#gemrb@irc.freenode.net logs for 6 Nov 2014 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage

[01:21:44] <-- raevol has left IRC (Quit: Leaving)
[01:24:22] --> brada has joined #gemrb
[01:47:12] <brada> im having trouble seeing how removing that list sort “broke” anything...
[01:47:23] <brada> the movie lists are not supposed to be sorted either
[01:47:51] <brada> they are supposed to appear in “chronological” order
[01:50:02] <brada> we do get some unwanted python “none”s at the end of the script list tho
[01:51:16] <-- brada has left IRC (Quit: brada)
[02:19:16] --> brada has joined #gemrb
[03:33:25] <brada> does anybody care to propose a better solution: http://pastebin.com/kyyHhJEu
[03:33:26] <Pepelka90> [Diff] diff --git a/gemrb/core/Palette.h b/gemrb/core/Palette.h index a06f1bc..0343cc5 - Pastebin.com
[03:33:39] <brada> preferably it wont me much more work :)
[03:35:08] <brada> the Palette semantics unfortunately prohibit constness :/
[03:37:23] <brada> also no sure how you are supposed to squelch leak warnings with statics like that. so yeah, hopefully a better way to share palettes.
[04:00:55] <-- brada has left IRC (Quit: brada)
[04:11:37] --> brada has joined #gemrb
[04:13:15] --> Eli2 has joined #gemrb
[04:15:32] <-- Eli2_ has left IRC (Ping timeout: 250 seconds)
[05:04:15] <-- brada has left IRC (Quit: brada)
[06:30:47] <-- Lightkey has left IRC (Ping timeout: 265 seconds)
[06:43:25] --> Lightkey has joined #gemrb
[08:32:09] --> edheldil has joined #gemrb
[08:32:09] --- ChanServ gives channel operator status to edheldil
[09:38:23] --> lynxlynxlynx has joined #gemrb
[09:38:23] <-- lynxlynxlynx has left IRC (Changing host)
[09:38:23] --> lynxlynxlynx has joined #gemrb
[09:38:23] --- ChanServ gives channel operator status to lynxlynxlynx
[10:04:39] <lynxlynxlynx> brad: so i went ahead an looked up what is sorted in master for you: character, charsound and portrait lists
[11:55:20] <-- Textmode has left IRC (Quit: "It was one dev, naked in a room with a carton of cigarettes, a thermos full of coffee and bourbon, and all his summoned angels.")
[12:50:58] <-- edheldil has left IRC (Ping timeout: 255 seconds)
[13:01:17] --> edheldil has joined #gemrb
[13:01:17] --- ChanServ gives channel operator status to edheldil
[14:22:24] --> brada has joined #gemrb
[14:24:49] <brada> lynx: those lists are all built in c++ and are still sorted
[14:30:17] <lynxlynxlynx> cool
[14:31:00] <lynxlynxlynx> so it seems you just made a circle
[14:45:56] <-- brada has left IRC (Quit: brada)
[15:06:25] <-- lynxlynxlynx has left IRC (Ping timeout: 255 seconds)
[15:20:21] --> brada has joined #gemrb
[15:20:32] <brada> a circle?
[15:34:03] --> lynxlynxlynx has joined #gemrb
[15:34:04] <-- lynxlynxlynx has left IRC (Changing host)
[15:34:04] --> lynxlynxlynx has joined #gemrb
[15:34:04] --- ChanServ gives channel operator status to lynxlynxlynx
[15:58:23] <lynxlynxlynx> lunch, then time to revisit the other branch
[16:58:56] <brada> there isnt supposed to be a newline after the “title” on pst journal entries btw
[16:59:04] <brada> i thought i already pushed a fix for that anyway
[17:00:08] <brada> oh that is iwd journal
[17:00:10] <brada> nevermind :)
[17:03:01] --> Marduk has joined #gemrb
[17:03:25] --- Marduk is now known as Guest42343
[17:04:01] <-- Guest42343 has left #gemrb
[17:05:13] --> Morgoth_deb has joined #gemrb
[17:16:23] --> jillsmitt has joined #gemrb
[17:16:26] <jillsmitt> hello
[17:17:21] <lynxlynxlynx> oj
[17:33:51] <brada> i am able to see favorites in iwd2. is there some special case this braks for?
