#gemrb@irc.freenode.net logs for 30 Mar 2013 (GMT)

Archive Today Yesterday Tomorrow
GemRB homepage


[00:00:33] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[01:54:45] --> brada has joined #gemrb
[01:55:16] <-- brada has left IRC (Client Quit)
[02:06:01] <-- Maighstir has left IRC (Read error: Connection reset by peer)
[05:42:16] --> brada has joined #gemrb
[05:55:12] <-- brada has left IRC (Quit: brada)
[08:01:56] --> edheldil_ has joined #gemrb
[08:06:23] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[08:50:26] --> Yoshimo has joined #gemrb
[08:51:41] <-- Yoshimo has left IRC (Client Quit)
[08:51:56] --> Yoshimo has joined #gemrb
[08:52:01] --> lynxlynxlynx has joined #gemrb
[08:52:01] <-- lynxlynxlynx has left IRC (Changing host)
[08:52:01] --> lynxlynxlynx has joined #gemrb
[08:52:02] --- ChanServ gives channel operator status to lynxlynxlynx
[09:22:43] --> edheldil_ has joined #gemrb
[09:42:13] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[09:51:56] --> edheldil_ has joined #gemrb
[09:52:30] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[09:57:13] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[10:47:56] <-- Gekz has left IRC (Ping timeout: 252 seconds)
[10:59:55] <-- |Cable| has left IRC (Ping timeout: 264 seconds)
[11:12:53] --> |Cable| has joined #gemrb
[11:13:17] --> rocket_hamster has joined #gemrb
[11:31:10] --> Gekz has joined #gemrb
[11:31:10] <-- Gekz has left IRC (Changing host)
[11:31:10] --> Gekz has joined #gemrb
[13:01:26] <-- Yoshimo has left IRC (Read error: Connection reset by peer)
[13:11:40] --> lynxlynxlynx has joined #gemrb
[13:11:40] --- ChanServ gives channel operator status to lynxlynxlynx
[13:33:36] --> edheldil_ has joined #gemrb
[13:54:55] --> Yoshimo has joined #gemrb
[14:36:15] <-- edheldil_ has left IRC (Ping timeout: 245 seconds)
[15:07:17] <-- Yoshimo has left IRC (Ping timeout: 252 seconds)
[15:51:29] --> brada has joined #gemrb
[15:59:19] --> edheldil_ has joined #gemrb
[16:13:16] <-- rocket_hamster has left IRC (Remote host closed the connection)
[17:09:08] <lynxlynxlynx> hah!
[17:09:58] <lynxlynxlynx> found an interesting bug, probably the cause of the intermittent "weapon unusable" messages
[17:10:30] <lynxlynxlynx> turns out one of the causes is us being too fair
[17:11:07] <lynxlynxlynx> we respect weapon breakability for enemies too, so they occasionally get theirs broken
[17:11:39] <lynxlynxlynx> there is no fist autoswap or anything like that, so they get pretty useless
[17:28:50] --> rocket_hamster has joined #gemrb
[18:12:09] <-- brada has left IRC (Quit: brada)
[18:18:03] <lynxlynxlynx> psch: here?
[18:19:35] --> brada has joined #gemrb
[18:20:40] <psch> briefly, yes
[18:20:45] <psch> lynxlynxlynx: what's up?
[18:21:17] <lynxlynxlynx> are there any blockers for including your work in the tree?
[18:21:48] <psch> well, what's left is brada's remarks about my prep_env.sh as it is now
[18:22:06] <psch> as in, include some means of providing prebuilt libraries instead of always checking out and building
[18:22:24] <brada> we dont need to provide libs
[18:22:33] <psch> for the user i mean
[18:22:36] <lynxlynxlynx> that can be added later
[18:22:37] <brada> we jsut ought to find them
[18:22:46] <-- brada has left IRC (Client Quit)
[18:23:11] <psch> i.e. the user tells the script where the libs are instead of the script just fetching them
[18:23:56] <psch> aside from that im just gotta look over it a bit, run it against latest tree, see if anything breaks
[18:24:06] <psch> but in general it should be fine about now
[18:24:38] <psch> there's a small thing wrt the SDL.h include, i.e. "SDL/SDL.h" from an installed SDL vs "SDL.h" from the headers pre-install
[18:27:29] <psch> i could get on testing once more in about an hour or so
[18:29:51] <lynxlynxlynx> ok, i'd just like to merge it soon
[18:30:11] <lynxlynxlynx> it's an excellent base if you don't consider it a solution it yet
[18:31:27] <psch> depending on how much breaks i could probably commit later this evening somewhen
[18:32:43] --> Maighstir has joined #gemrb
[18:39:47] <lynxlynxlynx> ok, great
[18:41:20] <lynxlynxlynx> it's one of the release blockers and since miha couldn't build a whole pelya build, this is the next best thing (beside being the platform for the future)
[18:42:11] <psch> what do you mean "whole pelya build"?
