#gemrb@irc.freenode.net logs for 7 Jan 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[00:08:41] --> traveler__ has joined #gemrb
[00:09:35] <traveler__> brada: currently don't have bg2 data extracted. as i never finished bg2, i fear i will get caught up ;)
[00:10:53] <-- traveler has left IRC (Ping timeout: 245 seconds)
[00:11:13] <brada> meh. lynx says its fine :p
[00:14:39] <traveler__> well, that potentially saves me from myself ;)
[00:17:12] <lynxlynxlynx> it was fine before, don't know about this commit ;P
[00:18:41] <lynxlynxlynx> not to say that i expect problems
[00:39:59] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[01:01:32] <-- Cable__ has left IRC (Ping timeout: 255 seconds)
[01:04:03] --> edheldil__ has joined #gemrb
[01:13:52] --> Cable__ has joined #gemrb
[01:36:13] --> CamDawg has joined #gemrb
[01:47:28] --> Liam has joined #gemrb
[01:47:31] --- Liam is now known as Kaeloree
[01:49:01] <Kaeloree> just wanted to pop in and say congrats on the awesome work you guys have done :) can't wait to check out the next version of gemrb!
[01:51:06] <-- Kaeloree has left IRC (Client Quit)
[01:56:53] <brada> fyi we only have one remaining use of audio::IsSpeaking()
[01:57:06] <brada> and there is a todo associated with it
[01:58:07] <brada> looks like we should be doing something else instead and I'd like to fix it
[02:00:20] <brada> maybe now that we can queue speech this is trivial, but i honestly dont quite know what is going on there
[02:01:04] <brada> i guess in this case it is wanting to queue text and not audio :(
[02:06:19] <-- edheldil__ has left IRC (Ping timeout: 246 seconds)
[02:14:57] <brada> i dont know about scripting but if the time to display is actually available just tell me how to access it and ill do it
[02:26:09] <brada> i imagine something like this should work: http://paste.debian.net/222378/
[02:26:12] <Pepelka> debian Pastezone
[02:29:09] <brada> or maybe the scriptable whould know the duration
[02:29:24] <brada> either way should be fairly simple *if* the duration is available
[02:31:14] <brada> ah CurrentActionTicks sounds about right
[03:13:11] <-- CamDawg has left IRC (Quit: Leaving.)
[03:49:27] <Pepelka> [commit] Brad Allred: TTF: return NULL if FT_Set_Pixel_Sizes fails http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=b501d54d680ed2dd5f5d111c9d35bc3f7c789e40
[03:49:28] <Pepelka> [commit] Brad Allred: Fix some dead stores http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=31117411a518eb15af2cadc587acbfbf16486442
[03:49:29] <Pepelka> [commit] Brad Allred: BAMImporter: plug potential leak in GetFramePixels http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=f1ec40b177bb182da23319b2395baa402e2ac8da
[03:49:31] <Pepelka> [commit] Brad Allred: Actor: prevent potential invalid memory access http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=d37a753fb1044015cb830eec4cadde185ee0c160
[04:21:42] <Pepelka> [commit] Brad Allred: CREImporter: no need to die. just return NULL http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=41368a40c8dd6b14d477d53a91c02409737e0fbe
[04:21:44] <Pepelka> [commit] Brad Allred: assign values to prevent warnings about garbage values http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=fe7408bf285d53a6c181ae53f563a84107d30f19
[05:00:11] <Pepelka> [commit] Brad Allred: move variables into the scope they are used in. http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=830d80272ab45951d4940b33df7d824e7b3ef367
[05:05:11] <-- brada has left IRC (Quit: brada)
[07:01:30] --> edheldil__ has joined #gemrb
[07:16:08] --> chiv has joined #gemrb
[07:18:54] <-- edheldil__ has left IRC (Ping timeout: 240 seconds)
[07:34:54] <-- edheldil_ has left IRC (Ping timeout: 240 seconds)
[07:47:02] --> avenger has joined #gemrb
[07:48:03] <avenger> this commit seems to break LoadWindow: http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=blobdiff;f=gemrb/core/Interface.cpp;h=8dc7d828d5c1271ae75ea49a160d1ab0ffb9fd02;hp=528ba2b3503e68dac181f8df8119b4ffeab4146a;hb=31117411a518eb15af2cadc587acbfbf16486442;hpb=b501d54d680ed2dd5f5d111c9d35bc3f7c789e40
[07:48:05] <Pepelka> SourceForge - gemrb/gemrb/blobdiff - gemrb/core/Interface.cpp
[07:53:16] <-- avenger has left IRC (Quit: Page closed)
[08:24:15] --> WingedHussar has joined #gemrb
[09:01:27] --> edheldil__ has joined #gemrb
[09:10:16] --> lynxlynxlynx has joined #gemrb
[09:10:16] <-- lynxlynxlynx has left IRC (Changing host)
[09:10:16] --> lynxlynxlynx has joined #gemrb
[09:10:16] --- ChanServ gives channel operator status to lynxlynxlynx
[09:20:07] --> miha has joined #gemrb
[10:20:51] <chiv> would it be any help if I brought pst/guicommonwindows more in line with the other one? I am trying to track down why portrait dragging is so annoying and I notice it lags behind a bit
[10:21:41] <lynxlynxlynx> if you can do it in a controlled, step by step manner — reviewable, then definitely yes
[10:22:06] <lynxlynxlynx> many small patches vs a patch bomb
[10:22:29] <chiv> Ok, I am willing to do that
[10:22:37] <lynxlynxlynx> excellent :)
[10:22:38] <chiv> but... there will be many patches,
[10:22:46] <chiv> its very different
[10:23:01] <lynxlynxlynx> you can't make it the same, the control ids and many strings are different
[10:23:26] <lynxlynxlynx> not sure what would affect dragging speed in there
[10:24:02] <chiv> this is something that I have been wanting to ask - wouldn't it be better to have the same code, but just store the strings in a table?
[10:28:54] <fuzzie> you can't make it identical anyway
[10:28:55] <lynxlynxlynx> an often floated idea
[10:29:34] <lynxlynxlynx> you'd need the same for the control ids and would end up with another abstraction layer that can't work with special casing
[10:29:44] <lynxlynxlynx> i mean without
[10:30:06] <lynxlynxlynx> guistore is probably a better example, it's practically identical
[10:30:14] <lynxlynxlynx> yet they had to move all the strrefs
[10:30:43] <lynxlynxlynx> i seriously can''t understand why they did that - it created extra work for them too
[10:30:48] <fuzzie> They might not have done so.
