[Back to Index]

  
[00:06] <-- SylvainTV left irc: Read error: Connection reset by peer
[00:15] <-- criezy left irc: Quit: criezy
[00:22] --> GitHub16 joined #scummvm.
[00:22] <GitHub16> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v69tt
[00:22] <GitHub16> scummvm/master 8b38295 Paul Gilbert: TITANIC: Cleanup of engine references
[00:22] GitHub16 (GitHub16@192.30.252.45) left #scummvm.
[00:36] <-- omer_mor left irc: Read error: Connection reset by peer
[00:43] <ScummBot> Port build status changed with 8b382954: Failure: master-wii, master-gamecube, master-debian-x86, master-ios7, master-mingw-w32, master-openpandora, master-gp2xwiz, master-android_arm, master-android_mips, master-amigaos4, master-ios, master-dingux, master-mingw-w64, master-debian-x86_64, master-webos, master-osx_intel, master-gp2x, master-android_x86, master-
[00:43] <ScummBot> debian-x86-clang
[00:44] <dreammaster> Crap
[00:47] --> GitHub33 joined #scummvm.
[00:47] <GitHub33> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v69mT
[00:47] <GitHub33> scummvm/master 2822fb5 Paul Gilbert: TITANIC: Compilation fix
[00:47] GitHub33 (GitHub33@192.30.252.45) left #scummvm.
[00:57] <ScummBot> Port build status changed with 2822fb58: Success: master-wii, master-gamecube, master-debian-x86, master-ios7, master-mingw-w32, master-openpandora, master-gp2xwiz, master-android_arm, master-android_mips, master-amigaos4, master-ios, master-dingux, master-mingw-w64, master-debian-x86_64, master-webos, master-osx_intel, master-gp2x, master-android_x86, master-
[00:57] <ScummBot> debian-x86-clang. Nice work, all
[00:57] <ScummBot> ports built fine now
[01:03] --> Vampire0_ joined #scummvm.
[01:07] <-- Vampire0 left irc: Ping timeout: 265 seconds
[01:36] <-- snover left irc: Quit: Leaving.
[01:43] <-- Littleboy left irc: Read error: Connection reset by peer
[01:49] <-- Dominus left irc: Ping timeout: 252 seconds
[01:49] --> Dominus joined #scummvm.
[01:49] <-- Dominus left irc: Changing host
[01:49] --> Dominus joined #scummvm.
[01:52] --> GitHub87 joined #scummvm.
[01:52] <GitHub87> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v69sW
[01:52] <GitHub87> scummvm/master df3e545 Paul Gilbert: TITANIC: Implemented more game classes
[01:52] GitHub87 (GitHub87@192.30.252.42) left #scummvm.
[02:01] --> Deledrius_ joined #scummvm.
[02:05] <-- Deledrius left irc: Ping timeout: 264 seconds
[02:19] --> GitHub46 joined #scummvm.
[02:19] <GitHub46> [scummvm] bluegr pushed 1 new commit to master: https://git.io/v69Gd
[02:19] <GitHub46> scummvm/master 1c19029 Filippos Karapetis: SCI32: Fix palette in SCI32 Mac games
[02:19] GitHub46 (GitHub46@192.30.252.45) left #scummvm.
[02:48] --> jammm joined #scummvm.
[02:48] <-- jamm left irc: Read error: Connection reset by peer
[03:38] --> Strangerke_ joined #scummvm.
[03:41] <-- Strangerke left irc: Ping timeout: 265 seconds
[03:41] Nick change: Strangerke_ -> Strangerke
[03:49] <-- dreammaster left irc:
[04:01] --> Polynomial-C joined #scummvm.
[04:03] <-- Poly-C left irc: Ping timeout: 244 seconds
[06:07] <-- Strangerke left irc: Ping timeout: 250 seconds
[06:15] --> GitHub67 joined #scummvm.
[06:15] <GitHub67> [scummvm] sev- pushed 1 new commit to master: https://git.io/v69Eo
[06:15] <GitHub67> scummvm/master 812a43d Eugene Sandulenko: DIRECTOR: Fix memory overwrite
[06:15] GitHub67 (GitHub67@192.30.252.41) left #scummvm.
[06:57] --> m_kiewitz joined #scummvm.
[06:57] <-- m_kiewitz left irc: Changing host
[06:57] --> m_kiewitz joined #scummvm.
[06:57] #scummvm: mode change '+o m_kiewitz' by ChanServ!ChanServ@services.
[07:00] --> _sev|work joined #scummvm.
[07:00] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[07:01] <-- _sev|work left irc: Client Quit
[07:02] <-- _sev left irc: Quit: Leaving
[07:14] --> _sev joined #scummvm.
[07:14] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[07:28] --> uruk-hai joined #scummvm.
[07:28] #scummvm: mode change '+o uruk-hai' by ChanServ!ChanServ@services.
[07:35] --> Strangerke|work joined #scummvm.
[07:35] <Strangerke|work> G'day
[07:36] <rootfather> hi Strangerke|work
[07:36] <Strangerke|work> hi rootfather :)
[07:54] <-- Mellified_Man_ left irc: Ping timeout: 244 seconds
[07:57] --> _sev|work joined #scummvm.
[07:57] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[07:58] --> Mellified_Man joined #scummvm.
