.
[00:32:01] <tomprince_loki> Plus I've been busy.
[00:32:19] <-- Maighstir_laptop has left IRC (Ping timeout: 260 seconds)
[00:34:17] <fuzzie> looking at IRC logs, the ZLIB_HOME=/usr/local in acinclude.m4 is also a problem.
[00:35:16] <fuzzie> or, well, maybe just the whole ZLIB_HOME idea.
[00:36:19] <fuzzie> no idea what they're trying to do there, but adding random library/include paths is not so great.
[00:36:31] <fuzzie> looks like the user disappeared before confirming it worked, though.
[00:37:34] <Edheldil_> we could switch to cmake completely, if that fixes things
[00:39:07] <tomprince_loki> Well, if we need to keep cmake, that would be my preference.
[00:39:32] <tomprince_loki> I have no idea how well cmake supports cross-compiling though.
[00:39:57] <fuzzie> 14:57 <fuzzie> if you use cmake, then it doesn't set SYSCONFDIR, so it doesn't know to look there
[00:40:23] <tomprince_loki> I think though, that a properly written autoconf build works with everything but MSVC.
[00:41:06] <tomprince_loki> It isn't like gemrb is does anything fancy.
[00:41:53] <Lightkey> your brain can't keep up with your fingers
[00:42:11] <fuzzie> I think the trouble is that a properly written autoconf build is beyond any of us.
[00:42:58] <tomprince_loki> Rather, by fingers with my brain : s/is does/is doing/ :)
[00:43:23] <tomprince_loki> Well, I have been trying to clean things up. Like I said, I have some patches that clean up more of the cruft.
[00:43:37] <fuzzie> I would be interested to see what you clean up :)
[00:43:45] <tomprince_loki> And if it doesn't work on Mac OSX, I am willing to work with someone to fix it.
[00:43:53] <fuzzie> But the autoconf troubles are that there just aren't cases for anything except pretending-to-be-Linux.
[00:45:19] <tomprince_loki> http://pastebin.ca/1850984 is the diff against the trunk (pre plugin stuff). At least on my laptop. I'm not sure which is more uptodate, my desktop or laptop.
[00:45:41] <tomprince_loki> Mostly getting rid of a bunch of checks that we don't pay attention to the result of.
[00:46:33] <tomprince_loki> Also, a fix for the cast need for dlsym that breaks on gcc-4.*.
[00:46:53] <tomprince_loki> And more consistent library checks.
[00:48:22] <fuzzie> Any idea if anything except LT_INIT is badly-supported?
[00:50:40] <tomprince_loki> No, all the machines I use run gentoo unstable, so I don't know anything about backward compatability.
[00:50:48] <tomprince_loki> Although, most of those changes are just deleting stuff.
[00:51:12] <tomprince_loki> Note, don't apply that as is, since some of the stuff depends on changes elsewhere.
[00:52:40] <fuzzie> Well, I'd like to avoid making any more rename/build changes until I see if anyone runs in here in the next few days and shouts about how I broke their build.
[00:52:41] <tomprince_loki> I think the new library checks are backward compatible.
[00:53:07] <fuzzie> I don't know if there's anything else I can apply in the meantime.
[00:53:10] <tomprince_loki> That's fine. I'm in no hurry.
[00:54:12] <fuzzie> I imagine I could apply the movie patch, although it doesn't look like it would work.
[00:54:23] <tomprince_loki> Maybe the getting rid of the hashtable in KeyImporter.
[00:55:02] <tomprince_loki> Or maybe not, that probably depends on the other changes to KeyImporter.
[00:55:22] <Edheldil_> fuzzie, what do you mean with isometric renderer? I thought that's what we have now ...
[00:55:31] <fuzzie> Edheldil_: rendering using isometric tiles.
[00:55:48] <Edheldil_> ah, for backhround?
[00:55:54] <fuzzie> and for walls etc.
[00:55:58] <fuzzie> so it's a bit complicated to do :)
[00:57:33] <fuzzie> tomprince_loki: do you know if the movie patch works?
[00:58:53] <tomprince_loki> I don't think I have had a working IWD2 to test against.
[00:59:11] <fuzzie> I don't see anything which loops on types.
[00:59:42] <fuzzie> So does it not just try the first plugin it finds and then fail?
[00:59:53] <tomprince_loki> What do you mean?
[01:00:03] <tomprince_loki> The loop should be in KeyImporter.
[01:00:16] <tomprince_loki> I think.
[01:00:17] <fuzzie> I don't see where it checks a return result.
[01:00:32] <fuzzie> It just seems to do 'return types[j].Create(ret);' or similar (typing from memory)
[01:03:29] <fuzzie> Oh, sorry.
[01:03:31] <fuzzie> I understand now
[01:03:54] <fuzzie> Files in archives don't matter, because they have a type.
[01:06:56] <tomprince_loki> That might actually break, if ie type id of acm == wav, and they are checked in the wrong order.
[01:06:59] <fuzzie> so IE_BLK_CLASS_ID should be .. 0x080F.
[01:07:11] <fuzzie> maybe.
[01:10:39] <fuzzie> should perhaps mention what those are in a comment.
[01:11:20] <fuzzie> i'd thought they were random numbers.
[01:12:09] <tomprince_loki> The last four digits aren't, if the type is something that can exist in a bif.
[01:12:30] <fuzzie> i see that now :) had better renumber some
[01:13:49] <tomprince_loki> Probably no reason to. As things move over to the plugin stuff, alot of those can go away. That is why I started the enum, for the stuff that isn't a file, or is never in a bif.
[01:14:20] <tomprince_loki> Because then we only need some unique ID.
[01:14:35] <fuzzie> Well, the bif format kept on being used post-IE, and the tools for those allow ogg/jpg/etc inside.
[01:14:47] <fuzzie> So it would seem silly to deviate from that.
[01:15:12] <tomprince_loki> Do people actually stick ogg/jpg and stuff in them? And do they have IDs allocated.
[01:15:16] <tomprince_loki> ?
[01:16:37] <Edheldil_> btw, on the pluggable script language, I have got lua embedded in GemRB as a plugin. But without a glue code, it's pointless and current GUIScript.cpp is a too thick glue :(
[01:17:05] <fuzzie> Edheldil_: you should commit it!
[01:17:24] <fuzzie> well, maybe not a good idea to build by default :)
[01:17:48] <fuzzie> tomprince_loki: yes, but i don't know how official the codes are.
[01:18:02] <fuzzie> i suspect they are grouped by game. i will have to try the tools.
[01:19:07] <fuzzie> As you say, it's pointless if nothing will actually bif the files.
[01:19:28] <fuzzie> ok, it is 2am, i am being shouted at, night :)
[01:20:30] <tomprince_loki> Edheldil_: I'd be interested in having a look at that code.
[01:23:04] <-- tomprince_loki has left IRC (Quit: tomprince_loki)
[01:24:01] <Edheldil_> it really is trivial, embedding lua is easy and there was no glue. But I will look into it
[01:39:38] <Edheldil_> good night
[01:39:58] <-- Edheldil_ has left IRC (Remote host closed the connection)
[02:17:10] <CIA-43> gemrb: 03edheldil * r7497 10/ie_shell/trunk/infinity/ (formats/__init__.py formats/baf.py formats/bcs.py stream.py):
[02:17:10] <CIA-43> gemrb: Commit before switching from SVN to Git
[02:17:10] <CIA-43> gemrb: Add not yet working BAF loader
[02:29:26] <-- _pickle has left IRC (Remote host closed the connection)
[02:37:18] <tomprince> Edheldil_: I think that my next project is probably to attack generating glue automatically again.
[02:38:11] <tomprince> So, ideally, whatever solution I come up with would work against all the scripting languages we might want to support.
[02:38:33] <tomprince> Which also raise the question, what other languages would people like to see support for?
[02:38:40] <tomprince> If any?
[03:04:09] <-- ratpor has left IRC (Quit: Get used to disappointment.)
[03:09:47] <Genraznx> what do you mean?
[03:28:19] <tomprince> All the gui code in GemRB is written in python.
[03:28:45] <tomprince> Right now, the python binding are all done by hand, and very much abstracted from the C++ code.