[18:43:03] <psch> pelya works 100% against sdl1.2
[18:43:16] <psch> but this release really wants multitouch i guess?
[18:50:40] <lynxlynxlynx> he couldn't get all the deps properly incorporated
[18:51:32] <lynxlynxlynx> at minimum freetype, but it's not that important
[18:51:41] <lynxlynxlynx> we'll provide both builds for this release
[19:19:21] --> nutron has joined #gemrb
[19:21:32] <edheldil_> Hi, folks
[19:27:18] <rocket_hamster> hello
[19:28:31] <psch> hi
[19:31:10] <psch> lines 187 and 201 in SDL20Video.cpp refer to SDL_color.unused, which doesn't exist in the latest sdl2 it seems
[19:31:22] <psch> i've uncommented the assignments and that seems to compile at least
[19:31:44] <psch> fwiw, the value that gets assigned is 0 and i haven't seen any references to the field at least in SDL20Video.cpp
[19:32:15] <psch> idk in how far we have a "stable" sdl2 version to build against, seeing as they still shuffle stuff around
[19:40:48] <lynxlynxlynx> it was added in a lumpy commit, so there's no specific reason for it
[19:41:12] <-- rocket_hamster has left IRC (Remote host closed the connection)
[19:41:15] <lynxlynxlynx> so if the base changed they should just be removed
[19:59:22] <psch> SDLActivity.java from upstream is broken :/
[19:59:42] <psch> so far i have found 2 "wrong case" errors
[20:00:14] <psch> well, at least ive got something to do
[20:02:11] <psch> ah, not actually a "wrong case" error, but sdl breaking against their declared minimum sdk version
[20:11:31] <psch> http://nopaste.info/d7b4c64868.html this is where i'm at right now - FATAL: no palette found
[20:11:32] <Seniorita> Nopaste - powered by project-mindstorm IT Services
[20:12:55] <psch> never saw that before, actually
[20:13:06] <psch> but then, i probably ain't seen nothin' yet
[20:14:10] <fuzzie> bad GameType? (bg1 vs vg2)?
[20:14:32] <psch> oh
[20:14:34] <psch> thanks
[20:15:38] <psch> yeah, gamepath points to bg1
[20:19:51] <psch> there's one issue that's kind of breaky in my opinion
[20:20:33] <psch> after being in the inventory or map (probably everywhere else except the actual "game view") and returning to the game it scrolls diagonally
[20:20:46] <psch> i have observed top left and bottom left up to now
[20:20:58] <psch> that also sometimes happens after scrolling the map with a 2 finger touch
[20:21:23] <psch> i don't remember if it always was like that
[20:21:58] <psch> single touching once fixes it
[20:22:14] <psch> it's probably not really breaking anything much, rather just an inconvenience
[20:23:35] <psch> aside from that, everything seems to be working
[20:24:05] <fuzzie> cool
[20:24:07] <psch> i had to change/add a few things to the prep_env script, but those were oversights in previous iterations
[20:24:34] <psch> like not actually copying the template gemrb.cfg or editing the SDLActivity line-based
[20:24:47] <psch> instead of appending after a match..
[20:25:12] <psch> im gonna add the actual gemrb icon too i think heh
[20:25:18] <psch> been meaing to do that for quite some time
[20:28:06] <viilee> Now this is annoying - GemRB does not save my options
[20:28:23] <viilee> Also, autopause settings have no effect
[20:28:40] <viilee> Are these known problems? Can't find via google...
[20:37:50] <-- |Cable| has left IRC (Ping timeout: 255 seconds)
[20:39:26] <psch> icons are 36x36, 48x48, 72x72 in the sdl project, should i add those to the artwork folder?