[07:58] --> GitHub108 joined #scummvm.
[07:58] <GitHub108> [scummvm] sev- pushed 1 new commit to master: https://git.io/v69o8
[07:58] <GitHub108> scummvm/master f8483d2 Eugene Sandulenko: DIRECTOR: Split score.cpp file into classes
[07:58] GitHub108 (GitHub108@192.30.252.45) left #scummvm.
[07:59] --> t0by joined #scummvm.
[07:59] #scummvm: mode change '+v t0by' by ChanServ!ChanServ@services.
[08:27] --> waltervn joined #scummvm.
[08:27] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[08:28] <waltervn> morning
[08:36] <[md5]> good morning waltervn
[08:36] --> GitHub83 joined #scummvm.
[08:36] <GitHub83> [scummvm] bluegr pushed 2 new commits to master: https://git.io/v69Pw
[08:36] <GitHub83> scummvm/master cd00132 Filippos Karapetis: SCI32: Also handle game checks for "torinsg.cat" in Torin...
[08:36] <GitHub83> scummvm/master 3d1f1a3 Filippos Karapetis: SCI32: Fix restoring games from Torin's game restore dialog...
[08:36] GitHub83 (GitHub83@192.30.252.40) left #scummvm.
[08:45] --> ajax16384 joined #scummvm.
[08:45] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[08:46] --> TMM joined #scummvm.
[08:56] <-- ced117 left irc: Remote host closed the connection
[09:05] <-- Lightkey left irc: Ping timeout: 258 seconds
[09:08] <m_kiewitz> wtf
[09:08] <m_kiewitz> 04250185212309
[09:08] <m_kiewitz> ah damn :P
[09:08] <m_kiewitz> silly copy+paste
[09:08] <m_kiewitz> https://medium.com/art-marketing/browsing-your-website-does-not-mean-i-want-your-spam-3821267e902#.8v3jkeeag
[09:09] <m_kiewitz> anyone not using an adblocker should maybe start using it now
[09:18] --> Lightkey joined #scummvm.
[09:18] --> ChrisKeys joined #scummvm.
[09:20] <t0by> anyone not using links as their web browser should switch now.
[09:21] <t0by> if can't do, at least install one of those extensions that hide all comments from web pages and block all scripts.
[09:21] <t0by> besides adblock, that is.
[09:22] --> GitHub154 joined #scummvm.
[09:22] <GitHub154> [scummvm] bluegr pushed 1 new commit to master: https://git.io/v69yf
[09:22] <GitHub154> scummvm/master 6ce472a Filippos Karapetis: SCI32: Update the SCI2/SCI2.1 games that use global 90 for speech/subs
[09:22] GitHub154 (GitHub154@192.30.252.34) left #scummvm.
[09:23] <t0by> m_kiewitz: http://motherfuckingwebsite.com
[09:52] <wjp> [md5]: about this list check
[09:52] <wjp> [md5]: I can't check right now, but what I'm _guessing_ is happening is that s->_segman has been reset entirely if it reloaded?
[09:53] <wjp> the check if a segment is a list segment is just weird
[09:53] <wjp> it's a very indirect way of checking whatever we should really be checking
[09:53] --> WooShell joined #scummvm.
[09:54] <wjp> also, this is an incredibly hacky situation
[09:54] <[md5]> wjp: that is correct, yes
[09:54] <wjp> please don't add such things with the comment "e.g. when restoring in Torin" sounding like an afterthought instead of a very serious issue
[09:55] <[md5]> hm
[09:55] <[md5]> what happens is: kListEachElementDo contains the draw list of the restore dialog
[09:55] <[md5]> so when it tries to load a game, that list is reset
[09:55] <wjp> or is it normal that restoring happens from inside a VM run by a kernel function?
[09:56] <[md5]> which is why it doesn't happen from the ScummVM load dialog
[09:56] <[md5]> it is normal to restore, but not to save
[09:56] <[md5]> let me check... there was a condition
[09:56] <[md5]> kSaveGame:
[09:56] <[md5]> if (s->executionStackBase) {
[09:56] <[md5]> warning("kSaveGame - won't save from within kernel function");
[09:56] <[md5]> return NULL_REG;
[09:56] <[md5]> }
[09:56] <wjp> yeah, I've seen that being triggered recently actually, now that I think about it
[09:57] <wjp> so that's something else to look at
[09:57] <[md5]> yeah
[09:57] <[md5]> that's a weird case, when it happens
[09:58] <[md5]> in this case, the same thing occurs when restoring, because drawing is done within a kernel function
[09:58] <[md5]> SSCI has no such checks
[09:58] <wjp> SSCI restoring works entirely differently anyway
[09:58] <wjp> so can't compare
[09:58] <[md5]> that is true, yes
[09:58] <[md5]> :/
[09:59] <wjp> I wonder if we should be doing a delayed restore instead here
[09:59] <[md5]> the other weird thing is, that the "Continue" button in Torin doesn't work for me
[09:59] <[md5]> hm
[09:59] <[md5]> or, to check if a game is being loaded?
[09:59] <[md5]> and bail out of the loop?