[03:29:02] <tomprince> I plan to implement bindings that are generated from the C++ interface.
[03:29:14] <tomprince> And there is talk of supporting more than just python.
[03:33:41] <Genraznx> I see
[03:33:55] <Genraznx> is it possible to write GemRB for Nintendo DS Format?
[03:36:28] <Genraznx> when you say bindings though
[03:36:34] <Genraznx> do you mean like buttom shortcuts?
[03:36:36] <Genraznx> or some such
[03:37:42] <tomprince> Well, no. By bindings, I mean the interface between the Python code and C++ code.
[03:38:21] <tomprince> I don't know anything about Nintendo DS, does it have support for SDL and OpenAL. Or OpenGL.
[03:38:39] <tomprince> Right now, GemRB depends on SDL for video and OpenAL for audio.
[03:38:52] <tomprince> There is talk of porting from SDL to OpenGL at some point.
[03:39:58] <Genraznx> I know there is a way to play PC games on the PSP
[03:40:04] <Genraznx> so porting it there is already possible
[03:40:07] <Genraznx> albeit limited
[03:40:14] <Genraznx> since the joystick crap
[03:41:17] * Genraznx muses
[03:41:19] <Genraznx> but if anything
[03:41:31] <Genraznx> How hard would it be to implement custom spells and the ike?
[03:41:35] <Genraznx> spell effects and such
[03:42:16] <Genraznx> Since it seems making a game for Gem RB seems relatively simple make a pretty map then animating sprites
[03:43:41] <tomprince> Spells shouldn't be too hard. Depending on how complex they are, it could be trivial. If the effect you want is the same as one in the original games, you can do it easily with scripting.
[03:43:53] <tomprince> If not, it would be more work, but not too difficult.
[03:44:09] <Genraznx> Well if one were to make a game and avoid liscensing
[03:44:18] <Genraznx> one would need to make a new 'stats' system
[03:44:25] <Genraznx> and spell effects be different
[03:44:30] <tomprince> Right now, if you were making a game, it would have to use the same rules.
[03:44:49] * Genraznx muses.
[03:44:53] <tomprince> Since all the rule code is currently spread through out the core.
[03:45:18] <Genraznx> Could I somehow 'disguise' it
[03:45:35] <Genraznx> as a different rule set?
[03:46:07] <tomprince> I don't know.
[03:46:13] * Genraznx muses
[03:46:28] <tomprince> Depends perhaps on how different it is from the original IE rules.
[03:46:53] <tomprince> Each of the games does have slightly different rules, already.
[03:47:10] <Genraznx> cant mix and match them?
[03:47:23] <Genraznx> the features of one said game to another
[03:47:52] <Genraznx> like placing the run feature from PS:T
[03:48:05] <Genraznx> to the pop up text
[03:48:19] <tomprince> Should be able to, alot of it is already controled through gemrb.ini.
[03:48:44] <Genraznx> Oh, that is interesting
[03:48:48] <tomprince> Well, PsT support is currently quite a bit behind the support for the other engines. I don't know if pop up text is supported.
[03:49:29] <Genraznx> Well making a game for GemRB seems relatively simple
[03:49:32] <Genraznx> then starting from scratch
[03:49:56] <Genraznx> only problem would be a artist and perhaps a programmer for special stuff I would like
[03:50:06] <tomprince> I suspect that if you are serious about creating a game from scracth for GemRB, you would be able to get a fair bit of support for getting features in.
[03:50:36] <Genraznx> I have some ideas floating on my head
[03:50:36] <tomprince> I know that I would be willing to help with that.
[03:50:40] <Genraznx> and decided on a villain
[03:50:57] <Genraznx> Keeping it a asian themed more so vietnam thailand and cambodia area
[03:51:00] <Genraznx> then china or japan
[03:52:52] <tomprince> You probably want to have a look at the IE-modding sites. Even if you are creating a game from scratch, you could probably use some of the community made resources. Also, they have all the documentation for creating resources for GemRB, since they are just the IE resources.
[03:53:15] <Genraznx> if anything I dont have a artist touch
[03:53:20] <Genraznx> so finding one would be key and or animator
[03:54:19] <tomprince> Although, if you want to use other formats, it has become not to diffcult to add support to GemRB. Although a lot of the code makes assumption about data formats.
[03:54:40] <Genraznx> I see I see
[03:54:46] <Genraznx> I just think its more productive to focus on story and setting
[03:54:58] <Genraznx> until I find people more qualified to fill in as programmer
[03:55:29] <Genraznx> and whatever else I would need
[03:55:39] <Genraznx> Seems easy enough to do something like scripting the maps and whatnot
[03:56:27] <tomprince> Well, I am willing to do programming on the engine.
[03:56:37] <Genraznx> Well that makes me quite happy
[03:57:02] <Genraznx> you know the engine or know how to program it.
[03:57:32] <Genraznx> want to care of the ideas lingering in my head for the setting and plot?
[04:06:44] <tomprince> I don't know how much useful input I could have.
[04:06:53] <Genraznx> If anything
[04:07:05] <Genraznx> Just getting a feel for the universe or having a interest is enough in my opinion
[04:07:49] <tomprince> Right now I am most interested in cleaning up the code base for GemRB. And being able to support a independent game requires that and helps show how to clean it up.
[04:07:59] * Genraznx nods.
[04:08:02] <Genraznx> Sounds good
[04:08:11] <Genraznx> One thing that I despise about NWN 2
[04:08:22] <Genraznx> its just feels like they just made a inferioir engine to BG2
[04:08:30] <Genraznx> and just added mehish 3d graphics
[04:09:07] <Genraznx> Though I do admit making a mod using NWN 2 creator is likely easy as pie. Considering all you need to do is put it at a angle you like take a screenshot and then do whatever needs to be done
[04:09:22] <Genraznx> which of coarse brings back to liscensing issues
[04:09:30] <Genraznx> being unable to use D&D rules is a pain considering Atari has it
[04:11:25] <Genraznx> But then I guess I just felt NWN 2 system is there just to hamper you
[04:12:44] <tomprince> Another thing to think about is that even if you do end up wanting to use another system, you can probably prototype quite a bit using the IE ruleset.
[04:13:08] <Genraznx> What do you mean?
[04:17:11] <-- tomprince has left IRC (Ping timeout: 265 seconds)
[04:18:21] <Genraznx> :(
[04:21:23] --> Gekz has joined #GemRb
[04:21:50] <-- Gekz_ has left IRC (Read error: Connection reset by peer)
[04:29:35] --> tomprince has joined #GemRb
[04:29:42] <Genraznx> welcome back
[04:29:53] <Genraznx> <tomprince> Another thing to think about is that even if you do end up wanting to use another system, you can probably prototype quite a bit using the IE ruleset.
[04:29:53] <Genraznx> <Genraznx> What do you mean?
[04:33:49] <-- tomprince has left IRC (Ping timeout: 245 seconds)
[04:40:35] --> tomprince has joined #GemRb
[04:40:46] <tomprince> Like you can have walking around, and talking, and quests and everything, which doesn't really depend on the ruleset. And you can fights and things, which you would need to rebalance if you changed the rulesets, but you would have something to show people.
[04:40:51] <tomprince> You could probably even get by with stick figures and art, if you have a compelling story line, to prototype it. And then get people to add nice graphics later.
[04:40:58] <tomprince> I think a good game is probably more about the story, than pretty graphics, or rules. At least the kind of game you would use GemRB for. So at the stage you seem to be at, it doesn't seem to make sense to worry about the rules over much.
[04:41:02] <tomprince> And flaky connection. :(
[04:41:50] <Genraznx> heh
[04:41:52] <Genraznx> I noticed
[04:42:02] <Genraznx> Indeed, for now I want to focus on setting and a general plot
[04:42:12] <Genraznx> I hafve a vilain but not so much a ending now or beginning
[04:42:15] <Genraznx> I have a vision of the world
[04:42:18] <Genraznx> a few quests in mind
[04:43:57] <Genraznx> But if anything
[04:44:18] * Genraznx muses. "I wonder where we could get a artist though."
[04:44:39] <Genraznx> Curious
[04:44:43] <Genraznx> how much free time do you have?