[20:39:38] <psch> seeing as the one in there is 64x64
[20:49:39] <lynxlynxlynx> viilee: what platform are you on?
[20:49:46] <lynxlynxlynx> and which game
[20:51:55] --> |Cable| has joined #gemrb
[20:53:16] <viilee> lynxlynxlynx: linux (gentoo). BGII(ToB) - with BGT and BG1. But same thing with plain BG1 (I just tested)...
[20:54:00] <viilee> I believe all the autopause settings are on, no matter what I choose in the options - and are not saved. There's a gem-baldur.ini created in the game directory, however everything is set to 0
[20:54:20] <viilee> If I delete it, I get music, until it get re-created... after that, everything is muted on start
[20:54:31] <lynxlynxlynx> awesome
[20:54:58] <lynxlynxlynx> someone reported a problem on windows and someone on android, but none of us was capable of reproducing it
[20:55:02] <lynxlynxlynx> this 0 config thing
[20:55:19] <lynxlynxlynx> are you building from git?
[20:55:30] <viilee> No, it's 0.7.2 release
[20:55:39] <viilee> But I could build from GIT if it helps
[20:56:30] <lynxlynxlynx> it would indeed
[21:01:07] <psch> http://i.imgur.com/KbZXAqQ.png i think the icon needs... something
[21:03:11] --> PixelScum has joined #gemrb
[21:03:43] <-- Drakkar has left IRC (Ping timeout: 256 seconds)
[21:06:10] <psch> i'll just leave it like that for now though, although it looks kind of not-so-nice
[21:06:31] <edheldil_> psch: best just make G white
[21:07:47] <psch> well, that's just the launcher there, users could have a white background on the home screen
[21:08:21] <psch> some kind of backdrop shadow or something might be the right idea, but i'm not really proficient with gimp
[21:26:01] --> Yoshimo has joined #gemrb
[21:41:05] <psch> alright, im done testing and merging
[21:45:08] --> brada has joined #gemrb
[21:46:30] <-- brada has left IRC (Client Quit)
[21:51:37] <psch> so what do i do now
[21:52:45] --> brada has joined #gemrb
[21:58:59] <-- brada has left IRC (Quit: brada)
[21:59:09] <lynxlynxlynx> ok
[21:59:24] <lynxlynxlynx> make commits if you haven't yet
[21:59:58] <lynxlynxlynx> i guess one for all the new stuff is ok, while any existing gemrb code mods should go into others
[22:00:14] <psch> ive made commits through
[22:00:21] <psch> merged master says im ahead by 33 comits
[22:00:29] <lynxlynxlynx> uff :)
[22:00:31] <psch> there was one rather big one in the beginning i think
[22:00:56] <psch> ive tried to say on a per-file level wrt commits
[22:01:14] <lynxlynxlynx> this is all new, so no need to bother with cleanup
[22:01:24] <lynxlynxlynx> git format-patch
[22:01:46] <psch> doesn't output anything
[22:01:51] <psch> and returns 0
[22:05:05] <lynxlynxlynx> it creates files
[22:05:12] <lynxlynxlynx> a patchset
[22:05:51] <lynxlynxlynx> either cat them together and put on a sane pastebin or send them to me by mail
[22:08:10] <psch> now i got them
[22:08:16] <psch> i merged my local branch into my local master
[22:08:35] <psch> and on my local master i ran format-patch
[22:08:43] <psch> which, of course, didnt find any difference against itsself
[22:14:40] <psch> 17 .patch files, at 2.5mb
[22:14:46] <psch> that fits in you mailbox?