[10:00] <wjp> checking that directly would be far preferable to checking it indirectly like now, yes
[10:00] <[md5]> i.e. if s->gameIsRestarting == GAMEISRESTARTING_RESTORE
[10:00] <[md5]> ok
[10:00] <[md5]> let me see
[10:00] <wjp> m_kiewitz probably understands the subtleties here better than me
[10:01] <m_kiewitz> huh
[10:01] <wjp> (hi :-) )
[10:01] <[md5]> hi :)
[10:01] <wjp> (see https://github.com/scummvm/scummvm/commit/3d1f1a3a21edd66f20f8de707ddcfcc8d97d13a6 and then the conversation here)
[10:02] <m_kiewitz> oh save+restore
[10:02] Action: m_kiewitz hides
[10:02] <m_kiewitz> :P
[10:02] <[md5]> :O
[10:03] <-- uruk-hai left irc: Ping timeout: 276 seconds
[10:06] <m_kiewitz> hmm i wonder if that's actually a script bug. In SCI16 access to deleted things worked almost all the time
[10:06] <m_kiewitz> in the original interpreter. That doesn't work in ScummVM of course
[10:09] --> GitHub107 joined #scummvm.
[10:09] <GitHub107> [scummvm] bluegr pushed 1 new commit to master: https://git.io/v69Qj
[10:09] <GitHub107> scummvm/master 6929cc6 Filippos Karapetis: SCI32: Explicitly abord kEachElementDo when a game is being loaded...
[10:09] GitHub107 (GitHub107@192.30.252.46) left #scummvm.
[10:09] <m_kiewitz> [md5] have you tried original restore/save too?
[10:09] <[md5]> argh... I meant "abort" :/ the grammar monster strikes again
[10:09] <m_kiewitz> or well is this even automatic save/restore?
[10:09] --> Henke37 joined #scummvm.
[10:09] <[md5]> m_kiewitz: the original restore does that
[10:10] <m_kiewitz> ScummVM dialog save/restore works?
[10:10] <[md5]> yes
[10:10] <[md5]> cause this is what is drawing the restore dialog
[10:10] <[md5]> i.e. this kListElementDo
[10:10] <[md5]> but the ScummVM restore dialog is handled in its own loop
[10:15] <-- Cruel` left irc: Quit: Leaving
[10:16] <m_kiewitz> yes, and i added the standard SCI script code for restoring as C++ code
[10:16] <m_kiewitz> ah i see
[10:17] <m_kiewitz> that's then possibly really a script bug, that just happened to work
[10:17] <m_kiewitz> does it call kRestoreGame before that?
[10:17] <[md5]> yes
[10:17] <[md5]> so the segment manager is reset
[10:17] <[md5]> so, it explodes
[10:26] <WooShell> good meowning =^.^=
[10:29] <-- jammm left irc: Read error: Connection reset by peer
[10:30] --> jamm joined #scummvm.
[10:32] <-- ChrisKeys left irc: Ping timeout: 240 seconds
[10:32] --> ChrisKeys joined #scummvm.
[10:41] <t0by> > a script bug, that just happened to work
[10:41] <t0by> ...unthinkable!
[10:41] <t0by> >:>
[10:47] <-- jamm left irc: Read error: Connection reset by peer
[10:48] --> jamm joined #scummvm.
[10:51] --> Mia joined #scummvm.
[11:06] --> GitHub111 joined #scummvm.
[11:06] <GitHub111> [scummvm] bluegr pushed 1 new commit to master: https://git.io/v69Aq
[11:06] <GitHub111> scummvm/master d15660e Filippos Karapetis: SCI32: Add a workaround for an uninitialized variable in Torin
[11:06] GitHub111 (GitHub111@192.30.252.41) left #scummvm.
[11:07] <[md5]> Happy to say that I've played Torin up to the beginning of chapter 3, and I've only experienced one graphical glitch ad the beginning of chapter 3 :)
[11:07] <[md5]> at *
[11:12] <LePhilousophe> oh nice!
[11:12] <LePhilousophe> last time I tried, there were glitches even in the intro video
[11:13] <[md5]> snover has done major rewrites
[11:14] <LePhilousophe> I didn't know
[11:14] <LePhilousophe> I once tried to fix some things but without a proper disassembly of Torin...
[11:14] <LePhilousophe> (never managed to handle the dos extender stuff)
[11:16] <LePhilousophe> [md5]: so is it playable after chapter 3 and you just stopped there? Or were you forced to stop?
[11:19] --> ced117 joined #scummvm.
[11:27] --> GitHub60 joined #scummvm.
[11:27] <GitHub60> [scummvm] bluegr pushed 1 new commit to master: https://git.io/v69pY
[11:27] <GitHub60> scummvm/master 14f163b Filippos Karapetis: SCI32: Allow kListAt to reference an empty list...
[11:27] GitHub60 (GitHub60@192.30.252.45) left #scummvm.
[11:27] <[md5]> LePhilousophe: up to now, I've experienced 2 graphical glitches in chapter 3
[11:27] <[md5]> I'm stuck in a puzzle now
[11:28] <[md5]> the phenocryst puzzle in chapter 3
[11:29] <LePhilousophe> oh yeah I think I see
[11:30] <[md5]> here:
[11:30] <[md5]> http://torinpassage.wikidot.com/puzzle-solutions
[11:30] <[md5]> I've placed the Pergolans like the picture... and nothing
[11:30] <[md5]> hm
[11:32] <LePhilousophe> oh so you think there is a bug
[11:32] <LePhilousophe> ok
[11:33] <[md5]> I'm not sure
[11:35] <wjp> [md5]: did you check what kListAt does in SSCI? I see an Alert() call in there, but can't check right now what it checks for exactly
[11:37] --> dreammaster joined #scummvm.