[10:30:59] <fuzzie> You have to remember that they forked pst way before bg1 was released.
[10:31:16] <fuzzie> I wouldn't be surprised if this stuff is just #defined in a header or something with an automated tool, anyway.
[10:31:47] <chiv> looking at guistore, what is wrong with that?
[10:32:02] <fuzzie> Is pst's store stuff all handled in the C++ side?
[10:32:04] <lynxlynxlynx> it's also lagging behind
[10:32:33] <fuzzie> I guess it is. That's convenient.
[10:32:34] <chiv> i mean the multi-game guistore
[10:32:55] <lynxlynxlynx> a few years back i fixed it by copying the bg2 version over and changing all the ids, but since then the now shared guistore received new love
[10:33:17] <fuzzie> well, the shared guistore already has game version checks all over
[10:34:00] <lynxlynxlynx> diff -u gemrb/GUIScripts/GUISTORE.py gemrb/GUIScripts/pst/GUISTORE.py
[10:34:03] <fuzzie> so it seems pretty harmless to add some more
[10:34:14] <fuzzie> also the string ids seem to often match bg2 anyway..?
[10:34:18] <lynxlynxlynx> ids and strrefs are different all the time
[10:35:57] <chiv> but i mean apart from the fact that there are no comments to say what is what, the storetips / roomtypes array seems to make sense
[10:36:13] <fuzzie> chiv: yes, but if you run the diff which lynx suggested, you can see that a whole bunch of other things are different
[10:36:32] <fuzzie> and if you put things like strings into an array, you just add a confusing layer of abstraction
[10:36:57] <lynxlynxlynx> pst guiscripts are bitrotten
[10:37:06] <fuzzie> the diff is pretty nice though
[10:37:19] <fuzzie> it looks like you can just go by that and add a bunch of 'if' checks for PST?
[10:38:00] <chiv> confusing? hey if I can understand it...
[10:38:03] <chiv> :)
[10:38:31] <fuzzie> well, sure, if you can add a non-confusing layer of abstraction..
[10:39:32] <chiv> i miss switches
[10:40:09] <fuzzie> just as long as modders don't have to keep flipping back and forth between yet more things when they try and work out how to change things, would maybe be a reasonable rule of thumb
[10:40:23] <fuzzie> but you can see all the existing GameIsBG2 etc checks in the shared guistore..
[10:40:39] <chiv> yeah, I am all for simplification too
[10:40:59] <fuzzie> poor lynx has been trying to finish this merging for ages
[10:41:04] <fuzzie> so help would be most welcome
[10:41:30] <chiv> i would consider futzing with guiscripts a waste of all of your talents...
[10:42:04] <chiv> I can at least bring them more in line anyway
[10:44:15] <lynxlynxlynx> testable chunks though please, that messagewindow dump is unreviewable
[10:44:23] <chiv> what about using simple defs, eg instead of button.settext(1337), button.settext(STR_LEET)
[10:44:43] <lynxlynxlynx> you still have to set them somewhere
[10:45:10] <lynxlynxlynx> could be a separate module, sounds cleaner than tens of new ifdefs
[10:46:13] <lynxlynxlynx> merging guiscripts was one of the first things i started doing when i became a contributor and now i have it on the todo to merge another file each release
[10:46:53] <lynxlynxlynx> only two remain that have more than 100 shared lines as reported by snake food :)
[10:47:11] <chiv> which are those?
[10:48:58] <lynxlynxlynx> actually no, those are occurence counts and the tool is simian
[10:49:12] <chiv> actually the more I think about it, the more I like it - you would have one file per game called Strings.py, where you could see every string the game uses with a plain comment...
[10:52:22] <lynxlynxlynx> i'd just start with whatever pst needs
[10:52:39] <lynxlynxlynx> i like the fact that strrefs are only one grep away usually
[10:56:49] <chiv> Anyway, I won't do any more patches like that message window, I was only trying to prove to myself it could be done easily...
[10:59:02] <lynxlynxlynx> ok :)
[10:59:16] <chiv> but hey, you already have ie_sounds etc, whats one more :) ?
[11:00:02] <lynxlynxlynx> in general, my merging approach is as follows
[11:00:43] <lynxlynxlynx> take the two gametypes that have the most similar file, update both of them to be identical and still work as before
[11:00:44] <fuzzie> chiv: the ie_sounds constants are purely internal, i.e. their value is irrelevant
[11:01:01] <chiv> I just saw...
[11:01:11] <fuzzie> ie_stats is better example maybe
[11:01:13] <lynxlynxlynx> if anything is found in one or the other to be a real bugfix, it's separated out of the merge
[11:01:49] <lynxlynxlynx> once they are the same, they can be removed - one moved to GUIScripts/ toplevel
[11:02:03] <lynxlynxlynx> at this point all the other games are still unaffected, since their file takes precedence
[11:02:30] <lynxlynxlynx> then you just repeat the same with each gametype, possibly quitting in rage after seeing the diff vs pst
[11:02:41] <chiv> thats the process I get - but I am not sure about sending patches
[11:02:52] <chiv> lol
[11:03:05] <chiv> yes, pst is a darling
[11:03:14] <lynxlynxlynx> kdiff3, meld or similar are pretty handy for this
[11:03:40] <chiv> no i mean, what is an example of a patch you like to recieve?
[11:03:46] <lynxlynxlynx> re patches, since you're on git it is easiest for everyone but you if you use git's mechanisms
[11:04:36] <lynxlynxlynx> eg. a separate patch for the merge and a separate one for the move/removal, otherwise the changes are lost in the noise
[11:05:52] <lynxlynxlynx> so let's say N patches syncing two scripts, 1 move/delete, goto start
[11:16:59] <chiv> bear with me, I am just not used to sharing my hacks in a big project, but can a patch be too small? eg http://pastebin.ca/2300083 is a patch to stop the load confirm dialog forgetting to toggle off. but in my own version, I have also renamed it to toggleloadmsgwindow to be more accurate, implemented a function in another file to center all windows in any res. it starts to get a bit hard to decide when to stop...