[04:45:29] <-- tomprince has left IRC (Ping timeout: 245 seconds)
[04:50:30] --> tomprince has joined #GemRb
[04:50:51] <tomprince> More than I should, probably. In other words, not a whole lot.
[04:51:02] * Genraznx nods.
[04:51:15] <tomprince> And I am by no means an artist.
[04:51:34] * Genraznx nods. "Never considered yo uwere."
[04:51:57] <Genraznx> Either or we have a start
[04:52:01] <Genraznx> and thats system work
[04:52:09] <tomprince> I am curious what you have in mind for needs from GemRB, that it doesn't already have.
[04:52:32] <Genraznx> Something to add in new spell effects would be key
[04:52:46] <Genraznx> Other then that...
[04:52:55] <Genraznx> Probably ease of use stuff
[04:53:22] <Genraznx> like not being forced to switch manually from melee to ranged weapons (Slings and such are fine able ot use a shield) But it would be nice to have 'weapon sets' to switch to ranged and unraged weapons.
[04:53:35] <Genraznx> Perhaps UI?
[04:53:46] <Genraznx> All in all whatever
[04:54:31] <tomprince> Spells aren't too hard. Visual effects might be, I don't know.
[04:55:28] <tomprince> Stuff like weapon sets should be doable too, and usefull for bioware games as well. That might take some work to clean up the code to the point where it could be implemented cleanly.
[05:02:04] <Genraznx> I just findi t irratating
[05:02:07] <Genraznx> that you are unable to wear a shield
[05:02:12] <Genraznx> and have a bow to switch to
[05:02:16] <Genraznx> pisses me off heh
[05:07:45] <Genraznx> another thing that I would like requested I suppose
[05:07:53] <Genraznx> Is graphic effects for aoe spells and cones?
[05:15:35] <Genraznx> suppose need to switch out the icons as wel...
[06:18:24] <tomprince> The icons are already supported.
[06:28:01] <Genraznx> kool
[06:28:41] <Genraznx> curious possible to
[06:28:50] <Genraznx> single out body parts?
[06:29:01] <Genraznx> or be able to attack a creature various body parts?
[06:33:17] <tomprince> No. There is no modelling of body parts stat wise.
[06:47:43] <tomprince> Could of course be implemented.
[06:55:11] * Genraznx nods
[06:55:13] * Genraznx muses
[07:05:29] <Genraznx> Tom you here?
[07:52:15] --> lynxlynxlynx has joined #GemRb
[07:52:16] --- ChanServ gives channel operator status to lynxlynxlynx
[08:22:18] <Lightkey> guys, the Nintendo DS hardware is ANCIENT, no OpenGL, no SDL, 4 MB RAM
[08:22:24] <Lightkey> please, just forget it
[08:22:40] <Gekz> IT CAN BE DONE
[08:22:46] <Gekz> haha, I'm joking.l
[08:22:53] <Gekz> there's no way in hell the DS can support GemRB
[08:22:55] <Lightkey> Gekz: something other maybe
[08:23:41] <Lightkey> Gekz: can you download https://espace.library.uq.edu.au/view/UQ:158682
[08:24:00] <Gekz> it seems rather slow
[08:24:06] <Gekz> which link?
[08:24:08] <Gekz> both?
[08:24:59] <Lightkey> sure
[08:25:21] <Gekz> arent you lucky I have a universal login?
[08:25:22] <Gekz> :P
[08:25:51] <Gekz> lets see if it worked
[08:26:28] <Lightkey> well, not for me
[08:27:11] <Gekz> its not downloadable
[08:27:14] <Gekz> they fucked it up
[08:27:59] <Lightkey> is that aussie for "I'm too stupid"? ;p
[08:28:08] <Gekz> no, they literally fucked it up
[08:28:17] <Gekz> I'm logged in, and it still says "You must login to download this"
[08:28:28] <Gekz> that's possibly why it has zero downloads
[08:28:39] <Gekz> perhaps email the uni requesting a copy of the thesis?
[08:29:40] <Lightkey> I think he is already asking the author himself, if he gets ahold of an address
[08:29:57] <Gekz> ok
[08:30:26] <Lightkey> you sure you have not disabled cookies? :p
[08:38:44] <Gekz> I am quite sure.
[08:38:50] <Gekz> please dont insult my intelligence l
[08:38:50] <Gekz> ol
[08:41:00] <Genraznx> fine doody head :(
[08:46:27] <Genraznx> suprising how bad the DS Hardware is
[08:47:25] <Gekz> lol
[08:47:29] <Gekz> want to buy one?
[08:47:35] <Genraznx> I do
[08:47:39] <Genraznx> but im a broke mahfuckah
[08:47:44] <Gekz> I am selling one of the old- damn you, sir
[08:47:45] <Genraznx> and get a r4 card or some other such thing
[08:58:08] <Edheldil> sigh, I can't get git post-commit hook running on SF :(
[09:14:02] <-- cfchris6 has left IRC (Changing host)
[09:14:02] --> cfchris6 has joined #GemRb
[09:26:33] <CIA-43> gemrb: 03lynxlupodian * r7498 10/gemrb/trunk/gemrb/GUIScripts/pst/GUICommonWindows.py: pst: made the pause button work and show the time
[09:29:20] <wjp> Edheldil: hm, strange. Things do look correct for the ie_shell repo
[09:31:49] <Edheldil> I have found out that the correct name is post-receive, not post-commit
[09:32:14] <wjp> yes
[09:32:27] <wjp> https://sourceforge.net/apps/trac/sourceforge/wiki/Git
[09:33:05] <Edheldil> and I have to integrate ciabot as well
[09:33:19] <wjp> so it works now?
[09:38:15] <Edheldil> It sent some mail to me, but I have to tweak it ... Don't you know from the top of your head how to display git revisions?
[09:38:43] <Genraznx> anyone know of a console command for BG2 to kill a enemy other then ctr y
[09:38:52] <Genraznx> I have this dumb mage who is completely invulnerable
[09:40:26] <-- |Cable| has left IRC (Remote host closed the connection)
[09:40:54] <wjp> Edheldil: what do you mean by display?
[09:41:50] <wjp> the patch?
[09:42:09] <wjp> if so: git show -C rev
[09:42:22] <Edheldil> I need oldrev and newrev as parameters to the post-receive hook. I thought they were commit ids, but it does not work with them
[09:42:39] <lynxlynxlynx> HEAD HEAD^ :)
[09:43:38] <wjp> post-receive gets oldrev, newrev, refname on stdin
[09:44:19] <wjp> (see the 'main loop' at the bottom of the email script)
[09:44:27] <wjp> while read oldrev newrev refname
[09:44:27] <wjp> do
[09:44:27] <wjp> generate_email $oldrev $newrev $refname | send_mail
[09:44:27] <wjp> done
[09:45:22] <Edheldil> I know, but "post-receive HEAD^ HEAD refs/heads/master" does not work either :(
[09:50:17] <lynxlynxlynx> let me check ours
[09:52:03] <lynxlynxlynx> no post-receive hooks
[09:55:50] <lynxlynxlynx> http://pastebin.com/335uP4e3 <-- this is what we use for the mail
[09:57:02] <Genraznx> HAH
[09:57:10] <Genraznx> IN YOUR FACE YOU STUPID INVULNERABLE MONSTER
[09:57:14] <Genraznx> Power Word Kill Scroll ftw
[09:58:39] <Edheldil> thank you, I will look into it
[09:58:47] <lynxlynxlynx> doesn't look like the full thing though, there should be a git whatchanged in there too
[09:59:30] <lynxlynxlynx> http://lists.ibiblio.org/pipermail/sm-commit/2010-March/028394.html <-- example output
[10:01:51] <Edheldil> grrrrrrr
[10:02:54] <Edheldil> post-receive sample lies, the args should be refname, old, new, not old, new, refname
[10:09:15] <wjp> the commandline and stdin orders are different
[10:12:08] --> Enverex has joined #GemRb
[10:12:39] --> barra_library has joined #GemRb
[10:20:23] <CIA-43> gemrb: 03lynxlupodian * r7499 10/gemrb/trunk/gemrb/GUIScripts/ (4 files in 2 dirs): pst: cache the strength tables here too
[10:21:58] <CIA-43> gemrb: 03lynxlupodian * r7500 10/gemrb/trunk/gemrb/GUIScripts/GUICommon.py: GUICommon: removed a now never matching check for the strength tables
[10:30:56] <fuzzie> morning
[10:31:21] <lynxlynxlynx> ;)
[10:32:06] <fuzzie> did the bg1 next thing get fixed
[10:32:54] <lynxlynxlynx> i added the import
[10:33:06] <fuzzie> ok :)
[10:34:29] <Edheldil> hi, fuzzie
[10:36:27] <CIA-43> gemrb: 03lynxlupodian * r7501 10/gemrb/trunk/gemrb/GUIScripts/iwd2/Start.py: iwd2: LoadCommonTables here too
[10:36:46] <fuzzie> (whee, we woke up the lurking lynx?)