[11:37] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[11:38] <[md5]> wjp: yep, will do
[11:38] <[md5]> here's the correct solution: http://lparchive.org/Torins-Passage-(by-Blind-Sally)/Update%2020/
[11:40] --> GitHub18 joined #scummvm.
[11:40] <GitHub18> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v69h7
[11:40] <GitHub18> scummvm/master a26a3d4 Paul Gilbert: TITANIC: Implemented CLift class
[11:40] GitHub18 (GitHub18@192.30.252.40) left #scummvm.
[11:42] --> WinterGrascph joined #scummvm.
[11:42] #scummvm: mode change '+v WinterGrascph' by ChanServ!ChanServ@services.
[11:44] <[md5]> ooh, chapter 4
[11:45] <Strangerke|work> md5: Which company?
[11:45] <[md5]> Strangerke|work: Torin's Passage, Sierra
[11:45] <Strangerke|work> oh, sorry.
[11:45] <wjp> ;-)
[11:45] <Strangerke|work> I thought I could negotiate some legal rights for a cheap price.
[11:46] <[md5]> wjp: that's what SSCI does with kListAt
[11:52] <wjp> ok, seems ok then
[11:55] <wjp> this error() call seems to have been a bit of an historical accident
[11:55] <[md5]> yep
[11:55] <wjp> your initial ListAt implementation accidentally did a null deref in this case
[11:56] <[md5]> oh
[11:56] <[md5]> could be, it's been years now :)
[11:56] <wjp> about 7, yes :-)
[11:56] <[md5]> :O
[11:57] <[md5]> hm
[11:57] <[md5]> ok, there's a weird error now
[11:57] <[md5]> > scummvm.exe!Sci::CelObjView::CelObjView(const int viewId, const short loopNo, const short celNo) Line 920 C++
[11:57] <[md5]> scummvm.exe!Sci::kBaseSetter32(Sci::EngineState * s, int argc, Sci::reg_t * argv) Line 77 C++
[11:57] <[md5]> and it errors out here:
[11:57] <[md5]> if (_info.celNo < 0) {
[11:57] <[md5]> error("Cel is less than 0!");
[11:57] <[md5]> }
[11:58] <[md5]> resource 42002, loop 1, cel -1
[11:58] <[md5]> ho hum
[11:58] Action: [md5] checks
[11:58] <[md5]> but wow... 7 years...
[11:59] <-- dreammaster left irc:
[12:00] --> dreammaster joined #scummvm.
[12:00] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[12:02] <-- dreammaster left irc: Client Quit
[12:09] <-- |Cable| left irc: Ping timeout: 265 seconds
[12:10] <m_kiewitz> wjp: yeah, make us all feel terribly old :P
[12:13] <[md5]> ok there's no check for cel index in SSCI
[12:14] <[md5]> but we do
[12:14] <[md5]> somehow that worked in the original by chance?
[12:20] <m_kiewitz> when does it happen?
[12:20] <m_kiewitz> is there a screenobj with cel = -1?
[12:20] <[md5]> chapter 4, Asthenia, when you put the first ball in the catapult
[12:20] <[md5]> and yes, there is
[12:22] <m_kiewitz> really weird. but i guess that would have caused invalid memory access in SSCI?
[12:22] --> |Cable| joined #scummvm.
[12:22] <m_kiewitz> i know for a fact that for example larry 6 SCI32 liked to crash a lot and i mean a lot
[12:22] <[md5]> this is where it happens:
[12:22] <[md5]> https://www.youtube.com/watch?v=BIDT0A_6cdQ#t=1h38m08s
[12:23] <m_kiewitz> im currently watching a let's play of that game and every episode it crashes at least once lol (the let's player used ScummVM, but for that one he uses DOS-Box and he gets quite annoyed lol)
[12:24] <m_kiewitz> is it one of those boulder animations?
[12:25] <[md5]> yeah
[12:25] <[md5]> and here's where it stops now:
[12:25] <m_kiewitz> hmmm
[12:25] <[md5]> (A bit after this spot)
[12:25] <[md5]> WARNING: Dummy function kBitmapCreateFromView[23:9] invoked. Params: 6 (0000:0005, 0000:9d6c, 0000:0001, 0000:0000, 0000:0000, 0000:0000)!
[12:25] <m_kiewitz> where does it get the -1 from?
[12:25] <m_kiewitz> ah
[12:26] <[md5]> I was able to get past this spot like this:
[12:26] <[md5]> if (_info.celNo < 0) {
[12:26] <[md5]> warning("Cel is less than 0!");
[12:26] <[md5]> _info.celNo = 0;
[12:26] <[md5]> }
[12:26] <[md5]> eh
[12:26] <[md5]> from kBaseSetter32
[12:26] <[md5]> reg_t object = argv[0];
[12:26] <[md5]> const GuiResourceId viewId = readSelectorValue(s->_segMan, object, SELECTOR(view));
[12:26] <[md5]> const int16 loopNo = readSelectorValue(s->_segMan, object, SELECTOR(loop));
[12:26] <[md5]> const int16 celNo = readSelectorValue(s->_segMan, object, SELECTOR(cel));
[12:27] <m_kiewitz> i mean where does the screen obj / view get that -1 from
[12:28] <m_kiewitz> is it right in the heap? or is it set?