[11:17:01] <Pepelka> pastebin - loadmsgwindowfix - post number 2300083
[11:18:15] <lynxlynxlynx> there is no such thing as a too small patch
[11:18:27] <chiv> right, no i am on your wavelength ;)
[11:18:32] <chiv> *now
[11:18:36] <lynxlynxlynx> you should see the linux or the wine process :D
[11:18:48] <edheldil> unless it is TOO small, of course ;-)
[11:19:49] <lynxlynxlynx> creating small changesets is something you need to keep in mind while hacking already
[11:20:17] <lynxlynxlynx> making real commits as soon as possible is a good way to avoid needing to cherrypick changes from a larger set later
[11:20:18] <chiv> so if I removed the comment, that's a step too far?
[11:20:41] <lynxlynxlynx> not if it's a useless comment
[11:21:45] <edheldil> splitting "tmp=a; a=b; b=tmp;" to two or more obviously is
[11:22:37] <chiv> its not really a useless comment... OpenOptionsWindow() to close it ? hehe
[11:23:41] <lynxlynxlynx> that's how it's wired, so i don't see why the comment should be removed
[11:25:57] <chiv> so anyway in summary, if I am going to do this merge, i should probably make patches function by function?
[11:27:12] <fuzzie> well, the most important thing is that the diff is readable
[11:27:32] <chiv> ok
[11:27:35] <fuzzie> there's not much use doing it function-by-function if no-one can review the result
[11:28:53] <lynxlynxlynx> have you used git as a developer before?
[11:29:29] <chiv> nah, hey im not even a developer, thats why I ask so many stupid questions :)
[11:31:51] <lynxlynxlynx> why would those two be mutually exclusive?
[11:32:25] <chiv> The last programming education I had was on an ACORN using LOGO, everything I actually know I have learned in my spare time
[11:32:51] <chiv> and you do have a fair point...
[11:34:03] <miha> source control can be used for any text, like wiki
[11:38:37] <chiv> I think there is a great irony in the fact that my good friend is a training software developer, went to college, got a degree etc, cushy job, owns a boat etc, but in contrast to me he wont touch anything that isn't flash
[11:39:02] <miha> flash is dying platform. apple killed it.
[11:40:06] <chiv> its not even as if he isnt capable of moving on, he just doesn't need to because his work don't force him to
[11:40:42] <miha> soon, very soon :)
[11:41:40] <chiv> one day, the era of my leverage over the boat will begin....
[11:42:06] <miha> besides, college teaches very little about software development
[11:42:32] <miha> "you already know how to code, so practise at home some more"..
[11:43:32] <miha> for example, there is quite a lot of math theory about 2D and 3D graphics. but very little coding :)
[11:45:26] <lynxlynxlynx> i'm also self taught and i sometimes miss annoying repetitive homework i'd get in real courses; eg. untargetted practice made me learn the memory model very slowly and i'm still not comfortable with it
[11:47:38] <miha> lynxlynxlynx: what memory model? pointers at programming 1. abstract theory at computer arhitecture? some malloc at "system programming"? very little low level stuff. rest everyone learns on his/her own? :)
[11:48:28] <miha> Fry: "I'm a certified college dropout."
[11:48:28] <miha> Leela: "Please... Everyone knows 20th century colleges were basically expensive day care centers."
[11:48:31] <miha> Farnsworth: "That's true. By current academic standards, you're merely a high school dropout."
[11:48:34] <miha> Fry: "What?! That's not fair. I deserve the same respect any other college dropout gets. By God, I'm going to enroll here at Mars University and drop out all over again!"
[11:48:34] <lynxlynxlynx> i don't know what they teach, but judging at least from my foreign colleagues, it's not so useless
[11:48:48] <lynxlynxlynx> haha
[11:49:09] <fuzzie> well, what gets taught in CS differs wildly between universities, I think
[11:49:29] <fuzzie> or even between different years at the same uni :)
[11:50:32] <lynxlynxlynx> i know i'm a bit unfair to myself, since often the comparisons are to older seasoned veterans, but still, targetted grind can get you far
[11:53:38] <miha> lynxlynxlynx: it's not job of undergrad CS course to teach you "memory model".
[11:53:53] <fuzzie> from my point of view, I'd much rather work with someone who knows how to use source control and do code reviews and etc, than someone who feels comfortable with memory model :)
[11:54:00] <miha> neither is to teach you "cryptography"
[11:54:18] <fuzzie> miha: The memory model is something which you should know after completing undergrad CS though.
[11:54:36] <miha> fuzzie: what memory model? memory is "allocated" by OS, somehow.
[11:54:40] <fuzzie> Since otherwise you can't have effectively understood operating systems, computer architecture nor compiler theory.
[11:55:02] <miha> fuzzie: without garbage collectory you need to free it yourself. that's it
[11:55:06] <miha> collector
[11:55:35] <miha> oh and there is cache, memory, disk. only cache is close to CPU speed. local code.
[11:56:08] <miha> fast code tries not to cause cache to swap all the time
[11:56:42] <miha> anything more is OS detail, not taught in CS course
[11:56:45] <lynxlynxlynx> luckily here we don't need to care about cache bounds
[11:56:55] <fuzzie> miha: What do you think *is* taught in CS? :)
[11:57:22] <miha> some programming, databases, graphics, you know normal apprentice-coder at business app :)
[11:57:42] <fuzzie> so, you mean, a not-CS course :)
[11:58:31] <chiv> i wish I knew what was, then I would have a convenient checklist...
[11:58:47] <miha> fuzzie: CS course != hacker course
[11:58:54] <fuzzie> miha: CS course != how to program :P
[11:59:14] <miha> fuzzie: exactly. it's about how to "manage software development project" :D
[11:59:21] <fuzzie> which is nothing to do with CS at all :)
[11:59:43] <chiv> why is that any surprise though :)?
[11:59:54] <fuzzie> well, there are plenty of CS courses which actually teach CS
[12:00:32] <miha> well, if you study biology or medicine, it's not theory about how to sort samples at lab
[12:00:33] <fuzzie> and how stacks and heaps work is pretty important for various CS courses, for example, and requires an understanding of memory models
[12:01:12] <miha> fuzzie: well we did stacks and heaps, at data structures and at algorithms. is that memory model?
[12:01:19] <miha> also all sorts of trees
[12:01:31] <miha> well heap is a tree, i think :)
[12:01:32] <fuzzie> no, no, heaps as in memory model heaps.
[12:01:38] <fuzzie> different thing :p
[12:01:52] <miha> isn't it a tree? :)
[12:01:59] <fuzzie> not in the memory model :)
[12:02:08] <miha> oh, stack on the other side of memory?
[12:02:08] <miha> :D
[12:02:12] * fuzzie tips miha gently
[12:03:10] <miha> well i know function arguments are on call stack, and thus destroyed when you exit function. you use heap for long-term objects. or in java, for all objects.
