ScummVM IRC Log - 17 Sep 2017

[Back to Index]

[00:26] --> GitHub154 joined #scummvm.
[00:26] <GitHub154> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5doL
[00:26] <GitHub154> scummvm/master cedd15f Paul Gilbert: TITANIC: Fix incorrect sound effect in German prologue
[00:26] GitHub154 (GitHub154@192.30.252.40) left #scummvm.
[00:38] <tsoliman> dreammaster: I looked at the differences between the titanic mac version and the windows version and the assets folders are similar but not identical.. I don't feel like "installing" the game (it wants OSX7) so I was wondering which files contain the scripts in the windows version?
[00:39] <tsoliman> the detection table is keyed of the EXE which doesn't exist in the mac version obviously, but I wanted to see how far it can go unmodified (and running off of the not-installed datafiles)
[00:39] <tsoliman> no it isn't the exe actually ... hmm
[00:42] <snover> i imagine wont go anywhere since the game engine is heavily designed around AVI files
[00:42] <snover> which i assume were rendered as quicktime movies in the mac version
[00:42] <dreammaster> Right, currently it uses the "newgame.st", which is basically the savegame representing the start of the game
[00:43] <dreammaster> You're probably right. Though depending whether they were bog standard or not, it might be possible to set up the AVISurface to instantiate a QuickTime decoder rather than an AVI one. And it might work
[00:44] <tsoliman> the text files are of the CR-only variety :) I don't think I've seen those before ever
[00:44] <snover> wonderful incompatible mac os classic.
[00:44] <dreammaster> :)
[00:46] <snover> cooperative multitasking and all the rest.
[00:48] <tsoliman> https://gist.github.com/tsoliman/d1377ef2de9e7d50e90709f0394bf80a
[00:49] <snover> wow!
[00:49] <tsoliman> Assets vs HDAssets :)
[00:49] <snover> are those really AVIs?
[00:49] <tsoliman> STCD3/Assets/z99.avi: RIFF (little-endian) data, AVI, 152 x 80, video: uncompressed
[00:49] <snover> wowww!
[00:49] <snover> thats really lucky.
[00:49] <dreammaster> Hot damn. Maybe it could be easily supported
[00:50] <tsoliman> HD here presumably means hard drive
[00:50] <dreammaster> Interesting as well that it has HDAssets for some of the videos
[00:50] <dreammaster> Oh. I thought it meant high definition. But you could be right
[00:50] <tsoliman> hmm ..
[00:51] <tsoliman> this is slightly weird
[00:51] <tsoliman> some of them are QT and some are regular AVI
[00:51] <snover> thats less lucky!
[00:51] <dreammaster> Blerg. Given the number of different encoding methods in the PC version, it's unsurprising that they might have had to reencode some videos for the Mac version
[00:51] <snover> how does ffplay or vlc feel about them?
[00:52] <tsoliman> vlc plays the AVI ones file
[00:52] <tsoliman> I updated the gist - check it out
[00:52] <tsoliman> (there's a second file in the gist now)
[00:52] <dafioram> tsoliman: The file size of my y222.avi is 4,463 KB what is the mac size of that file?
[00:53] <snover> these guys were so crazy
[00:53] <tsoliman> dafioram: 4571136
[00:54] <tsoliman> (bytes)
[00:55] <dafioram> mine is same so i guess its not high definition
[00:55] <snover> there doesnt seem to be any particular rhyme or reason on the quicktime re-encodes so far
[00:55] <tsoliman> snover: mpv plays "Apple QuickTime movie (unoptimized)" and "RIFF (little-endian) data, AVI, 64 x 64, video: uncompressed" fine
[00:55] <snover> in the wintel version, some of them are indeo 4.1, some of them are cinepak
[00:55] <tsoliman> it even plays the "data" one (file couldn't figure it out)
[00:55] <snover> some of them are uncompressed
[00:55] <snover> so it seems kind of random that they got re-encoded
[00:56] <tsoliman> the install dir contains an indeo extension installer
[00:57] <tsoliman> that is empty apparently :)
[00:57] <snover> not empty, the data would be stored in the resource fork
[00:57] <tsoliman> this is where stuff like forks and other mac legacy voodoo comes in
[00:57] <tsoliman> yep
[00:58] <tsoliman> the only time I dealt with forks was with Pegasus .. and that had a nice linux script to do it for me :)
[00:58] <tsoliman> now I am on an actual (modern) mac
[01:01] <tsoliman> hmm .. yeah .. the GOG (windows) version has a bunch of avi files that are also of mixed types and encoding
[01:03] <-- Dubberino left irc: Remote host closed the connection
[01:04] --> Dubberino joined #scummvm.