[10:37:29] <lynxlynxlynx> i'm fed up with the pebkac i'd have to deal with otherwise ;)
[10:39:19] <fuzzie> no sign of Avenger?
[10:42:04] <lynxlynxlynx> not today
[10:45:48] <CIA-43> gemrb: 03fuzzie * r7502 10/gemrb/trunk/gemrb/ (8 files in 4 dirs):
[10:45:48] <CIA-43> gemrb: Make movies use new plugin system.
[10:45:48] <CIA-43> gemrb: Signed-off-by: Tom Prince <tom.prince@ualberta.net>
[10:45:54] <fuzzie> ack.
[10:46:42] <fuzzie> i mean, Signed-off-by: is still good for the git-svn import, but why From yesterday and Signed-off-by today?
[10:47:07] <fuzzie> oh, because tomprince's commit already had it.
[10:49:09] <fuzzie> tomprince's jpeg plugin is using exceptions, i probably can't commit that..
[10:52:21] <fuzzie> pity, because it's so stand-alone
[10:55:55] <wjp> I wonder if libjpeg likes that use of exceptions
[10:56:51] <wjp> hm, but I guess it should be ok since the docs suggest using longjmp from an error callback is ok too
[10:58:51] <fuzzie> i don't know if all the target platforms support exceptions at all
[10:59:16] <Enverex> Can someone point me to the bug-tracker? I couldn't find it from the main Wiki site (only the Howto)
[11:00:20] <wjp> https://sourceforge.net/tracker/?group_id=10122&atid=110122
[11:00:29] <Enverex> Ta
[11:01:40] <fuzzie> be sure to change Status to 'Open' and hit 'filter' if it's showing closed bugs
[11:06:21] <Edheldil> example m,ail sent to gemrb-commits@
[11:06:28] <Enverex> Hrm, I noticed there aren't any bugs open for the "colour icons are too big" and "save game thumbnails overlap their box". I'm guessing people are aware of these
[11:06:44] <fuzzie> which colour icons?
[11:09:06] <Enverex> I'm just getting screenshots now
[11:09:13] <fuzzie> thankyou :)
[11:10:18] <fuzzie> i think thereotically i am busy today, but i imagine i will manage without learning about what the existational and universal quantifiers are..
[11:10:48] <fuzzie> maybe i can commit the pst stuff
[11:16:03] <Enverex> Ok, here are the 5 minor graphical bugs I found a few minutes ago (their individual title is on the image page) -- http://xnode.org/i/?p=92 == http://xnode.org/i/?p=93 == http://xnode.org/i/?p=94 == http://xnode.org/i/?p=95 == http://xnode.org/i/?p=96
[11:16:57] <fuzzie> the last one is bg2?
[11:17:06] <fuzzie> i am forgetful
[11:17:07] <Enverex> They are all BG1
[11:17:12] <fuzzie> ok.
[11:17:27] <fuzzie> i don't think any of those should be hard to fix
[11:17:53] <fuzzie> well, maybe the doll..
[11:18:46] <lynxlynxlynx> link them on the wiki, it'll be faster than sf
[11:18:50] <Enverex> Feel free to use xnode.org/i/ as an image bin if you need somewhere to stick them that isn't riddled with adverts and crap, I made it for such occasions, lol
[11:18:56] <Enverex> lynxlynxlynx, On the "ToDo
[11:18:58] <Enverex> " page?
[11:19:06] <lynxlynxlynx> yes
[11:19:14] <Enverex> Ok, doing that now
[11:20:05] <fuzzie> i wonder what the bug is with the colour picker - is COLGRAD the wrong BAM, or are we using the wrong frames, or just changing the wrong control?
[11:24:51] <Enverex> Hrm, this things slow with the emailing..
[11:25:03] <-- raevol has left IRC (Quit: Leaving.)
[11:25:38] <fuzzie> lynxlynxlynx: add tomprince to AUTHORS?
[11:26:03] <lynxlynxlynx> sure
[11:26:41] <lynxlynxlynx> tomprince: any prefered mail over the one in the commit?
[11:30:44] <Enverex> Is there a problem with the signup on the Wiki? It doesn't seem to send the email
[11:33:15] <lynxlynxlynx> not to my knowledge
[11:33:26] <lynxlynxlynx> but i don't know if anyone tried to signup since the move to sg
[11:33:28] <lynxlynxlynx> sf
[11:33:38] <lynxlynxlynx> do know that sf mails can take a while though
[11:33:46] <lynxlynxlynx> sometimes the commit mails lag for a day
[11:34:16] <fuzzie> maybe manually activate the account?
[11:34:16] <-- Enverex has left IRC (Read error: Connection reset by peer)
[11:34:33] --> Enverex has joined #GemRb
[11:36:18] <lynxlynxlynx> too soon to bother
[11:36:19] <CIA-43> gemrb: 03lynxlupodian * r7503 10/gemrb/trunk/gemrb/plugins/GUIScript/GUIScript.cpp: fixed an optional paramater of GemRB_SetButtonTextColor not being marked as such
[11:38:57] <Enverex> Ouch, ah well, I cant add those to the Wiki till it comes through
[11:40:50] <lynxlynxlynx> remind me later if you still don't get the mail
[11:41:08] <lynxlynxlynx> starting ssh on sf is slow now, since they fire up a vm just for you
[11:47:03] <fuzzie> was simpler in the day when we just all lived on one machine :)
[11:55:25] <CIA-43> gemrb: 03lynxlupodian * r7504 10/gemrb/trunk/AUTHORS: added tomprince to AUTHORS :)
[11:57:46] <lynxlynxlynx> iwd2 is mocking me again
[11:57:52] <fuzzie> oh?
[11:58:09] <lynxlynxlynx> doesn't see a table that's supposed to be loaded
[11:58:38] <lynxlynxlynx> one of those from r7501 and reimporting doesn't help either
[11:59:30] <lynxlynxlynx> it doesn't look any different than let's say bg2, but I guess there's some bad loop and guiinv get's the values before Start.py can set them
[12:04:34] <lynxlynxlynx> ahh
[12:07:29] <CIA-43> gemrb: 03lynxlupodian * r7505 10/gemrb/trunk/gemrb/GUIScripts/ (GUICommon.py iwd2/GUIINV.py pst/GUIINV.py pst/GUIWORLD.py): merged all the encumbrance text setting
[12:08:31] <fuzzie> did the inventory messages get seperated from the other messages yet?
[12:08:41] <fuzzie> i never looked into that, but it seems very strange that they'd get mixed up at all
[12:08:56] * fuzzie runs off for a bit
[12:10:19] <lynxlynxlynx> you mean the bg2 inventory message window?
[12:10:36] <fuzzie> i seem to recall it happening elsewhere too
[12:10:39] <fuzzie> was it just a bug?
[12:11:30] <lynxlynxlynx> i think the message windows share the same text
[12:11:52] <lynxlynxlynx> since the inventory one only has one row and we force a pause, it is not immediately obvious
[12:30:27] <fuzzie> something which needs fixin g?
[12:36:09] <lynxlynxlynx> sure
[12:36:24] <lynxlynxlynx> bbl
[12:36:27] <CIA-43> gemrb: 03lynxlupodian * r7506 10/gemrb/trunk/gemrb/GUIScripts/iwd2/ (9 files): iwd2: kill the extra loading of the races table
[13:14:29] <tomprince> lynxlynxlynx: No, that one is good. Theoretically, it is for life.