[12:03:22] <miha> struct on stack, class on heap in c++?
[12:03:31] <fuzzie> but if you design a simple processor, or write a simple compiler, or do paging/etc for operating systems, you need to know how they work
[12:03:57] <miha> fuzzie: but you dont learn to do "paging for OS" in undergrad course. perhaps as your thesis?
[12:04:00] <fuzzie> and those are all generally considered essential CS courses (although the extent of compiler theory is argued over, probably only parser theory is essential)
[12:04:10] <miha> we did learn parsers yes
[12:04:15] <fuzzie> miha: I think every OS course I've seen has a mandatory component around paging or similar.
[12:04:31] <miha> fuzzie: how old are you?
[12:04:36] <fuzzie> Too old. :P
[12:04:51] <miha> fuzzie: since 32 bit and 64 bit OS, we dont switch memory to tape all that often
[12:04:59] <fuzzie> heh
[12:05:10] <miha> and if we do, it's "virtual memory" done by OS
[12:05:11] <fuzzie> so you have no paging on your system, right? :)
[12:05:39] <miha> sure there is, but is implicit to me. cache<=>memory<=>swap
[12:05:48] <miha> i never did that myself :)
[12:05:56] <fuzzie> right, but that just means you didn't learn computer architecture nor operating systems :)
[12:06:22] <miha> i learned a lot of theory at computer arhitecture and operating systems. but we didnt code that.
[12:06:37] <miha> i just know there is some gnome that does it
[12:06:38] <fuzzie> sure, you don't have to *code* it, but the theory is pretty essential.
[12:07:44] <miha> for lynxlynxlynx http://www.emka.si/arhitektura-in-organizacija-racunalniskih-sistemov/PR/34422,562 if you read whole book, you'll learn more than most cover at class :)
[12:07:46] <Pepelka> Arhitektura in organizacija računalniških sistemov: Dušan Kodak: Knjiga | Emka.si
[12:08:08] <miha> of computer arhitecture. i read whole book multiple times cause i didnt know only first two chapters apply for exam :p
[12:08:35] <chiv> you can never have too much knowledge
[12:08:37] <fuzzie> well, the usual standard textbook for CA is Hennessy&Patterson, with the MIPS architecture etc
[12:09:29] --> CamDawg has joined #gemrb
[12:09:33] <miha> kodak wrote about some imaginary RISC system in detail and briefly about most popular RISC and CISC
[12:09:45] <miha> also floating point IEEE standard
[12:09:46] <miha> and such
[12:10:09] <fuzzie> my old version is really lacking on the memory model front though, the cache stuff is hard to understand without
[12:10:10] <miha> fuzzie: i doubt we missed much since it was about most detail book we did study :D
[12:10:42] <fuzzie> it has some 10 pages on paging
[12:10:45] <fuzzie> no tapes involved ;)
[12:10:47] <miha> as i said, i was supposed to only read 2 chapters. i skipped lectures so i didnt know :)
[12:11:05] <miha> and then i read it about 5 times all over caused it scared me
[12:11:16] <fuzzie> I actually still haven't passed Computer Architecture here.
[12:11:36] <fuzzie> There's an exam on the 15th, that sounds like fun.
[12:12:12] <miha> fuzzie: i do agree computer arhitecture is good to know. not just for programming. also helps to understand what they want to sell you at your new computer
[12:13:06] <miha> people are lost now, it used to be simple: more Mhz, better
[12:13:32] <miha> Mhz
[12:44:15] <traveler__> MHz?
[12:44:46] <traveler__> only when speaking of early pentium /586 clones
[12:45:34] <miha> up to dual cores
[12:48:46] <traveler__> don't remember amd64?
[12:49:10] <traveler__> that is, overblown p4 clock with lonng pipelines
[12:49:41] <traveler__> which just didn't perform well in real world usually
[12:50:08] <traveler__> apart from video encoding iirc
[12:50:11] <miha> or pentium 4 hyper-threading, dual core without dual cores :)
[12:51:20] <traveler__> well yes, different thing but too very questionable if this was ever usable in real world
[12:51:55] <traveler__> people are still usually disable this iirc if stuck with p4/ht
[12:58:14] <traveler__> and cisc cpu were never directly comparable with other arch. clocks but that's kind of obvious, but you had some apple ppc/sun workstations at the time of original pentium too I think.
[13:06:55] <traveler__> dis intel finally admit total failure with epic/itanium? they were stuck with ancient technology and only supplying as replacement parts for vendors like hp last time i checked
[13:07:00] <traveler__> *did
[13:09:18] --> duckpunch has joined #gemrb
[13:32:41] <-- CamDawg has left IRC (Quit: Leaving.)
[13:56:02] --> CamDawg has joined #gemrb
[13:56:47] <-- CamDawg has left #gemrb
[14:41:31] <-- miha has left IRC (Quit: Lost terminal)
[14:48:21] <traveler__> 2012 February: Court papers were released from a case between HP and Oracle Corporation that gave insight to the fact that HP was paying Intel $690 million to keep Itanium on life support
[14:48:22] <traveler__> hah
[15:01:16] <chiv> is http://pastebin.ca/2300134 a more agreeable sort of size for a patch?
[15:01:19] <Pepelka> pastebin - Something - post number 2300134
[15:08:10] <edheldil> miha: we had to tinker with educational language and operating system in my CS courses :)
[15:08:26] <edheldil> and memory model is pretty important if you want to do RE :)
[15:09:55] --> kida has joined #gemrb
[15:11:27] <Pepelka> [commit] Jaka Kranjc: admin: added details on how to generate a gource video http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=03adea72de3cfcf2a53a770c12c8d6bdc8fbf8a8
[15:11:28] <Pepelka> [commit] chiv: pst: close also the subwindow when loading a game http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=19abf0caa40d62548536997b8ce41660a53fce64
[15:11:29] <Pepelka> [commit] Jaka Kranjc: only display the spell failure roll if there is a chance to fail http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=1308c9c245a2eb920da2d8fcff92b675de818912
[15:12:01] <-- kida has left IRC (Client Quit)
[15:12:12] <edheldil> speaking of that, we had to calculate electric properties of a transistor from slab of silicon with given sizes and enrichment, but I skipped that to save my sanity
[15:12:14] --> kida has joined #gemrb
[15:12:21] <-- kida has left IRC (Client Quit)
[15:12:38] --> kida has joined #gemrb
[15:13:01] <-- kida_laptop has left IRC (Remote host closed the connection)
[15:14:01] <lynxlynxlynx> chiv: that's fine, but you'll need to repeat the process, since the two files are out of sync after your changes too
[15:17:13] <edheldil> looking at 1308c9c245a2eb920da2d8fcff92b675de818912 (only display the spell failure roll ...), we will have to move all this into a rule plugin sometime in the future
[15:17:54] <chiv> I'll stick to about that size then, aiming to make them identical
[15:18:52] <lynxlynxlynx> sometime
[15:20:53] <edheldil> yes, but if would be better to have our simple rules than relying on ADnD for a free game
[15:22:10] <chiv> I'm not convinced the madness of dnd rules isn't a scheme to sell rulebooks...