[12:28] <m_kiewitz> maybe there is something else going wrong
[12:29] <wjp> Wouldn't surprise me if a different bug is causing this -1, indeed
[12:30] <[md5]> hm
[12:30] <[md5]> there's code that fixes invalid cel counts in SCI16
[12:30] <[md5]> which is different than the one in SCI32
[12:31] <[md5]> SCI16:
[12:31] <[md5]> if (it->celNo >= viewCelCount) {
[12:31] <[md5]> it->celNo = 0;
[12:31] <[md5]> writeSelectorValue(_s->_segMan, it->object, SELECTOR(cel), it->celNo);
[12:31] <[md5]> } else if (it->celNo < 0) {
[12:31] <[md5]> it->celNo = viewCelCount - 1;
[12:31] <[md5]> }
[12:31] <[md5]> SCI32:
[12:31] <[md5]> if (_celInfo.celNo >= celCount) {
[12:31] <[md5]> const int maxCelNo = celCount - 1;
[12:31] <[md5]> _celInfo.celNo = maxCelNo;
[12:31] <[md5]> writeSelectorValue(segMan, object, SELECTOR(cel), maxCelNo);
[12:31] <[md5]> }
[12:31] <m_kiewitz> yes, Sierra actually did exactly that in SCI16
[12:32] <m_kiewitz> and some games even need exactly those fix ups
[12:32] <wjp> Signed or unsigned comparison with celCount?
[12:32] <m_kiewitz> i think Sierra even changed some inbetwen versions and games also require those
[12:32] <[md5]> oh
[12:32] <m_kiewitz> no idea about SCI32 behavior
[12:32] <[md5]> _celInfo.celNo is int16
[12:33] <[md5]> celCount is uint8
[12:33] <m_kiewitz> oh right, that one could cause celNo to be -1
[12:33] <m_kiewitz> can you set a breakpoint if its triggered right before?
[12:33] <m_kiewitz> that would explain what happens
[12:33] <[md5]> hm, let's see
[12:41] <[md5]> I think I found it
[12:41] <[md5]> yeah
[12:42] <[md5]> celNo is an unsigned integer in SSCI
[12:42] <[md5]> by mistake, perhaps
[12:42] <[md5]> so, this comparison will trigger when it is -1:
[12:42] <[md5]> if (_celInfo.celNo >= celCount) { ... }
[12:42] <[md5]> so, it worked by mistake :P
[12:43] <m_kiewitz> i guess maybe one should add an error in there too in case celCount is 0
[12:43] <m_kiewitz> so that celNo is never set to -1
[12:43] <m_kiewitz> or maybe not, idk. at least a warning should be added
[12:45] <[md5]> ok, so that fixes this :) a very subtle bug
[12:53] --> GitHub123 joined #scummvm.
[12:53] <GitHub123> [scummvm] bluegr pushed 1 new commit to master: https://git.io/v6HLv
[12:53] <GitHub123> scummvm/master 511d9f1 Filippos Karapetis: SCI32: Fix crash in Torin, chapter 4, catapult scene (via ScreenItem)...
[12:53] GitHub123 (GitHub123@192.30.252.45) left #scummvm.
[12:53] <[md5]> thanks for the help guys :)
[12:53] <[md5]> meh that got lost in the chat
[12:53] <[md5]> I said... thanks for the help guys :)
[13:04] <wjp> yes, that's why I asked if it was an unsigned comparison :-)
[13:05] <wjp> and it's not so relevant in general what the type of the variables is, it's relevant what the signedness of the comparison is
[13:09] <wjp> (jb vs jl and such)
[13:14] <-- Strangerke|work left irc: Ping timeout: 264 seconds
[13:19] --> snover joined #scummvm.
[13:19] #scummvm: mode change '+o snover' by ChanServ!ChanServ@services.
[13:21] <wjp> morning
[13:22] <[md5]> hello
[13:53] <-- WinterGrascph left irc: Quit: Leaving
[13:55] <wjp> [md5]: please (again) consider letting commits sit for a little while longer before pushing them to master
[13:55] <[md5]> which ones?
[13:55] <wjp> almost all of us do this to have some chance to consider the impact of them
[13:55] <wjp> well, most of them
[13:56] <[md5]> you mean in pull requests? I thought we agreed to do that only on major changes?
[13:56] <wjp> or just on a branch
[13:56] <[md5]> again, I thought we agreed to do that only for major changes?
[13:57] <wjp> as you noticed, several commits today warranted some discussion
[13:57] <[md5]> don't they always?
[13:57] <wjp> exactly
[13:57] <wjp> hence, please consider letting them sit before a while before pushing them to master
[13:58] <wjp> s/before/for/
[13:58] <-- Henke37 left irc: Ping timeout: 264 seconds
[13:58] <[md5]> but there was only one commit that warranted discussion... the rest were straightforward. So why do you mention that this was for most of them?