[17:35:03] <lynxlynxlynx> dunno, it's a pretty fresh gemrb save
[17:35:35] <brada> also im still unable to figure out how to get verbal actions to work at all in pst
[17:35:43] <brada> the guiopt surely doesnt work
[17:35:58] <lynxlynxlynx> it was another settext error, so it could be a remnant of the delimited_str foo
[17:36:10] <lynxlynxlynx> first one fails and the rest doesn't matter in the window
[17:36:47] <lynxlynxlynx> re vbs: no idea what went wrong anyway
[17:37:14] <lynxlynxlynx> i'll bisect it if needed, but there's other stuff to do anyway
[17:37:47] <jillsmitt> i was unable to find any centos 7 rpm package of gemrb and now compiling it from source(
[17:37:50] <brada> it would still be nice to know how to enable those in pst
[17:38:17] <brada> or rather fix the guiopt so that the correct setting is used?
[17:39:44] <brada> the opengl driver is unfortunately not very usable
[17:42:26] <brada> we can probably just comment out that rudamentary screen clipping i jammed in there for the short term since the font no longer attempts to draw outside the screen clip
[17:42:41] <brada> that would (probably) fix the worst of its glitchyness
[17:43:12] <lynxlynxlynx> what does guiopt have to do with vbs (besides being able to tune verbosity)?
[17:43:44] <brada> i thought the problem was that i couldnt enable them. is the problem that we just dont support them at all for pst?
[17:44:06] <lynxlynxlynx> no, it's a crash
[17:44:16] <brada> right a crash that i cant reproduce
[17:44:32] <brada> and the fact that the sound feedback doesnt work in pst
[17:44:45] <lynxlynxlynx> i can, but only on the new branch
[17:44:51] <lynxlynxlynx> but it's odddd
[17:45:02] <brada> and im wondering why it doesnt work since it works fine for other games
[17:47:12] <lynxlynxlynx> i think i even checked with valgrind
[17:47:33] <lynxlynxlynx> looking with dltcep or ni would work too
[17:48:59] <brada> also wondering if maybe i dont understand the mapnote thing. i can create a custom note, save the game, then exit, and reload and the note is still there. I definately didnt notice any gui errors.
[17:49:11] <brada> that was BG2 tho
[17:50:01] <lynxlynxlynx> i was trying in bg2 too
[17:50:10] <lynxlynxlynx> tob
[17:50:31] <lynxlynxlynx> try in the pocket plane if you have any saves
[17:50:37] <lynxlynxlynx> or perhaps it was only hell
[17:51:37] <brada> i regretfully never got around to playing tob :(
[17:52:15] <brada> tho a gui error is certainly unrelated to the original worry
[18:04:24] <jillsmitt> guys, i never played BG or something, it will be first time
[18:04:41] <jillsmitt> description is nice and i want to try
[18:04:52] <brada> its a great game :)
[18:05:22] <brada> after this font/text business im gonna get the EE edition to work in gemrb
[18:10:44] <lynxlynxlynx> jillsmitt: try with the git version, it's better
[18:11:56] <Morgoth_deb> is EE natively for linux?
[18:14:40] <lynxlynxlynx> at least iwdee
[18:17:18] <jillsmitt> lynxlynxlynx, i download and compile 0.8.1, is it good?
[18:17:47] <lynxlynxlynx> a few more bugs
[18:18:03] <lynxlynxlynx> a silly one in bg2, but it's circumventable
[18:27:50] <jillsmitt> lynxlynxlynx, is there version without bugs?
[18:28:36] <lynxlynxlynx> haha
[18:29:18] <lynxlynxlynx> there're probably bugs even in /bin/true
[18:40:16] <jillsmitt> maybe proprietary systems does not have bugs?
[18:42:20] <brada> ^ seriously?
[18:42:44] <Morgoth_deb> what do you mean by "proprietary os"?