[13:17:36] <tomprince> fuzzie: the jpeg stuff could be swtiched to use longjmp. The thing is, error_exit isn't allowed to return.
[13:19:33] <fuzzie> exceptions might be fine too, i don't know.
[13:21:09] <tomprince> We could just even say that we don't support jpeg on platforms that don't support exceptions. The exception never leaves that particular plugin.
[13:22:41] <fuzzie> well, if people can't rely on it, it makes it somewhat less valuable :)
[13:23:46] <fuzzie> but i think it ought to be fine on anything gemrb is going to run on.
[13:24:09] <wjp> I don't think I'd mind using exceptions internal to a plugin as an experiment to see how portable it turns out to be
[13:24:33] <wjp> if we get complaints it's relatively easy to convert it into longjmps
[13:24:58] <fuzzie> there's a scummvm irc log snippet of someone asking there, and it's all platforms we don't care about, except winCE?
[13:26:33] <fuzzie> and i'm pretty sure any modern winCE device is going to have an exceptions-capable compiler, and the platform is kind of dead anyway now..
[13:29:41] <fuzzie> where do i link for libjpeg, ijg.org?
[13:30:58] <wjp> yeah
[13:33:08] <Edheldil> are the exceptions working on Avenger's MSVC6? ;-)
[13:33:15] <fuzzie> we'll see :)
[13:33:30] <fuzzie> i imagine in particular we will hit problems with throwing around char* exceptions, if anything
[13:36:38] <tomprince> Well, doesn't matter what we throw, since there is one throw and one catch. :)
[13:37:19] <wjp> a 'const char*' sounds more correct incidentally (if that works)
[13:37:54] <tomprince> True.
[13:38:14] <fuzzie> i am just testing the cmake side
[13:38:41] <fuzzie> hm, doesn't work.
[13:39:03] <fuzzie> Oh, right, I applied this against svn.
[13:41:41] <fuzzie> tomprince: http://fuzzie.org/nfs/gemrb/cmake_jpeg.txt
[13:54:31] <tomprince> Looks sound.
[13:55:14] <fuzzie> well, I thought you might like to take it for your tree for now.
[14:01:50] <tomprince> Pushed.
[14:03:12] <tomprince> So, are there any other scripting language we want to support, besides python and lua?
[14:05:07] <fuzzie> well, if you're trying to make it general-purpose, you might want to pick something like scheme.
[14:05:14] <Gekz> SCHEME
[14:05:14] <Gekz> NO
[14:05:36] <Edheldil> KISS
[14:05:53] <Gekz> me.
[14:05:56] <Edheldil> beware the feature-bloat
[14:06:04] <fuzzie> all of our code is going to still be in python anyway, and the ScriptEngine interface is trivial.
[14:07:02] <Edheldil> do we I am not sure we want to support Lua ;-)
[14:07:12] <Edheldil> s/do we//
[14:07:37] <tomprince> Well, I think we want to support two, to make sure our interface is sufficently general.
[14:08:22] <fuzzie> Lua is not really sufficiently different from Python to ensure that, though.
[14:08:49] <tomprince> Well, I'll also have a look at what it takes to embedded scheme.
[14:14:15] <fuzzie> do you remember why you changed the pperm values in PLTImporter?
[14:17:01] <tomprince> If I rember correctly, it was because I was inverting the mapping. That we need to map in the opposite direction now.
[14:17:58] <tomprince> So, ppermnew[i]=j iff ppermold[j]=i.
[14:18:57] <fuzzie> aha.
[14:22:00] <fuzzie> I don't understand the check marked with '// Use this?', either.
[14:25:10] <fuzzie> I mean, it seems fine, it just changes behaviour, I think? And I am not awake enough to understand if it's fine.
[14:29:20] <tomprince> Well, it doesn't change behavior, it just moved it around. The code originally used to be in the constructor.
[14:29:45] <tomprince> Although I can't remember what the comment is for. Probably wondering if that case ever happens.
[14:30:37] <tomprince> There is a memset(-1) in GUIScript, which is the only place that function is called.
[14:30:49] <fuzzie> I don't see any 0xFFFFFFFF checks on the PLT codepath in the original code.
[14:34:08] <fuzzie> I mean, not that it seems to make any sense *without* one, but the old code seems to loop with GetPalette and stomp over the palIndices in all cases.
[14:36:09] <tomprince> I think that check might actually be useless.
[14:36:39] <tomprince> As well as the memset in GemRB_SetButtonPLT
[14:37:06] <tomprince> Which would make the array intial useless?
[14:37:49] <fuzzie> That is how I readit.
[14:38:18] <fuzzie> And so I would consider that a bug, fixed by your version..
[14:38:53] <tomprince> Although, if anybody calling button.SetPLT from python, with the first color set to -1 would trigget that check.
[14:40:21] <fuzzie> All the callers do pass the colour parameters, though.
[14:41:10] <tomprince> But is the colour paramater ever -1?
[14:41:28] <fuzzie> Not that I can see. I'm just wondering whether those parameters should be optional.
[14:43:48] <tomprince> They aren't, right now. I suspect that what happened is that they were, at somepoint, and then that was changed, but not all of the code to deal with it was removed.
[14:45:01] --> Maighstir_laptop has joined #GemRb
[14:46:25] <tomprince> intial was added in commit r2864
[14:47:28] <fuzzie> r2860 made the parameters optional.
[14:51:30] <tomprince> Yes, I see. I didn't notice the |.
[15:11:56] <-- Maighstir_laptop has left IRC (Quit: Maighstir_laptop)
[15:14:43] --> maighstir has joined #GemRb
[15:41:22] --> tomprince_loki has joined #GemRb
[15:44:05] <-- barra_library has left IRC (Quit: Verlassend)
[15:49:03] <tomprince_loki> fuzzie: As far as I can tell WeiDU doesn't support ogg or jpeg in biffs. And that seems to be the primary tool used to create them.
[15:50:56] <tomprince_loki> Which makes sense, as there is no tool that knows how to use them in a bif, since we don't.
[15:51:00] <fuzzie> weidu doesn't support anything post-IE,though?
[15:51:42] <tomprince_loki> Is there newer engines that use bif?
[15:52:01] <fuzzie> Yes.
[15:52:07] <tomprince_loki> Ah.
[15:52:11] <tomprince_loki> I wasn't aware of that.
[15:52:39] <tomprince_loki> What else uses it, and is it documented somewhere.
[15:53:29] <tomprince_loki> If we support stuff that newer games put in biffs, we should probably support them in biffs too. Since it isn't hard.
[15:54:10] <fuzzie> All Bioware's games since use it.
[15:54:41] --> kettuz has joined #GemRb
[15:54:53] <fuzzie> Not all of them support third-party content, though, so I don't know how complete the tools are.
[15:55:18] <fuzzie> And I seem to recall someone saying that the released Dragon Age doesn't use them at all.
[15:56:00] <fuzzie> They've certainly fixed some other horrors (eg, 2DA files are finally replaced with something else, even if they are still called 2DAs everywhere).
[15:56:53] <fuzzie> But as I said the other day, I'm not sure anything supports ogg files with a generic id; the uses I've found seem game-specific.
[15:56:59] <fuzzie> So it may not be worthwhile.
[15:58:28] <fuzzie> It is really weird poking around their new games; it's all still the same structure, 'override' folder and TLK files with string ids and all..
[15:59:19] <tomprince_loki> Well, I think it is more or less a sound design.
[15:59:43] <tomprince_loki> Obviously, it is, since they keep using it.
[16:01:41] <tomprince_loki> Do you know if there are any emulators for their newer games? A number of our plugins could be reused for newer games without much change, if they have the same data layout.
[16:03:08] <fuzzie> I think there are already various libraries for at least the Arora Engine versions of the files.
[16:03:38] <fuzzie> Bioware provided file format specifications, so it's a bit easier for them. :)
[16:06:01] <tomprince_loki> I wonder if we could support, for example, the creatures from the Aurora engine. I rember a discussion about how sprite animations was one of the more difficult things to mod.
[16:06:01] <fuzzie> For example, http://git.swordcoast.net/?p=nwn/nwn-lib.git;a=blob;f=lib/nwn/res.rb;h=880f48beef94ea51e49d5286642016ffcbb3c7c7;hb=HEAD is one such; their list has no 'ogg', but does have the 'bik' and 'jpg' entries which are in a couple of the other tools too.