[13:59] <wjp> because it's not really easy to tell right away if a commits turns out to be straightforward or not
[13:59] <wjp> often the "oh, wait, forgot about X" realization comes a bit later
[14:00] <[md5]> true, but I didn't see a similar reaction when 10 or so commits were added some days ago
[14:00] <wjp> which ones?
[14:02] <[md5]> https://git.io/v61RA
[14:03] <wjp> those had mostly been sitting in snover's working branch for days already
[14:04] <wjp> (plus one or two minor ones, and some we had discussed on IRC already)
[14:05] <[md5]> yeah, some of these were not straightforward either. Will making them sit in a local working branch change anything? I'm just trying to figure out what warrants a pull request in SCI and what doesn't
[14:06] --> Henke37 joined #scummvm.
[14:08] <wjp> "oh wait, that should've been different"-commits tend to indicate a commit should've had some more time to mature, but it's of course nowhere near black and white
[14:08] <wjp> it's just nicer if that happens in branches than in master :-)
[14:09] <wjp> I'd suggest a non-local working branch over a local one to gives others the chance to think about it as well
[14:09] <wjp> s/gives/give/
[14:09] <[md5]> agreed :) There was only one commit that warranted a discussion, the one about kEachElementDo. The first commit was not that great, I agree
[14:10] <[md5]> e.g. the multi-CD support feature was absolutely great, but it wasn't discussed in a pull request... I still have questions about that one
[14:10] <[md5]> i.e. the resource files in some multi-CD games are different for each CD... can we safely ignore these differences, too?
[14:11] <[md5]> IIRC, the resource.001 files in GK2 are like that
[14:12] <[md5]> plus, the discussion we had for the changes in the robot player was never finished, and the pull request was merged, and I never got any answer
[14:20] <t0by> I have no idea what you are talking about, but try not to get too emotional about the issue. Sort this out quickly instead, so that you can get back to work on SCI2. I demand GK2 support asap :P
[14:21] <[md5]> heh :)
[14:21] --> uruk-hai joined #scummvm.
[14:21] #scummvm: mode change '+o uruk-hai' by ChanServ!ChanServ@services.
[14:32] --> uruk_hai joined #scummvm.
[14:32] #scummvm: mode change '+o uruk_hai' by ChanServ!ChanServ@services.
[14:35] <-- uruk-hai left irc: Ping timeout: 276 seconds
[14:40] <-- ajax16384 left irc: Quit: Leaving
[14:50] <-- uruk_hai left irc: Read error: Connection reset by peer
[14:56] <m_kiewitz> [md5] it's only about the audio + sfx resources
[14:56] <m_kiewitz> and those are actually different on each CD. It seems snover's version of GK1 even has different maps on each discs. My 2 versions have the same map on each disc
[14:56] <snover> (phant1 :))
[14:57] <m_kiewitz> was it phantasmagoria 1? i thought you said gk1
[14:57] <snover> GK1 is a 1-CD game :)
[14:57] <m_kiewitz> oh right, my bad urgh
[14:57] Action: m_kiewitz hides
[14:58] <[md5]> :P
[14:59] <[md5]> IIRC, the issue was with Phant1 and GK1, cause they had files with the same names and different content
[14:59] Action: [md5] checks
[15:00] <m_kiewitz> gk2, i think i checked gk2 too and the maps were also the same
[15:00] Action: [md5] mumbles
[15:00] <[md5]> Lighthouse is OK...
[15:00] <m_kiewitz> ok? as in maps are the same?
[15:01] <[md5]> no, different names
[15:01] <m_kiewitz> ah that
[15:03] <[md5]> Phantasmagoria 1 has audio files with the same name, but different names for resource files
[15:04] <[md5]> (just checking the CDs now)
[15:04] <m_kiewitz> for gk2 we could add an info text about the renaming in case resaud.001 etc. are not available
[15:04] <[md5]> so that works with snover's changes
[15:04] <[md5]> (well done!)
[15:04] <m_kiewitz> i mean tell the user to rename resource.aud on the first CD to resaud.001, etc.
[15:05] <[md5]> we could throw a warning and refuse to run the game
[15:05] <[md5]> if the files are missing or not named properly
[15:05] <[md5]> Phantasmagoria 2 has different file names
[15:10] <[md5]> GK2 has different resource.aud and resource.sfx files in each CD
[15:10] <[md5]> with the same name
[15:11] <[md5]> Phant1 has the same resource.aud in each CD, but different resource.sfx files
[15:13] <[md5]> so that probably doesn't work with the current changes
[15:14] <[md5]> cause the code assumes that the AUD files are different, whereas the SFX ones are the same (but the opposite is true)
[15:15] <[md5]> ah, no, it handles both cases
[15:21] --> Cruel` joined #scummvm.
[15:21] <[md5]> I don't understand why 37.map should be omitted
[15:22] <[md5]> and we could skip 65535.map in multi-cd games ourselves
[15:33] <-- TMM left irc: Quit: Ex-Chat
[15:36] <m_kiewitz> [md5] yeah, the gk2 files need to be renamed - resaud.001, ressfx.001, resaud.002, ressfx.002, etc.
[15:37] <m_kiewitz> and our code should then switch around according to the requested CD number by the scripts
[15:39] <[md5]> there is a similar functionality used for MUMG deluxe (for the English/Spanish toggle)
[15:51] <-- snover left irc: Quit: Leaving.