[15:23:28] <chiv> along the lines of 'they'll never make a decent phone' theory
[15:26:15] <lynxlynxlynx> of course, but it's a bit early imho, since the combat system is not final yet
[15:26:28] <lynxlynxlynx> at least two more intrusions for iwd2 are needed
[15:26:54] <lynxlynxlynx> adding callbacks everywhere will make more sense when we are stable
[15:30:14] <chiv> what's the realistic chances of being able to make a mod that converts up to 3e rules like icewind gate?
[15:30:58] <chiv> i suppose the obvious answer is it depends on my ability...
[15:34:25] --> CamDawg has joined #gemrb
[15:36:27] <lynxlynxlynx> how would it be different from iwg2?
[15:36:41] <lynxlynxlynx> and the chances are definitely better with gemrb
[15:37:07] <lynxlynxlynx> some things would get fixed by just a flick of the switch
[15:38:51] <chiv> it would be different from iwg2 in that it would not get stalled by broken hardcoded crap....
[15:39:26] <chiv> actually i see what you are saying, i guess someone could start from where iwg2 left off
[15:40:52] <lynxlynxlynx> yeah, a review would be needed and then it can be started
[15:41:15] <lynxlynxlynx> i guess it already does the tedius job of converting items, spells, scripts, whatever to the new formats
[15:41:32] <chiv> it probably wouldnt be that far off, i seem to remember it was only evil party joining that was breaking the whole thing
[15:45:59] <chiv> actually i can't wait to see what you guys come up with for game rules...
[15:46:46] <lynxlynxlynx> i doubt we'd go invent anything
[15:47:19] <lynxlynxlynx> we "just" need to make it possible, then whatever people want to develop on top of it can easily spring up
[15:47:37] <lynxlynxlynx> using pathfinder or some other freely usable system would be easiest
[15:47:58] <chiv> I have always thought the dnd rules were written while on acid though...
[15:54:17] <chiv> for instance, its not ok for an priest to stab someone, but its fine for a priest to club things to death
[15:55:47] <lynxlynxlynx> and you can backstab with qstaffs but not with flails
[15:55:49] <CamDawg> Meh, it's just a classic case of a balance change being justified post-hoc. Most 2e item restrictions are equally senseless.
[15:56:51] <chiv> its fine in a living room, because you tell the rules where to shove it, that all falls apart with computer games though...
[16:03:40] <edheldil> well, that DnD is not very computer friendly is given. Neither WotC is :)
[16:06:01] --> brada has joined #gemrb
[16:08:18] <CamDawg> Yeah, there's a reason why I openly mock folks who come to the Fixpack/BGEE bug forums and want PnP implementations. It's house rules, at best, just as any 2e game would be.
[16:08:57] <chiv> actually some pnp implementations are an improvement, eg mage spellcasting in armor
[16:09:01] <CamDawg> And lynxlynxlynx, I'd be happy to host an IWD2 Fixpack, but not develop one. I've never even finished it :)
[16:09:09] <chiv> as opposed to NO YOU CANT!
[16:09:49] <CamDawg> Yeah, but then we're cutting into downloads of BG2 Tweaks and hence my fame, power, and chick-magnet-ness decrease.
[16:10:01] <CamDawg> (From 0 to, well, still 0, but still...)
[16:10:07] <chiv> heh
[16:11:04] <lynxlynxlynx> CamDawg: mostly gauging community interest there, plus lazy-checking what else exist
[16:13:05] <CamDawg> I love the idea of G3 being the go-to place for Fixpacks--and truth be told I do like building them--it's just that I have zero knowledge of the game and that's unlikely to change any time soon.
[16:14:54] <chiv> g3 is pretty much my first stop after installation normally
[16:15:13] <chiv> the game is filthy and dirty without the fixpacks
[16:16:23] <CamDawg> Yeah, there are times I kinda sit back and go: wait, Fixpack touches 4800+ files? Holy crap.
[16:17:52] <chiv> it really should be an officially endorsed patch
[16:19:30] --> miha has joined #gemrb
[16:22:27] <lynxlynxlynx> which reminds me, i haven't finished 10pp either
[16:22:52] <lynxlynxlynx> so much stuff to do :)
[16:24:34] <CamDawg> And the stuff turning up in BGEE makes me convinced that BG may have been worse. So many quests rely on bad assumptions and are easily broken.
[16:25:51] <brada> how is it possible for http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=blobdiff;f=gemrb/core/Interface.cpp;h=8dc7d828d5c1271ae75ea49a160d1ab0ffb9fd02;hp=528ba2b3503e68dac181f8df8119b4ffeab4146a;hb=31117411a518eb15af2cadc587acbfbf16486442;hpb=b501d54d680ed2dd5f5d111c9d35bc3f7c789e40 to break anything?
[16:25:53] <Pepelka> SourceForge - gemrb/gemrb/blobdiff - gemrb/core/Interface.cpp
[16:26:10] <brada> local var is assigned to and never read....
[16:26:29] <lynxlynxlynx> i didn't notice any problems
[16:26:50] <lynxlynxlynx> which reminds me, i was in a middle of a debugging session
[16:36:27] <wjp> brada: maybe Avenger just thought that was a patch for the identical snippet in LoadWindow? (Where slot _is_ used after that assignment)
[16:37:28] <brada> thats what happens when you dont use {} for all your ifs!
[16:37:32] <brada> :p
[16:38:52] <fuzzie> the GemRB_GetStoreItem in that patch is not necessarily a no-op..