[16:08:39] <fuzzie> (ogg is supposedly 2078, but it's only used officially in The Witcher, I thin?)
[16:09:39] <tomprince_loki> Well, the new plugin system makes it trivially easy to add support.
[16:10:08] <tomprince_loki> It is a one or two liner, just passing that id to the ResourceDesc constructor.
[16:11:13] <fuzzie> The trouble with starting to add things like 3D rendering is that at some point it seems simpler just to write an engine from scratch and copy from gemrb anything useful..
[16:12:37] <tomprince_loki> Well, for the creatures, the IE modding community was talking about extracting creature animations from NWN*. It was just a random thought of something we could support. It does seem a bit crazy to add 3D rendering though.
[16:14:36] <tomprince_loki> Just generating ideas randomly.
[16:14:41] <tomprince_loki> Brainstorming.
[16:14:43] <fuzzie> :)
[16:22:48] --> barra_desktop has joined #GemRb
[16:23:00] --> Avenger has joined #GemRb
[16:23:04] --- ChanServ gives channel operator status to Avenger
[16:23:08] <Avenger> hi
[16:26:08] <fuzzie> aha
[16:26:09] <fuzzie> hello
[16:26:32] <tomprince_loki> Hello.
[16:27:29] <tomprince_loki> I'm am busy moving GetFactoryResource from ResourceMgr to GameData.
[16:27:57] <fuzzie> following in the footsteps of wjp's tidying up :)
[16:28:26] <tomprince_loki> He started moving stuff to GameData then?
[16:28:50] <fuzzie> yes
[16:29:38] <tomprince_loki> That does really belong in core. Especially since I plan to break up KeyImporter into three plugins.
[16:30:35] <fuzzie> well, the searchpath stuff certainly doesn't belong there :|
[16:31:37] <tomprince_loki> Well, I was going to create a plugin to handle searchpaths, which would take a list of ResourceMgr, and be itself a resourcemgr, then have one for handling directories, and one for handling chitin.key+biff.
[16:32:00] <fuzzie> do those really need plugins?
[16:32:27] <tomprince_loki> Well, they all can reasonably present ResourceMgr interfaces.
[16:32:46] <tomprince_loki> Then my thought was that we could add a zip ResourceMgr.
[16:32:54] <tomprince_loki> That could be added to the search path.
[16:33:20] <fuzzie> cute :)
[16:33:50] <tomprince_loki> They don't really need to be in seperate plugins. That just inforces discipline, that the interfaces have to be only the published ones.
[16:42:09] <Avenger> i don't think we need to explode gemrb into smaller pieces, i wish we use the plugins only when replacing one with another by configuration is needed. (optional stuff, driver stuff and file importers)
[16:43:24] <Avenger> i like you removed those CD files, hope this still compiles on msvc6 though
[16:43:48] <Avenger> i bet i will have to work on the windows project files
[16:43:55] <tomprince_loki> Well, like I said, they don't need to be seperate plugins, but I do think they should be seperate classes, that derive from ResourceMgr.
[16:44:19] <tomprince_loki> And the current setup has everything the behaves that way as a plugin.
[16:44:48] <fuzzie> Avenger: the VC++ build should still work, except you have to remove all the deleted files.
[16:44:48] <tomprince_loki> There is an argument to be made that a lot of the stuff we have that are plugins really shouldn't be.
[16:45:04] <Avenger> just don't add interfaces for beauty :) this is a game which needs to be efficient, being nice is secondary ;)
[16:45:05] <fuzzie> I modified the patch to be as small as sensible.
[16:45:24] <Avenger> that's ok, i can easily remove them
[16:45:51] <fuzzie> But the interfaces are useful; BIKPlayer should work by default, now, without having to remove plugins.
[16:46:13] <tomprince_loki> Well, the latest profiling that has been done seems to show that effeceincy isn't much of a concern, except for blitting, and a couple of other things.
[16:46:29] <Avenger> well removing plugins in iwd2's config was not a big deal?
[16:46:52] <Avenger> its fine we don't need now, though
[16:47:13] <tomprince_loki> And apparently, at least one of the changes that I made purely from a beauty standpoint apparently cause a significant speed up.
[16:47:27] <Avenger> that's good, speed ups are welcome ;)
[16:47:34] <Avenger> which one
[16:47:40] <fuzzie> not a committed one
[16:47:52] <Avenger> i just say: lol :)
[16:48:05] <Avenger> but you seen it, Fuzzie?
[16:48:14] <fuzzie> he fixed some of the map stuff to be a direct Bitmap class, and not go via BMPImporter
[16:48:21] <fuzzie> which is one of the things i had on my todo list
[16:48:25] <tomprince_loki> Avenger: If we do apply the rename patch, you might want to hold off updating the MSVC files. Because most of the renames rename files to names that got deleted in the the big patch.
[16:48:29] <Avenger> the searchmap?
[16:48:43] <fuzzie> I guess so?
[16:48:53] <fuzzie> I don't remember, I only checked the speedup.
[16:49:08] <Avenger> ok, tom, just tell me when the big storm is over
[16:49:11] <fuzzie> We should fix the exploration map to not be ieByte* too, probably.
[16:49:30] <fuzzie> But I'd like to know if svn works with MSVC, before making any other changes.
[16:49:48] <tomprince_loki> Chicken and egg. :)
[16:50:03] <Avenger> speedups are always welcome, i'm sure we can do better than the original engine :)
[16:51:11] <Avenger> fuzzie, what about the curved path magic missile?
[16:51:30] <tomprince_loki> And I suspect that a layer of indirection in file loading would be lost in the noise from disk access anyway.
[16:51:50] <Avenger> it doesn't need to be bitexact with the original, just a kind of non-straight line which is based on projectile id
[16:52:04] <fuzzie> Avenger: you are right, i should do that.
[16:52:36] <fuzzie> tomprince: yes, all the plugin/file changes are fine, I think.
[16:53:19] <fuzzie> Well, I don't know about the hash table.
[16:55:18] <fuzzie> Avenger: did we decide whether it should change the drawn position or the actual path?
[16:55:39] <Avenger> the drawn position, based on its distance from the target
[16:55:55] <Avenger> but if you got a better idea now, anything is fine
[16:56:36] <Avenger> i thought it is easier to have the straight line, modified based on the distance and the projectile id. But i never finished it
[17:08:59] <Edheldil> Hi, Avenger
[17:09:13] <Avenger> hi ed
[17:09:56] <Edheldil> I am preparing a switch to Git, as everybody seemed to be ok with that
[17:10:17] <Edheldil> At least I hope they were :)
[17:12:04] <fuzzie> still got to make it work for Avenger :)
[17:12:12] <Avenger> we'll see how that will impact me
[17:12:39] <fuzzie> i think it is no problem changing the drawn position, but i can't try it for a few hours.
[17:13:22] <fuzzie> but finding the right angle for an arc and then calling sin() is the easy bit. :)
[17:15:29] <tomprince_loki> The issue is that only bg1 and bg2 uses that script during chargen, and only
[17:15:43] <tomprince_loki> ignore that. :)
[17:15:53] <tomprince_loki> Avenger: http://git.or.cz/course/svn.html
[17:16:38] <Edheldil> I would rather got rid of SVN after the switch
[17:17:34] <tomprince_loki> No, that is a tutorial for git, for people familiar with svn.
[17:17:38] <Edheldil> as far as I could see, syncing them (even just git->svn) is PITA
[17:17:47] <Edheldil> ah
[17:17:55] <fuzzie> tomprince: not really a very friendly introduction for a Windows user who just wants to treat it like svn :)
[17:20:33] <tomprince_loki> Well, I don't have a windows install anywhere.
[17:20:44] <tomprince_loki> Look, there is http://code.google.com/p/tortoisegit/
[17:20:50] <Edheldil> me neither, hehe
[17:21:16] <tomprince_loki> s/Look/Looking/
[17:22:05] <fuzzie> huh, that looks quite nice.