[18:42:59] <brada> im currently working on the *worst* code i have ever seen and you’d better believe it is proprietary
[18:44:03] <kujeger> Morgoth_deb: EE is native on linux for IWD and BG. BG2 is not, currently.
[18:44:45] <Morgoth_deb> through steam or gog?
[18:44:49] <Morgoth_deb> both of them?
[18:44:54] <kujeger> at least on steam. GOG only has IWD right now
[18:45:53] <kujeger> I would assume BGEE will be for linux on GOG soonish, and BG2EE on both steam and GOG (maybe with the fabled 1.3 patch)
[18:45:57] <Morgoth_deb> jillsmitt: you mean windows doesn't have bugs...? :D
[18:47:37] <Morgoth_deb> kujeger: maybe they will also remake PST on linux soon?
[18:48:52] <kujeger> Morgoth_deb: PST is by far the "strangest" of the IE engine games, so I would assume it would be the last to be given the EE treatment, after IWD2
[18:49:13] <kujeger> (all speculation of course)
[18:49:37] <kujeger> in any case, wine handles all the IE games very well -- better than windows
[18:49:49] <kujeger> in most cases
[18:50:40] <Morgoth_deb> without BSOD every few hours of gameplay? ^^
[18:52:19] <kujeger> heh.. in my experience, the biggest issue is GPU drivers not handling the old ways of doing things
[18:52:29] <kujeger> nvidia seems to be the biggest offender
[18:54:18] <Morgoth_deb> true. I've got FreeBSD, Linux and Windows on different quite new computers and bg1 runs smoothly only on linux and BSD through wine or gemrb
[18:54:30] <Morgoth_deb> bg2 is ok on windows
[18:54:49] <Morgoth_deb> but when it comes to a bit older PC, bg1 runs well :)
[18:55:23] <Morgoth_deb> *on windows
[18:55:51] <Morgoth_deb> so... bg1 isn't in good relationships with new GPU's
[19:13:31] <-- Morgoth_deb has left IRC (Quit: Page closed)
[19:29:35] --> Morgoth_deb has joined #gemrb
[19:35:28] <jillsmitt> Morgoth_deb, and Mac?
[19:35:40] <jillsmitt> Morgoth_deb, i never used this 2
[19:36:05] <Morgoth_deb> me too
[19:36:13] <Morgoth_deb> im not keen on apple software
[19:36:49] <jillsmitt> ever used Windows?
[19:37:48] <Morgoth_deb> for fiveteen years then I switched to linux
[19:37:54] <Morgoth_deb> and bsd
[19:41:49] <Morgoth_deb> *fifteen arr
[20:05:07] <lynxlynxlynx> interesting, the gl code doesn't compile on font-rewrite
[20:08:04] <brada> it does for me
[20:10:00] <lynxlynxlynx> found the problem already
[20:10:06] <lynxlynxlynx> silly diagnostic
[20:10:28] <lynxlynxlynx> still see a performance drop
[20:11:09] <brada> somebody that actually know what they are doing needs to deal with that
[20:11:13] <lynxlynxlynx> intro screen was practically the same, while loaded game at atleast 50% less
[20:11:35] <brada> it cant help that we are creating so many duplicate palettes there
[20:11:40] <brada> gl driver those are all textures
[20:12:13] <lynxlynxlynx> ... however, it is marginally faster than plain sdl2
[20:13:56] <brada> the performance of plain sdl is better for me than on master… but open gl is the same
[20:14:52] <lynxlynxlynx> oh and of course DEBUG_FONT was one, so some cycles were wasted on the rects
[20:15:44] <brada> oh, yeah, we can probably turn that off now if all the asserts and things are fixed
[20:17:28] <lynxlynxlynx> ==18253== Invalid read of size 2
[20:17:28] <lynxlynxlynx> ==18253== at 0x4F330E9: GemRB::Font::CreateAliasForChar(unsigned short, unsigned short) (Font.cpp:210)
[20:17:48] <lynxlynxlynx> that line is chock full of stuff
[20:20:28] <brada> im surprised then that it doesnt explode
[20:26:31] <lynxlynxlynx> just luck perhaps
[20:29:11] <lynxlynxlynx> valgrind also complains about 172
[20:29:22] <lynxlynxlynx> Address 0x19d56fe2 is 18 bytes inside a block of size 32 free'd
[20:29:48] <lynxlynxlynx> it's a simple resize, so i'm baffled
[20:30:07] <brada> if it were just luck id assume that i would have seen a crash by now. at first glance it appears that the memory is allocated, just uninitialized.