[22:15:03] <psch> i tar'd it, so it's just one file, but the size is what gets filtered anyway
[22:15:59] <lynxlynxlynx> sure
[22:16:27] <psch> alright, sent
[22:22:07] <lynxlynxlynx> still flying
[22:22:31] <psch> oh shit
[22:22:35] <psch> i actually messed up
[22:22:43] <psch> wrong signature on the forums
[22:22:49] <psch> namely one that has an email
[22:22:54] <psch> i sent it to the bigg :|
[22:23:02] <lynxlynxlynx> hah
[22:27:31] <psch> yeah so
[22:27:42] <psch> i hope the wiki has uptodate email addresses
[22:27:55] <lynxlynxlynx> it's in AUTHORS and git log btw
[22:28:01] <psch> right
[22:28:51] <psch> the one in AUTHORS seems to be the same as on the wiki
[22:30:19] <psch> that's the one i send it to, anyway
[22:37:11] <lynxlynxlynx> the wiki links to that file directly
[22:37:40] <lynxlynxlynx> i don't remember any other addresses published on the wiki
[22:38:08] <psch> in that case you should recieve the email, i think
[22:43:09] <lynxlynxlynx> it's taking its time unfortunately
[22:43:42] <psch> to: lynxlupodian@users.sourceforge.net
[22:43:46] <psch> that's right isn't it
[22:43:55] <psch> i'm resorting to direct communication!
[22:45:57] <-- Yoshimo has left IRC (Read error: Connection reset by peer)
[22:46:53] --> Yoshimo has joined #gemrb
[22:50:21] <lynxlynxlynx> mhm
[22:51:29] <-- Yoshimo has left IRC (Ping timeout: 255 seconds)
[22:59:45] <viilee> lynxlynxlynx: I made the git build (I needed to bypass some compiling warnings - actually I made a live ebuild for it, seemed easier than rewriting the config file).
[22:59:52] <viilee> Same issue persists
[23:00:34] <viilee> But it's quite late here, so maybe we'll look into this tomorrow / another day... I need to sleep too ;-)
[23:00:37] <lynxlynxlynx> we have a switch to disable -werror btw
[23:00:47] <lynxlynxlynx> but it's good that you confirmed it
[23:00:56] <viilee> Yeah, I found a way.
[23:01:01] <lynxlynxlynx> i'll need you another day for further analysis
[23:01:44] <viilee> ah, ok. We'll get back to it
[23:01:51] <lynxlynxlynx> i don't have gcc 4.8 yet, so no idea if it emits anything new for us
[23:03:23] <-- Radegast has left #gemrb
[23:06:37] <viilee> I have 4.6.3 (the newest stable on Gentoo)
[23:09:30] <lynxlynxlynx> do you have the warning handy?
[23:12:02] <viilee> You mean the compile warning? Yes. It was Font.cpp, unused variable "test" - I'll paste the full line in a second...
[23:12:47] <fuzzie> master builds fine with gcc 4.8
[23:13:02] <fuzzie> you'll get test warning if you build in release mode
[23:13:03] <viilee> /var/tmp/portage/games-engines/gemrb-9999/work/gemrb-9999/gemrb/core/Font.cpp:619:10: virhe: käyttämätön muuttuja "test" [-Werror=unused-variable]
[23:13:06] <fuzzie> please don't build in release mode :-p
[23:13:13] <fuzzie> but of course we should fix it anyway
[23:13:15] <viilee> Sorry about the finnish
[23:13:24] <viilee> I should set LANG=C somewhere
[23:13:37] <viilee> It says error: unused variable (most probably)
[23:14:08] <fuzzie> yes
[23:14:14] <fuzzie> /home/fuzzie/src/gemrb/gemrb/core/Font.cpp:619:10: error: unused variable ‘test’ [-Werror=unused-variable]
[23:14:18] <fuzzie> ieWord* test = dbString;
[23:14:20] <fuzzie> ^
[23:14:52] <fuzzie> is that the only warning you get?
[23:14:59] <viilee> I believe there was more warnings, after I removed the -Werror
[23:15:06] <viilee> No
[23:15:50] <fuzzie> hm, in fact, we should unset NDEBUG in all builds
[23:15:56] <fuzzie> how are you building?