[17:22:58] <Edheldil> bye, later
[17:23:26] <tomprince_loki> https://git.wiki.kernel.org/index.php/MSysGit:Home
[17:24:30] <fuzzie> The trouble with msysgit is that you start to use it like it's actually git, and then you do something and it spends an hour churning away at a 10-second operation.
[17:24:52] <fuzzie> ok, bbiaw.
[17:26:31] <Avenger> fuzzie: i'm a linux user who wants to treat it like svn
[17:27:07] <Avenger> as in: i can get the changes from the repo, and then i can commit my changes back
[17:27:10] <Avenger> nothing else
[17:28:34] <wjp> that's certainly possible :-)
[17:28:42] <fuzzie> easygit - http://people.gnome.org/~newren/eg/git-for-svn-users.html - is the laziest way for that.
[17:29:04] <tomprince_loki> Well, have a look at that first link I posted.
[17:30:22] <tomprince_loki> git://repo.or.cz/gemrb.git is the url to clone.
[17:31:01] <fuzzie> it really isn't.
[17:31:07] <fuzzie> it doesn't have a master branch.
[17:31:27] <fuzzie> so git rambles at you about unborn things when you clone it.
[17:31:44] <tomprince_loki> Not anymore. :)
[17:31:50] <tomprince_loki> Just fixed it.
[17:32:01] <fuzzie> ok, good :)
[17:32:02] <tomprince_loki> It has the trunk, from when I last updated from svn.
[17:32:21] <tomprince_loki> I made sure to do that before posting a link for someone unfamiliar with git. :)
[17:33:00] <tomprince_loki> Don't want to scare them away right from the get go.
[17:33:37] <fuzzie> that pages rambles for .. four pages? .. before explaining how to commit, bit late not to scare people :p
[17:34:56] <tomprince_loki> Well, hard for me to judge. I think I started using git within a month or two of the very very first version.
[17:35:34] <tomprince_loki> Back when git was just going to be plumming, and other people would write nice frontends for it.
[17:37:47] <Avenger> see you later!
[17:37:49] <-- Avenger has left IRC (Quit: bye!)
[17:39:10] <fuzzie> I found it easy to learn, but I was doing so surrounded by experienced git users, so I could just shout out confused questions and get someone scurrying over.
[17:40:30] <fuzzie> But my experience in trying to explain it to other people is that trying to introduce the concepts first only works if people actually care. Then again, I am not a wonderful explainer.
[17:41:15] <wjp> the (high level) internals are very elegant
[17:41:45] <wjp> (haven't looked under the hood much yet, so can't comment on the low level internals)
[17:42:24] <tomprince_loki> I haven't done any coding for git, but I did follow the mailing list for the first couple of years.
[17:43:49] <tomprince_loki> I got the impression that it is very well coded, and stood the test of time. It wouldn't suprise me if there is quite a bit of very low level code that hasn't been touched for years. I think there may be a little bit of code that dates back to the very first commit.
[17:44:21] <tomprince_loki> The core design is just so simple and obvious, that the code hasn't needed to be changed.
[17:44:24] <tomprince_loki> I think.
[17:50:34] * lynxlynxlynx starts to read the backlog
[18:03:04] <lynxlynxlynx> yeah, the design made for a funny flamewar with the hg folks :)
[18:04:51] <-- tomprince_loki has left IRC (Ping timeout: 240 seconds)
[18:09:52] --> tomprince_loki has joined #GemRb
[18:17:30] --> Edheldil_ has joined #GemRb
[18:36:01] <-- tomprince_loki has left IRC (Ping timeout: 264 seconds)
[18:40:01] <CIA-43> gemrb: 03lynxlupodian * r7507 10/gemrb/trunk/gemrb/GUIScripts/iwd2/ (7 files): iwd2: remove over-loading of classes.2da
[18:42:55] <fuzzie> so is Edheldil going to construct an authorative git repository?
[18:43:40] <fuzzie> because we probably want to rewrite some of the authors, at least..
[18:44:22] <CIA-43> gemrb: 03lynxlupodian * r7508 10/gemrb/trunk/gemrb/GUIScripts/iwd2/Enemy.py: iwd2: remove over-loading of clskills.2da
[18:51:28] <CIA-43> gemrb: 03lynxlupodian * r7509 10/gemrb/trunk/gemrb/GUIScripts/iwd2/Abilities.py: iwd2: remove over-loading of kitlist.2da
[18:55:12] <CIA-43> gemrb: 03lynxlupodian * r7510 10/gemrb/trunk/gemrb/GUIScripts/iwd2/GUIREC.py: iwd2: remove over-loading of xplevel.2da
[18:55:44] <lynxlynxlynx> LoadCommonTables ftw, even though the last three were more or less cosmetic
[19:05:11] <Edheldil_> what authors?
[19:06:54] <fuzzie> Edheldil_: my commits of tomprince's patches.
[19:06:58] --> |Cable| has joined #GemRb
[19:07:23] <fuzzie> In the latest two, '--use-log-author' will fix it, but not before that.
[19:07:57] <lynxlynxlynx> it's not such a big deal
[19:08:18] <lynxlynxlynx> afterall, you had a big role in it too
[19:08:30] <fuzzie> well, he seemed a bit upset about it, and it's easy to patch up
[19:08:53] <lynxlynxlynx> ok
[19:09:20] <lynxlynxlynx> btw, what is the easy way, because i can only think of recommiting
[19:09:40] <lynxlynxlynx> (--amend)
[19:09:59] <fuzzie> well, that is what i'd do, --amend --author=bla
[19:10:33] <fuzzie> but someone elsewhere said there was a cleverer way
[19:11:53] <fuzzie> although the amend+rebase idea is presumably the only actual option
[19:12:40] <lynxlynxlynx> and a lot of branching
[19:14:07] <fuzzie> the automated way is using git filter-branch, i suppose.
[19:16:00] <fuzzie> but maybe tomprince would be fine with just a --use-log-author import.
[19:16:04] <wjp> ideally we'd want to rewrite authors to whatever we'll be using in the future
[19:16:27] <fuzzie> well, the git-svn import can handle the rewrites for svn authors
[19:16:33] <wjp> yes
[19:17:08] --- barra_desktop is now known as barraAway
[19:17:20] <fuzzie> a bit annoying how changing the metadata changes the sha :)
[19:17:57] <wjp> we might also want to change the default git-svn blurb at the end of the commit messages
[19:18:25] <fuzzie> the filter-branch manpage gives an example for that.
[19:19:23] <fuzzie> although if there's going to be a fin al import you might as well drop the whole thing, i guess.
[19:19:32] <fuzzie> i'll leave it in expert hands..
[19:19:37] <Edheldil_> you should not amend or rebase remote repository
[19:19:53] <fuzzie> Edheldil_: this is for during the import.
[19:20:20] <wjp> it would also be nice to clean up the tags
[19:21:14] <Edheldil_> when I switched ie_shell from svn to git, it preserved commit authors and I got rid of git svn metadata
[19:21:31] <wjp> there's some cvs2svn stuff left-over in the tags
[19:23:56] <wjp> hm, only GEMRB_0_3_2 had commits to the tag it seems
[19:24:21] <wjp> I wouldn't mind too much if that was a tag not on any branch
[19:24:40] <fuzzie> in any case, i don't mind at all as long as someone remembers to pass --use-log-author.
[19:25:13] <fuzzie> i don't know if it would be clever to keep revision ids, do we reference them anywhere in comments etc?
[19:25:46] --> raevol has joined #GemRb
[19:25:48] <Edheldil_> the solution to that would be to push up to tomprince's commits and manually fix those
[19:25:56] <lynxlynxlynx> wjp: i agree; even further, i find it silly to make branches at release time (unlike in svn, tags are cheap in git)
[19:26:51] <wjp> well, branches for releases make a lot of sense in some development models
[19:27:09] <Edheldil_> we do not support older versions yet, so making tags is enough, yes
[19:27:26] <wjp> e.g., scummvm branches about a month before release and then has users playtest that branch while development continues on trunk
[19:27:33] <Edheldil_> we can always branch from the tags in git
[19:27:48] <wjp> tags are cheap in svn too, by the way
[19:27:56] <fuzzie> $ grep "[^a]r[0-9][0-9][0-9]" gemrb/plugins/*/*.cpp
[19:27:56] <fuzzie> gemrb/plugins/Core/Spell.cpp: //FIXME (r7193):
[19:28:02] <fuzzie> ^- i guess not worth worrying about :)
[19:28:24] <wjp> bug tracker items might refer to specific revisions?