[20:30:31] <brada> clearly something is wrong tho
[20:31:18] <brada> except resize should also initialize, no?
[20:31:28] <brada> ill look into that later
[20:32:08] <lynxlynxlynx> i can give you the full stack if you want
[20:32:30] <lynxlynxlynx> the only other thing is an old warning in sdl
[20:32:34] <lynxlynxlynx> so a good result overall
[20:34:36] <lynxlynxlynx> http://paste.debian.net/130634/
[20:34:37] <Pepelka90> debian Pastezone
[20:38:21] <lynxlynxlynx> disabled font debugging helped with almost 10 fps ingame
[20:38:38] <lynxlynxlynx> on the other hand, not noticeable on intro
[20:38:47] <lynxlynxlynx> back to normal stuff
[21:23:40] <lynxlynxlynx> the ogl driver also has some problems with text drawing - visible in iwd2 guirec main stat values
[21:23:55] <lynxlynxlynx> as if it forced more padding and chars overlapped
[21:34:13] <brada> its the thing i was complaining of earlier
[21:34:26] <brada> i lack the ability to do anything about it
[21:36:19] <brada> btw its not really a text drawing problem
[21:36:24] <lynxlynxlynx> sdl2 on the other hand has problems with scrolling
[21:36:25] <brada> its quite noticable in other place
[21:36:28] <brada> places
[21:36:49] <lynxlynxlynx> i can't scroll item descriptions except by clicking on the scrollbar
[21:37:04] <lynxlynxlynx> guirec at least works with mouse wheel after that
[21:37:26] <lynxlynxlynx> we need a toggle for perpixel scrolling if we don't have it yet, it sucks for mice
[21:38:27] <brada> thats nothing new with this tho. been that way since forever. shouldnt be using sdl2 for anything but touchscreens atm
[21:38:45] <brada> trackpads are even worse ;)
[21:43:58] <brada> im not sure a toggle is the best method anyway. that would just give android users more headache when its not enabled. probably can distinguish between input devices instead
[21:44:44] <lynxlynxlynx> yeah
[21:44:54] <lynxlynxlynx> sdl is bound to have something already
[21:45:19] <lynxlynxlynx> an interesting thing happened, someone shouted and managed to squeeze himself between the dialog text and options
[21:49:59] <Pepelka90> [wiki] developers:lynx http://www.gemrb.org/wiki/doku.php?id=developers:lynx&rev=1415310585&do=diff
[21:50:37] <brada> does that not happen on master too?
[21:50:43] <brada> ive seen that in pst
[21:50:58] <brada> but pst shouldnt record shouts to the textarea in the first place
[21:51:40] <brada> anyway i think i have an easy and good solution for that
[21:55:41] <lynxlynxlynx> it shouldn't happen at least if the dialog induces a pause
[21:56:11] <lynxlynxlynx> no idea how the originals behaved
[21:56:39] <lynxlynxlynx> queued for the interim?
[21:58:37] <brada> that would probably be the best, but since the seem pretty unimportant, i was thinking just insert them before the dialogBeginNode
[21:59:14] <brada> pluss queuing could potentially spam
[21:59:48] <brada> we could also drop them right?
[22:00:22] <brada> but i dont know if important things could potentially be dropped, so i think the insert is maybe the best/safest aproach
[22:22:09] <lynxlynxlynx> yeah, by queueing i meant up to the next change, in this case your reply
[22:22:39] <lynxlynxlynx> before i think they were appended at the end, which made it scroll
[22:37:28] <-- Morgoth_deb has left IRC (Quit: Page closed)
[23:04:01] <-- jillsmitt has left IRC (Quit: Backspace.Quit())
[23:51:39] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:59:07] <-- brada has left IRC (Quit: brada)