[16:39:00] <-- kida has left IRC (Quit: Leaving)
[16:41:10] <fuzzie> presumably it's harmless but it's not so nice to change behaviour with commit message "fix some dead stores" :P
[16:42:48] --> kida has joined #gemrb
[16:44:01] <wjp> hm, different behaviour for amounts >= 2^31 I suppose
[16:44:16] <fuzzie> mmhm
[16:46:01] <fuzzie> you all have way too much time, there's no way I can review all these commits :/
[16:46:27] <lynxlynxlynx> i have an idea!
[16:46:29] <fuzzie> nice work on many fronts though.
[16:46:53] <lynxlynxlynx> look at leaveare instead ;)
[16:47:09] <wjp> great to see so many commits recently :-)
[16:47:10] <fuzzie> it isn't particularly complicated I thought
[16:47:28] <lynxlynxlynx> you have it all in your head
[16:48:22] <lynxlynxlynx> 1472 Commits: Up +20 (1%) from previous 12 months <-- december saved us
[16:48:53] <lynxlynxlynx> 21 Contributors: Up +9 (75%) from previous 12 months <-- much more important trend
[16:50:06] <wjp> hm, looking at those, isn't avenger's email address invalid?
[16:50:21] <fuzzie> well, you can make a big improvement to leavearea by just removing the EveryoneStopped check from NIDSpecial2
[16:50:35] <brada> wjp: i believe it is
[16:50:50] <brada> i know people have complained about not being able to send him mail before
[16:51:26] <fuzzie> there's a lot more to be done for a proper fix, but I think just that is worth a try first
[16:51:55] <lynxlynxlynx> the sf one?
[16:52:10] <fuzzie> it should be @users.sourceforge.net presumably
[16:52:21] <lynxlynxlynx> fuzzie: i doubt it solves the mayor's house issue
[16:52:28] <fuzzie> lynxlynxlynx: what is the mayor's house issue? :)
[16:52:33] <lynxlynxlynx> small door, but you can get very near, yet still not travel
[16:52:45] <fuzzie> i think it should
[16:52:46] <lynxlynxlynx> not even with jumping into the region
[16:52:51] <fuzzie> because after that it just checks EveryoneNearPoint
[16:52:51] <lynxlynxlynx> could be something else though
[16:52:59] <fuzzie> oh no, it checks the actor pos. argh :P
[16:53:23] <fuzzie> but, I guess if there's a region then NIDSpecial2 isn't involved
[16:53:32] <fuzzie> anyway I have to go cook. bother me in an hour if possible?
[16:54:41] <lynxlynxlynx> hmm, noticed this now: Position: 0.0 <--
[16:55:04] <lynxlynxlynx> sure, easily ;P
[16:56:16] <lynxlynxlynx> brada: ah, in the git log; yes - AUTHORS contains the right one
[16:57:59] <lynxlynxlynx> i'll fix it
[17:00:24] <edheldil> ah, forum is up
[17:02:19] <lynxlynxlynx> i can only fix it for the shortlog, git log ignores the mailmap by default %an vs %aN
[17:05:10] <lynxlynxlynx> git log --pretty="%aE %aN" # <--
[17:05:53] <Pepelka> [commit] Jaka Kranjc: mailmap: fixed avenger's sf email http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=351ae28c1e7bc3a647d42b42ffe8b9f4e1f15bcb
[17:06:03] <brada> lynx: id like to fix the todo in GameScript::DisplayStringWait
[17:06:26] <brada> i monologued about it earlier while you all slept
[17:06:59] <lynxlynxlynx> i read it in the morning, don't remember much though
[17:07:54] <lynxlynxlynx> ah, i guess this is responsible for cutscenes to be faster if there's no audio - NullSound probably doesn't implement IsSpeaking
[17:11:03] <lynxlynxlynx> i'd store it in int1param, since it is unused
[17:11:09] <lynxlynxlynx> no need to use statics
[17:11:26] <lynxlynxlynx> *int2param
[17:13:09] <brada> where do i get the value from in the first place? i know nothing at all about how scripting works
[17:13:17] <lynxlynxlynx> i don't see any uses where int1param is set, where did you get it from?
[17:13:33] <brada> that was psudo code
[17:13:53] <brada> i just assumed that since the comment was refrencing the length then it must exist
[17:13:53] <lynxlynxlynx> ok
[17:14:12] <brada> then there is a field in scriptable for something along those lines too
[17:14:16] <lynxlynxlynx> all these have strrefs with sounds attached, so that's where to get the duration
[17:15:24] <lynxlynxlynx> eg DisplayStringWait(Myself,2834) // ~No matter, they will only prove a slight delay.~ [DPJON162]
[17:15:35] <lynxlynxlynx> DPJON162 being the sound resource
[17:16:15] <lynxlynxlynx> we have code for this in core, IE_STR_SPEECH maybe
[17:16:37] <brada> ill look into it
[17:20:34] <lynxlynxlynx> it's tied in with the text code, so you'll likely have to duplicate or split part of it
[17:22:43] <traveler__> re: avenger mail should be fine, he has been blocking mail not from sf though...
[17:25:43] <traveler__> my mail on avenger_teambg@users.sourceforge.net went through
[17:25:51] <traveler__> after a while
[17:38:11] <-- brada has left IRC (Quit: brada)
[17:48:30] --> Cuvieronius has joined #gemrb
[17:56:52] <lynxlynxlynx> fuzzie: time's up. Press 0 to wake up, 1 to snooze
[18:02:21] <chiv> how does this look to you? http://pastebin.ca/2300179
[18:02:23] <Pepelka> pastebin - option strings - post number 2300179
[18:02:33] <chiv> by the way, is there a better way than pastebin...
[18:04:06] <tomprince> github pull request?
[18:06:14] <lynxlynxlynx> the other file is missing, plus RepositionWindow is not related
[18:06:38] <chiv> sorry, just asking your opinion
[18:06:42] <chiv> ignore that
[18:07:02] <chiv> i mean from line 70
[18:09:32] <chiv> because setting up the strings like that also makes SetupIWD2MenuControls redundant
[18:10:20] <chiv> arg, i mean SetupIWD2WindowControls
[18:10:31] <chiv> I am too hungry to think straight...
[18:11:12] --> brada has joined #gemrb
[18:13:27] <brada> i defiantly prefer the dict approach for strings compared to using raw strrefs
[18:13:45] <brada> you of course realize you will now have to do the same for everything :p
[18:13:51] <brada> mwahahaha
[18:14:24] <chiv> thats the thing though, its exactly what I want to do...