[16:11] <-- _sev|work left irc: Quit: This computer has gone to sleep
[16:11] --> ignalina joined #scummvm.
[16:23] Action: t0by would merrily volunteer for some gk2 grunt work btw
[16:23] Action: t0by loves gk2
[16:23] Action: t0by loves gk-anything
[16:31] --> ny00123 joined #scummvm.
[16:34] --> ajax16384 joined #scummvm.
[16:34] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[16:40] --> Littleboy joined #scummvm.
[16:40] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[16:56] <bgK> t0by: are you interested in joining efforts with the guy from gk3tools and starting a ResidualVM engine? :)
[16:57] <t0by> bgK, caveat: *right now* I have no free time, no life, no money and little will to live. But I also love gk3, so :)
[16:58] <t0by> I'd love to contribute what I can/
[16:59] <t0by> bgK, please don't hesitate to drop me a line or whatever.
[17:02] <bgK> well the SVN repo at https://sourceforge.net/projects/gk3tools/ has a lot of resources (level viewer, script engine, ...) that could be used as a base to reimplement the game
[17:07] --> Strangerke joined #scummvm.
[17:07] <t0by> bgK, bookmarked, will tinker sometime in the fall
[17:07] <t0by> no earlier than mid-september, though
[17:07] <-- ignalina left irc: Quit: ignalina
[17:08] <t0by> bgK, are there specific plans or ideas being thrown around?
[17:08] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[17:09] <bgK> t0by: you'd have to create plans should you want some, but I think GK3 is more about maps ;)
[17:09] <t0by> bgK, ah, so you didn't actually have talks with gk3tools guy?
[17:10] <bgK> t0by: not at all
[17:10] <t0by> sorry, misinterpreted
[17:11] <t0by> bgK: Still, it's very interesting, didn't know about the project. Certainly I will have a look when I have a bit of free time and see what the situation is.
[17:13] --> GitHub191 joined #scummvm.
[17:13] <GitHub191> [scummvm] sev- pushed 3 new commits to master: https://git.io/v6H7T
[17:13] <GitHub191> scummvm/master 3d6d6fa Eugene Sandulenko: DIRECTOR: Reduce header dependency
[17:13] <GitHub191> scummvm/master dbf2876 Eugene Sandulenko: DIRECTOR: Further reduction of include dependency
[17:13] <GitHub191> scummvm/master 1bbbb14 Eugene Sandulenko: DIRECTOR: Added test D3 target
[17:13] GitHub191 (GitHub191@192.30.252.40) left #scummvm.
[17:13] <t0by> Don't set your expectations too high, though
[17:13] <t0by> Yay more director.
[17:13] <t0by> So many cool engines
[17:13] <t0by> so little time
[17:13] <t0by> :(
[17:16] <bgK> t0by: that is so true
[17:17] <bgK> nevertheless I will send a message to the gk3tools guy to see what he thinks about ResidualVM
[17:17] <t0by> bgK, neat. BTW honestly I thought Residual's scope was limited to 2.5D. No, I have no idea where I got that idea from.
[17:20] --> ignalina joined #scummvm.
[17:24] <bgK> well, Myst III is not 2.5D
[17:29] --> ComradeAnderson joined #scummvm.
[17:32] <t0by> well, it's round(2.5)D
[17:33] <t0by> do we have to start splitting hairs now.
[17:56] --> SylvainTV joined #scummvm.
[17:56] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[18:02] <-- ajax16384 left irc: Quit: Leaving
[18:08] --> ajax16384 joined #scummvm.
[18:08] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[18:09] <-- ignalina left irc: Quit: ignalina
[18:29] --> Sylvain joined #scummvm.
[18:37] <-- Sylvain left irc: Quit: User pushed the X - because it's Xtra, baby
[18:47] --> ignalina joined #scummvm.
[18:48] --> jammm joined #scummvm.
[18:52] <-- jamm left irc: Ping timeout: 244 seconds
[18:56] <-- jammm left irc: Quit: Leaving
[18:57] --> jamm joined #scummvm.
[19:21] --> criezy joined #scummvm.
[19:21] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[19:24] --> LyTchet joined #scummvm.
[19:30] <-- t0by left irc: Quit: Bye!
[19:31] <-- LyTchet left irc:
[19:33] --> frankyboy_ joined #scummvm.
[19:44] <-- ajax16384 left irc: Read error: Connection reset by peer
[19:58] <-- ignalina left irc: Quit: ignalina
[20:05] --> t0by joined #scummvm.
[20:05] #scummvm: mode change '+v t0by' by ChanServ!ChanServ@services.
[20:13] <-- jamm left irc: Ping timeout: 240 seconds
[20:13] --> jamm joined #scummvm.
[20:27] --> snover joined #scummvm.
[20:27] #scummvm: mode change '+o snover' by ChanServ!ChanServ@services.