[23:18:36] <viilee> Umm., I'm quite novice in building (and now next to nothing about programmin) ... so I'm not quite sure what you want to know. I made a live ebuild (from the 0.7.2 ebuild already in portage). I guess it's quite straigforward cmake style
[23:19:00] <fuzzie> well, so, your error is caused because the assert() isn't being compiled in
[23:21:31] <fuzzie> which is deliberate sabotage by cmake-utils.eclass I guess
[23:23:20] <fuzzie> the cmake-utils docs suggest you can set the CMAKE_BUILD_TYPE variable to something more appropriate (e.g. RelWithDebInfo) to deal with that ("before invoking cmake-utils_src_configure")
[23:24:37] <viilee> I'm a bit lost here... I thought the error is caused because -Werror is in CXXFLAGS? What's assert() and where it is.... the sabotage by cmake-utils.eclass I can understand, though ;-)
[23:24:44] <fuzzie> no
[23:24:51] <fuzzie> it's caused because cmake-utils.eclass is passing -DNDEBUG
[23:25:05] <viilee> Hmm, OK
[23:25:08] <fuzzie> which sabotages assert(), which causes the 'unused variable' error because the variable is now unused :)
[23:25:32] <fuzzie> (since 'test' is only used by assert(), as you can see in the source)
[23:26:13] <viilee> Yes, I did look at it (though I don't understand it)
[23:26:27] <fuzzie> well this specific example seems a really silly and weird one
[23:26:34] <fuzzie> which is why I asked if you get other warnings
[23:26:48] <fuzzie> but we don't want people using builds without assert, really
[23:29:12] <fuzzie> since they're a lot less useful to test with, so if you can make your build include those it would be appreciated..
[23:29:26] <fuzzie> but of course you can also disable -Werror if you just want it to work, sure :)
[23:30:47] <viilee> Hmm., I'm still a bit lost here, but I put CMAKE_BUILD_TYPE="RelWithDebInfo" into the ebuild, and now it at least compiles. Is that enough, is the "assert()" (whatever that is) now compiled in?
[23:30:59] <fuzzie> yep
[23:31:20] <lynxlynxlynx> oddly still no mail
[23:31:25] <lynxlynxlynx> i'll just go to sleep
[23:31:28] <fuzzie> simplifying a bit: 'assert' is a function which causes an error if you pass it something which is false
[23:31:42] <fuzzie> so you can do "assert(something which must be true here);" and it will error out if it isn't true
[23:31:57] <fuzzie> which is a helpful way to make sure things aren't going horribly horribly wrong somewhere
[23:31:59] <psch> lynxlynxlynx: im puzzled
[23:32:06] <lynxlynxlynx> could be my end
[23:32:07] <viilee> Ok. Well, we can continue from this another day (if not tomorrow... or today depending on where you live)
[23:32:21] <psch> the bit i pasted was from gmail, so if the address was right there it should come
[23:32:46] <psch> i'd suspect sourceforge not allowing 2.5mb attachements, but i don't know and you're saying it should go through
[23:32:47] <fuzzie> but the documentation for cmake-utils.eclass does mention the possibility of needing to change CMAKE_BUILD_TYPE for silly projects like ours which match it
[23:32:56] <fuzzie> so it sounds like it might be the 'best' fix
[23:33:01] <lynxlynxlynx> gmail has a limit at 20 ir 25M, so that's fine
[23:33:22] <lynxlynxlynx> the sf one is only forwarding
[23:33:25] <psch> i see
[23:33:54] <lynxlynxlynx> fuzzie: i started working on the on-sight autopause, is this roughly what you had in mind? http://sprunge.us/hKcf?diff
[23:34:22] <fuzzie> well, I have totally forgotten what I had in mind
[23:34:27] <fuzzie> but the diff seems to make sense
[23:34:36] <lynxlynxlynx> I think I concluded the easiest solution is just to have an IF_WASVISIBLE. If an actor is visible, and it isn't already set, then set it and do the autopause.
[23:34:36] <lynxlynxlynx> And I guess for now you can just make it so that if it isn't visible and it is set, then unset it.
[23:34:36] <lynxlynxlynx> Maybe the latter is best done elsewhere, but that's a story for another time.
[23:34:58] <lynxlynxlynx> the first time we talked about it was more confusing
[23:35:11] <fuzzie> sounds right, then :)
[23:35:15] <psch> i guess if the mail didnt go through tomorrow i'll just upload it somewhere
[23:35:17] <psch> or we just wait more
[23:35:20] <psch> whatever
[23:35:45] <lynxlynxlynx> ok, just need to verify it then
[23:35:59] <lynxlynxlynx> from a quick look i didn't discover when we draw
[23:36:27] <lynxlynxlynx> but that's just something for day-eyes
[23:40:13] <-- lynxlynxlynx has left IRC (Remote host closed the connection)
[23:56:27] --> vampi-the-frog has joined #gemrb