[19:28:34] <Edheldil_> btw, I am not a git expert :)
[19:30:29] <fuzzie> well, i am going to leave it to others to worry about.
[19:30:31] <fuzzie> must go.
[19:48:49] --> tomprince_loki has joined #GemRb
[19:51:07] <tomprince_loki> Well, I am willing to do the official import, and fix everything up.
[19:51:59] <tomprince_loki> I'm just curious what everybody wants for email on all the commits.
[19:54:29] <fuzzie> Not sure it matters that much, it's clear who we all are.
[19:55:00] <fuzzie> I certainly don't use my sourceforge address for commits, but it's harmless having it on the old ones..
[19:56:27] <-- tomprince_loki has left IRC (Ping timeout: 260 seconds)
[19:56:41] <wjp> same for me
[19:56:58] <wjp> I'd be willing to do an official conversion
[19:59:17] --> tomprince_loki has joined #GemRb
[20:04:05] <tomprince_loki> No, it won't hurt. If that is the way people want to go, that is easy enough. If people want something else, that is jus as easy.
[20:06:06] <lynxlynxlynx> history is not so important for emails and you can always claim everything via the mailmap
[20:06:39] <lynxlynxlynx> one thing i'm especially looking forward to, is the removal of $Id$ :)
[20:12:23] <-- tomprince_loki has left IRC (Ping timeout: 246 seconds)
[20:19:44] <Edheldil_> hehe
[20:20:24] <Edheldil_> I am willing to do the official import as well :-P. But now I want to get ciabot mailer working
[20:21:28] <lynxlynxlynx> divide and conquer
[20:22:05] <Edheldil_> I will start with the other moduls and leave gemrb as the last one
[20:22:17] <Edheldil_> but first the ciabot
[20:26:51] <-- maighstir has left IRC (Quit: buh-bye)
[20:33:59] --> barra_away has joined #GemRb
[20:36:30] <-- barraAway has left IRC (Ping timeout: 240 seconds)
[20:52:15] <CIA-43> GemRB: 03edheldil 07master * rfa1e52d917fb 10/infinity/stream.py: Another insignificant change
[20:53:01] <Lightkey> O_O
[20:53:41] <Lightkey> I should lay down the crack pipe, I see colours
[20:54:10] <Edheldil_> I apologize for test messages, others will follow
[21:15:46] <-- cfchris6 has left IRC (Ping timeout: 246 seconds)
[21:17:40] --> cfchris6 has joined #GemRb
[21:24:33] <CIA-43> GemRB/ie_shell: 03edheldil 07master * rfa1e52d917fb 10/infinity/stream.py: Another insignificant change
[21:27:46] <lynxlynxlynx> the gemrb module name is gemrb?
[21:28:01] <lynxlynxlynx> i think you can kill the prefix there then
[21:37:03] <Edheldil_> I don't know, but possibly it's empty
[21:37:32] <Edheldil_> or do you mean in SVN/git?
[21:38:15] <tomprince> I think the CIA prefix.
[21:38:24] <tomprince> So ie_shell, instead of GemRB/ie_shell?
[21:39:18] <lynxlynxlynx> yes
[21:40:55] <tomprince> http://pastebin.ca/1851866
[21:42:03] <fuzzie> seems to make the most sense :)
[21:42:23] <fuzzie> looks like i have no time to finish anything tonight..
[21:42:57] <tomprince> Well, I am running an import with that right now, but if somebody wants a real email address there ...
[21:45:05] <lynxlynxlynx> looks fine to me
[21:45:13] <lynxlynxlynx> finally the wolf will be gone
[21:46:37] <tomprince> Wolf?
[21:46:42] <fuzzie> tomprince: do you think applying your configure.in vorbis change along with your other build system changes would make sense?
[21:46:54] <lynxlynxlynx> lupo, lupus
[21:48:08] <tomprince> Well, you should really have the Makefile.am changes too. Adding the if VORBIS.
[21:48:10] <fuzzie> i am figuring there's no point applying more patches to svn now, if there is actually going to be a move
[21:49:28] <CIA-43> GemRB: 03edheldil 07master * rfa1e52d917fb 10ie_shell/infinity/stream.py: Another insignificant change
[21:50:17] <fuzzie> but once we move to git, it still seems to make more sense bundling the build system changes together, rather than trying to make them at the same time as rearranging things.
[21:51:20] <tomprince> That is fine.
[21:51:21] <fuzzie> (am seeing three things here: your proposed configure.in changes, the vorbis changes, and the ones from the 'Rework plugin interface'?)
[21:52:16] <tomprince> Well, the configure changes a posted are actually sitting in a couple of commits that I haven't commited and pushed yet.
[21:52:38] <tomprince> Some of them are already included in the patches on or.cz
[21:52:46] <fuzzie> well, sorry, this is what I mean: it seems to make sense to just bundle it all as one commit, from here.
[21:54:49] <fuzzie> but if it's a problem then they can be seperate, just pondering.
[21:55:04] <fuzzie> in the absence of my having anything except a web browser :)
[21:55:17] <tomprince> Well, some of the changes to configure.in go with changes to other files too.
[22:00:10] <fuzzie> why are GetWidth() and GetHeight() ever not unsigned?
[22:01:50] <tomprince> I don't think there is any reason.
[22:04:34] <tomprince> They were added as int originally.
[22:12:51] <Genraznx> how goes work on GemRB?
[22:16:04] <lynxlynxlynx> good
[22:18:28] <Genraznx> what are you folks working on anywho?
[22:19:13] <lynxlynxlynx> currently much of it goes into cleanup
[22:19:51] <Genraznx> Like the original code?
[22:20:07] <Genraznx> or just revising your own?
[22:20:21] <lynxlynxlynx> 100% of 28 users recommend this project :)
[22:20:37] <lynxlynxlynx> our own is the original code
[22:23:19] <lynxlynxlynx> so, wjp, you'll handle the main module migration?
[22:23:59] <lynxlynxlynx> ed is on the others and avenger is not interested, which leaves just you ;)
[22:26:17] <lynxlynxlynx> Enverex: got the mail yet?
[22:26:25] <Edheldil_> I have +- completed the cia mailer, now I have to integrate them into post-receive
[22:36:40] <lynxlynxlynx> nice playmode patch tomprince, I'll check if all the guiscript stuff is ok once the migration is over
[22:37:37] <lynxlynxlynx> some parts looked iffy in the diff, but it's hard to judge from six lines of context ;)
[22:38:47] <Enverex> lynxlynxlynx, Nope
[22:39:30] <lynxlynxlynx> ok, i'll ssh in
[22:40:20] <lynxlynxlynx> doh, i can do it from the web
[22:42:45] <tomprince> I'll push the svn-import to or.cz later tonight. Or directly to SF.
[22:42:50] <lynxlynxlynx> new password generated and the mail sending supposedly worked too
[22:46:50] <-- kettuz has left IRC (Quit: Leaving)
[22:48:52] <Enverex> lynxlynxlynx, Nope, let me check the mail log and make sure it's not hitting a wall somewhere
[22:54:21] <Edheldil_> what mail sending?
[23:00:04] <Enverex> Ok, added those bugs + image links
[23:05:21] <lynxlynxlynx> wiki thing
[23:10:44] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:26:15] <CIA-43> GemRB: 03edheldil * rb9a9a01c90e4 10ie_shell/infinity/formats/baf.py: Another test commit
[23:27:12] <Edheldil_> ok, simultaneous mail to cia and mailing list is working
[23:28:47] <Edheldil_> I wonder if I have wiki access
[23:29:49] <-- Enverex has left IRC (Read error: Connection reset by peer)
[23:30:38] --> Enverex has joined #GemRb
[23:30:57] <fuzzie> hm
[23:31:11] <fuzzie> is the first magic missile curved or fired ina straight line?
[23:50:09] --> ratpor has joined #GemRb
[23:52:03] <Genraznx> probably in a straight line
[23:52:54] <fuzzie> not getting them to curve at all right now, is very odd.