[20:56] <-- ny00123 left irc: Quit: Leaving
[20:58] <-- frankyboy_ left irc: Remote host closed the connection
[21:31] <-- criezy left irc: Quit: criezy
[21:39] <DrMcCoy> SDL2's SDL_SetWindowIcon() for some reason behaves very weird with my WM Enlightenment DR16
[21:40] <DrMcCoy> (SDL2 uses the _NET_WM_ICON Atom instead of XSetWMHints() like SDL 1.2)
[21:40] <DrMcCoy> Allegedly, e16 supports _NET_WM_ICON
[21:41] <DrMcCoy> But I never got it working within xoreos and I saw other programs using SDL2 where the icon just wouldn't stick
[21:41] <DrMcCoy> In e16, the icon isn't displayed on the window itself, but only in the iconbox when the program is minimized. When there's no icon, e16 shows a small version of the window's contents instead
[21:41] <DrMcCoy> Interestingly, ScummVM does show the icon there, even when build with SDL2
[21:42] <DrMcCoy> But: only when I start the game directly by giving a target on the command line
[21:43] <DrMcCoy> When I start ScummVM without one on the command line, i.e. when using the launcher, the icon doesn't show up when ScummVM is minimized
[21:43] <DrMcCoy> Neither in the launcher nor after having started a game
[21:43] <DrMcCoy> Frankly, I don't get it :)
[21:48] --> GitHub47 joined #scummvm.
[21:48] <GitHub47> [scummvm] bluegr opened pull request #812: Torin's Passage maze functionality (master...torin_maze) https://git.io/v6Q6q
[21:48] GitHub47 (GitHub47@192.30.252.41) left #scummvm.
[22:10] --> TMM joined #scummvm.
[22:11] <-- Cruel` left irc: Quit: Leaving
[22:13] --> GitHub146 joined #scummvm.
[22:13] <GitHub146> [scummvm] sev- pushed 1 new commit to master: https://git.io/v6Q1Z
[22:13] <GitHub146> scummvm/master 754907f Eugene Sandulenko: DIRECTOR: Gather statistic on correct bitmap sizes
[22:13] GitHub146 (GitHub146@192.30.252.34) left #scummvm.
[22:16] --> Cruel` joined #scummvm.
[22:23] --> dreammaster joined #scummvm.
[22:23] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[22:31] <m_kiewitz> [md5] what's so special about that Torin's room that they had to do it that way?
[22:31] <m_kiewitz> any idea what happened?
[22:31] <m_kiewitz> too complicated for the normal way?
[22:31] <m_kiewitz> or is it randomly generated?
[22:39] <[md5]> m_kiewitz: it was a special room designed by Al Lowe's daughter (!!)
[22:39] <m_kiewitz> you kidding?
[22:39] <[md5]> nope
[22:39] <m_kiewitz> so wait, they added a special kernel function for that?
[22:39] <[md5]> it was probably easier to create this netmask logic (again), than use polygons
[22:39] <[md5]> yes
[22:39] <m_kiewitz> or she created the room in the most complicated way possible?
[22:39] <[md5]> the latter :P
[22:39] <m_kiewitz> lol
[22:40] <m_kiewitz> i didn't even know that he had a daughter
[22:40] <[md5]> http://lparchive.org/Torins-Passage-(by-Blind-Sally)/Update%2023/
[22:41] <[md5]> "Anyways, welcome to the lava maze. This part sucks. A lot. It's horrible. We are unable to right click on the map to have Torin quickly teleport to that area. The game has disabled it so we don't "cheat" through the maze. Thus, we actually have to click Torin through every step of the way.
[22:41] <[md5]> Fun fact: the design of the maze was drawn up by Al Lowe's daughter. I don't blame her, though, I blame Lowe and the developers for deciding to put the maze in in the first place."
[22:42] <m_kiewitz> i guess we could add a debug command to disable sierra disabling right click :P
[22:43] <-- WooShell left irc: Quit: Zu gotdy od mpy nrmy stpimf. Zu drvpmf zrsmd aogy jrt iq pt viy jrt yp yjr htpimf.
[22:43] <m_kiewitz> does psychedelic mode work already correctly?
[22:43] <m_kiewitz> well "correctly"
[22:44] <m_kiewitz> "Actually there *IS* another Sailor Moon egg in Torin's Passage. But it's
[22:44] <m_kiewitz> nearly impossible to get to. You have to work Torin into a invisible maze
[22:44] <m_kiewitz> that goes over into the lower left corner of a lava screen. If you do so,
[22:44] <m_kiewitz> his legs will dissappear and as he walks in the corner, "Sailor Moon" in
[22:44] <m_kiewitz> Japanese will appear in place of his legs as he walks there."
[22:44] <m_kiewitz> WOW
[22:46] <[md5]> yeah :P
[22:46] <[md5]> and one of the artists replied below
[23:08] --> GitHub17 joined #scummvm.
[23:08] <GitHub17> [scummvm] waltervn pushed 1 new commit to master: https://git.io/v6QQ8
[23:08] <GitHub17> scummvm/master 3d558fe Walter van Niftrik: ADL: Make the optional scanlines less harsh
[23:08] GitHub17 (GitHub17@192.30.252.45) left #scummvm.
[23:24] <-- t0by left irc: Quit: Bye!
[23:30] <-- m_kiewitz left irc: Quit: technology isn't intrinsically good or evil. It's how it's used. Like the Death Ray.
[23:31] <-- waltervn left irc: Quit: Leaving
[23:43] <-- SylvainTV left irc: Read error: Connection reset by peer
[00:00] --- Wed Aug 24 2016