[18:14:49] <chiv> but its sort of pointless unless its right by you guys :)
[18:15:13] <lynxlynxlynx> looks fine from there on
[18:18:58] <brada> chiv: re pastebin: either what tomprince said or upload a patch file
[18:19:25] <chiv> so how do I go about the git way?
[18:19:40] <brada> github? there is a pull request button on the top
[18:20:14] <brada> that is great for review, but for us to actually pull it in and use it i think we need to use a github remote
[18:20:25] <brada> but we should have one we can use iirc
[18:22:30] <chiv> well, apparently, I am already registered... I wonder if I was drunk at the time...
[18:22:33] <brada> heh
[18:22:47] <brada> just for the gemrb github project
[18:23:02] <fuzzie> lynxlynxlynx: still snoozing a bit, apparently
[18:23:12] <brada> *fork
[18:24:03] <brada> then on your computer add a new remote for gemrb
[18:24:15] <brada> dont know the commands since i use a git gui
[18:24:34] <lynxlynxlynx> might be over your head if you haven't used git much
[18:24:51] <lynxlynxlynx> especially once i ask you to fix a commit that is not the latest
[18:25:26] <lynxlynxlynx> pastebin works for me now
[18:25:59] <brada> chiv seems interested in learning tho
[18:26:50] <chiv> well i already use git command line
[18:28:06] <lynxlynxlynx> ok, let's start it slowly then
[18:28:34] <lynxlynxlynx> set up .git/config to use the name and email address (preferably the same as in our repo)
[18:28:55] <lynxlynxlynx> when you're done with a part, use git commit -p # to select parts of your changes to add to the commit
[18:28:59] <brada> he can just use the same clone he already has
[18:29:05] <lynxlynxlynx> sure
[18:29:24] <lynxlynxlynx> ... then add the commit message and save
[18:29:40] <lynxlynxlynx> then git log -p # to see what you did and verify
[18:30:03] <lynxlynxlynx> then you can use git format-patch to dump it as a nice enriched patch with all the commit info
[18:31:12] <lynxlynxlynx> once you master git log -p # with 's', reordering and amending, you're well prepared to use your own branch
[18:31:28] <lynxlynxlynx> you can always fix the latest commit by using git commit --amend
[18:31:40] <lynxlynxlynx> all the same commandline options apply
[18:32:31] <lynxlynxlynx> to fix something earlier than HEAD, a rebase is needed, but let's not go there yet
[18:34:31] <chiv> i feel like I may have just had an epiphany
[18:35:19] <chiv> i have definitely been doing it wrong
[18:39:44] <-- kida has left IRC (Ping timeout: 260 seconds)
[18:42:00] <brada> the way i did it i only had to add a new remote for the github fork to my existing sourceforge clone. then do a checkout for github/HEAD
[18:42:24] <lynxlynxlynx> ok, found the problem in creimporter
[18:43:38] <brada> chiv: for something like what you are doing with many comits over many files over many days you should probably create a new branch
[18:44:16] <brada> that way you dont have to worry about us changing things underneeth you until you are done
[18:44:30] --> Yoshimo has joined #gemrb
[18:44:38] <lynxlynxlynx> no sense in doing that until he has the main workflow in his blood
[18:46:06] <chiv> yeah I started keeping seperate branches
[19:11:49] --> traveler has joined #gemrb
[19:14:14] <chiv> well I have all that figured out, but what I really wanted to know was where's the best place to send a patch :)
[19:14:38] <-- traveler__ has left IRC (Ping timeout: 245 seconds)
[19:16:05] <lynxlynxlynx> github is fine
[19:23:59] <brada> yeah if you have github set up then just use a pull request
[19:24:16] <brada> much easier to review there
[19:24:54] <chiv> i am just about to test it
[19:25:18] <tomprince> (And maybe mention it here, since we don't currently use github very actively)
[19:26:34] <lynxlynxlynx> not sure if a pull request on github would work
[19:26:35] <chiv> eh, something wierd happened
[19:28:17] <chiv> its somehow shoved in a load of other stuff in the pull request
[19:31:44] <chiv> I think I know what happened, is the github behind by a few commits?
[19:32:13] <lynxlynxlynx> quite possibly
[19:37:00] <tomprince> I think if you refresh your pull request, they will go away.
[19:40:57] <brada> chiv: thats why i said to use github/head for your work if possible
[19:41:27] <brada> but maybe they will go away like tom said
[19:47:04] <chiv> i dont seem to be able to do anything but add even more to it...
[19:48:32] --> edheldil_ has joined #gemrb
[19:50:59] <lynxlynxlynx> you need to wait for the mirror to sync
[19:54:59] <brada> you can avoid it in the future tho by working off github/HEAD instead of sourceforge/HEAD
[19:55:31] <chiv> thats what i will do
[19:56:27] <brada> and when you make a new branch branch from github/HEAD
[19:56:33] <lynxlynxlynx> that works if nobody else is touching the same code
[19:57:02] <lynxlynxlynx> you could rebase, but it's dangerous
[19:57:11] <brada> i would avoid that
[19:58:03] <brada> if they dont disappear when the mirror syncs you have some options
[19:58:31] <brada> if you dont have many commits just do a soft reset to github/master and delete your branch locally and remotely
[19:58:37] <brada> then do a new branch and re-commit
[19:59:05] <brada> or you can do more complicated things like do a new branch and cherry pick the commits from your old branch
[19:59:52] <brada> if you do the soft reset it is safe for you to discard the "changes" that arent yours
[20:00:06] <brada> they are already commited
[20:04:40] <chiv> I could just start again, doesnt really bother me
[20:04:52] <brada> well you already have a patch file on pastebin
[20:05:05] <brada> so yeah you can just delete your branch and start a new one
[20:05:10] <brada> then apply your patch to it
[20:05:59] <brada> do a git push after deleting the branch to kill the remote branch
[20:06:05] <lynxlynxlynx> ouch
[20:06:21] <brada> its not that painful :p
[20:06:34] <brada> rebasing can be much more difficult
[20:06:40] <lynxlynxlynx> how so?
[20:07:21] <brada> if there are conflicts
[20:07:29] <chiv> how do I just delete my fork off github, it seems like it is idiot proofed...
[20:07:44] <brada> when i was new to git it wasnt apparent how to resolve them
[20:07:47] <lynxlynxlynx> git push hub :branch
[20:08:02] <lynxlynxlynx> he shouldn't get any now anyway
[20:08:06] <brada> where 'hub' is the name of your github remote
[20:08:28] <brada> yeah he shouldnt i was talking in general
[20:10:28] <chiv> ah found it, the Danger Zone (tm)
[20:12:13] <lynxlynxlynx> you can just do git rebase --interactive HEAD~50 # guessing 50 commits back was already in sync on both repos
[20:12:30] <lynxlynxlynx> and then delete all the new commits that caused the out-of-sync
[20:12:37] <lynxlynxlynx> then just push again
[20:12:58] <brada> wouldnt git reset --soft github/HEAD do that too?
[20:13:08] <brada> then discard all the changes he doesnt need for his work
[20:13:50] <brada> i dont know if thats the proper command since i use a gui :p
[20:16:18] <-- traveler has left IRC (Ping timeout: 245 seconds)
[20:16:47] <lynxlynxlynx> reset would require recommitting
[20:17:30] <brada> no because he should branch off of github/HEAD which he is resetting to
[20:17:30] <chiv> thats not exactly terrible, since i made a patch anyway
[20:17:39] <brada> chiv: right
[20:17:49] <chiv> anyway its all a learning excercise
[20:17:52] <brada> yes
[20:18:09] <brada> i gues lynx mean you would have to recommit your own work
[20:18:40] <brada> but like we were saying just apply your own patch to the new branch and you should be good to go
[20:18:54] <chiv> i have always had trouble with commitment
[20:19:06] <chiv> badum tish
[20:19:12] <brada> heh
[20:19:43] <lynxlynxlynx> fuzzie: rrrring
[20:23:02] <-- duckpunch has left IRC (Quit: leaving)
[20:24:25] --> duckpunch has joined #gemrb
[20:34:29] <edheldil_> chiv: why do you use dict and not 'constants'? e.g. st. like Button.SetText (STR_GAME) ?
[20:34:32] <fuzzie> sorry. buried in uni stuff. can you describe testcase (or point me to log)?
[20:35:53] <chiv> just seemed right at the time
[20:37:12] <chiv> any reason not to?
[20:37:47] <Pepelka> [commit] Jaka Kranjc: SpellEntry::AddLevel: added a log message if we encounter duplicate entries http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=f2de40d27641681663378185f9c7344a94818964
[20:37:48] <Pepelka> [commit] Jaka Kranjc: CREImporter: translate the kit values when looking up iwd2 spells http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=743c8846148ebddf4c7b42c4350a2ae9078c0abf
[20:37:49] <Pepelka> [commit] Jaka Kranjc: CREImporter::GetActorIWD2: put our money where our mouth is - assert http://gemrb.git.sourceforge.net/git/gitweb.cgi?p=gemrb/gemrb;a=commitdiff;h=8618a91bec631e7f6fbb0d4c7ac64be666dc81fc
[20:38:10] <lynxlynxlynx> i kind of like it, since it's a bit more explicit without_extra_long_names
[20:38:33] <chiv> that was my thinking, it just sort of looks nice
[20:38:43] <lynxlynxlynx> fuzzie: for the normal leavearea stuff or something else?
[20:38:54] <fuzzie> for normal stuff.
[20:40:39] <lynxlynxlynx> one is that it is tricky to get in/out of narrow passages to travel areas, but that could also be fixed by just increasing the radius (iwd1 kuldahar eastern stone bridge is an example)
[20:41:30] <lynxlynxlynx> one other thing, i sometimes had to select a single char to make the group travel, even though everyone was still already and relatively near
[20:42:03] <lynxlynxlynx> without jumping, party travel can be very annoying
[20:42:28] <fuzzie> well, not all the games did that
[20:43:14] <lynxlynxlynx> did what?
[20:43:48] <chiv> pst you can travel with a single character if your buddies are on the other side of the map
[20:43:59] <chiv> but that is sort of meh
[20:44:32] <chiv> I vote to port 'you must gather your party' sound into pst
[20:45:03] <chiv> only make it louder and repeat more often
[20:45:39] <fuzzie> did jumping/
[20:45:50] <fuzzie> sorry, on tablet atm.
[20:45:51] <lynxlynxlynx> ah, i meant player induced
[20:45:56] <fuzzie> makes me terse.
[20:45:57] <lynxlynxlynx> eg. me
[20:46:04] <fuzzie> oh, i see
[20:46:15] <fuzzie> yes, i agree that it's v.bad atm
[20:46:30] <fuzzie> it just sounded like you had a case where it was plain broken
[20:46:38] <fuzzie> which would be nice to test againsy
[20:46:45] <lynxlynxlynx> the iwd2 one is probably different
[20:47:00] <lynxlynxlynx> for some reason it has Pos(0,0), so likely a range check fails
[20:47:50] <lynxlynxlynx> leavearea is just annoying in general in all games and you know what's broken and where, so that's a better start
[20:48:54] <fuzzie> yes.
[20:51:35] <-- miha has left IRC (Quit: Lost terminal)
[21:08:47] * lynxlynxlynx summons avenger
[21:09:40] <chiv> wild surge: you have been buried under a pile of watermelons
[21:20:50] <edheldil_> grrr, dltcep's are editing is ... braindead
[21:28:36] <lynxlynxlynx> silly how it keeps popping up the wed
[21:28:46] <edheldil_> yeah
[21:29:13] <edheldil_> or rather, it's forgetting its tileset
[21:29:45] --> traveler has joined #gemrb
[21:35:17] <lynxlynxlynx> fuzzie: the actual reason you can't go through that iwd2 door is indeed distance related, but not how i expected — TrapLaunch is out of the bbox and ends up being too far away
[21:35:27] <-- Yoshimo has left IRC (Quit: Yoshimo)
[21:36:20] <lynxlynxlynx> the actor misses the outline check for a small margin
[21:37:58] <lynxlynxlynx> TalkPos is also 0,0, so indeed a true distance check vs the outline or bbox is needed
[21:40:57] <lynxlynxlynx> actually the actor is in the bbox of the outline, so we could use that as a cheap way out too
[21:45:13] <lynxlynxlynx> http://sprunge.us/WWaD?diff fixes it in that manner
[21:56:01] <-- Cuvieronius has left IRC ()
[22:05:34] <-- CamDawg has left #gemrb
[22:19:51] <-- brada has left IRC (Read error: Connection reset by peer)
[22:21:20] --> brada has joined #gemrb
[22:39:56] <-- Coriander has left IRC (Read error: Connection reset by peer)
[22:41:16] --> Coriander has joined #gemrb
[23:04:39] --- ermo^ is now known as ermo
[23:40:16] <-- brada has left IRC (Quit: brada)