[01:04] <-- Polynomial-C left irc: Ping timeout: 255 seconds
[01:05] <-- Dominus left irc: Ping timeout: 260 seconds
[01:06] <tsoliman> according to the READMEs .. the GOG version is "v1.00.42" and the mac version is "1.00.42mc" .. which makes me think they're the same .. of course GOG could have done stuff without updating the readme
[01:06] --> Dominus joined #scummvm.
[01:07] <snover> but is it v1.00.42a, v1.00.42b, or v1.00.42c? :P
[01:14] <tsoliman> oh nice .. the windows detection entry detects the mac version because the newgame.st is identical :D
[01:15] <snover> so far so good
[01:15] <dreammaster> Hah. Well at least that means they didn't introduce any new fields to any of the game classes, like they did for the German version
[01:16] <tsoliman> WARNING: Failed to find RIFF header!
[01:16] <tsoliman> Could not open video - ycursors.avi!
[01:17] <dafioram> sounds like a bug
[01:17] Nick change: Storm-AFK -> Stormkeeper
[01:17] <dreammaster> As far as I'm aware, all AVI files should start with RIFF, irrespective of what frame or audio formats they use internally
[01:17] <snover> ycursors.avi is one of the QT files
[01:17] <tsoliman> yeah file output for gog is : RIFF (little-endian) data, AVI, 64 x 64, 15.00 fps, video:
[01:18] <snover> or is this happening with the GOG release?
[01:18] <tsoliman> but for the mac version is : Apple QuickTime movie (fast start)
[01:18] <tsoliman> no the GOG release works fine :)
[01:18] <dreammaster> Right, the file encodes each cursor as a 64x64 frame
[01:18] <tsoliman> I am trying to see how far I can go with the mac version
[01:18] <tsoliman> switching the platform to "macintosh" was a bad idea so I kept it at windows (because it cannot find game data otherwise somehow)
[01:20] <snover> if its not a RIFF, see if its a&whatever the first QT atom is
[01:20] <snover> mdat? moov?
[01:20] <snover> &and if so then change decoder to QuickTimeDecoder
[01:22] <-- ignalina left irc: Quit: ignalina
[01:22] <snover> or i guess in the interest of hacking it to see what works, just change it to QuickTimeDecoder if RIFF isnt there
[01:24] <snover> the first dword is BE atom size, second dword is BE fourCC-style atom type
[01:25] <tsoliman> this is engines/titanic/support/avi_surface.cpp correct?
[01:27] <snover> yeah
[01:27] <snover> seems like it may be a little trickstery since Video::AVIDecoder is inherited instead of used as a concrete implementation
[01:28] <dreammaster> Right. There's various changes I do.
[01:29] <snover> but it looks pretty well-encapsulated so it should be easy enough to address that
[01:29] <dreammaster> Come to that, simply replacing AVIDecoder with a quicktime decoder wouldn't work, as there are AVI specific things I do, like getting reference to transparency surface video tracks, that a QT decoder wouldn't have
[01:29] <snover> of course. but at least that could be stubbed just to see what, if anything, can get rendered right now.
[01:30] <tsoliman> I just realized the inherited decoder that has extra gubbins
[01:30] <dreammaster> It may be simpler before going all out to just try hacking a simple quicktime video playback loop in titanic.cpp that loads in the ycursors.avi file and plays it. That way you'd know whether ScummVM can handle it or not, before you start fiddling around with AVISurface
[01:31] <snover> what a rational suggestion :)
[01:31] <tsoliman> I mean I could also replace it with the other ycursors.avi - you know the one from GOG :)
[01:31] <tsoliman> I guess I want to compare the game scripts - which I am still not sure where they are .. the savegame is the same I guess?
[01:31] <dreammaster> That too, I guess. Though that it failed on ycursors.avi means it may well fail on any other QT AVI file
[01:32] <dreammaster> Well, at least the first 5kb are.
[01:32] <tsoliman> yep - I mean there's also all the business with reverse playback
[01:32] <tsoliman> I wonder if that's why a bunch of stuff is uncompressed
[01:32] <dreammaster> Which bodes well. All the scripts are in the various game object classes and NPC script classes
[01:32] <tsoliman> no I md5summed the entire file .. it is identical
[01:33] <dreammaster> Ah, good then. So my earlier comment remains valid.. given that they're identical, it's more than likely the game object classes have identical logic. It's only the video decoding and playback that differed for the Mac version.
[01:34] <dreammaster> Which would make sense since they're both English. There's probably only the video playback differences, and other minor things like for CD swapping, etc.
[01:34] <tsoliman> so as a quick and dirty test .. I'll just overwrite the AVIs from GOG on top of this to verify this theory
[01:34] <snover> did you hash and compare the avi files?
[01:35] <tsoliman> yeah and some were the same and some were different
[01:35] <tsoliman> and some were missing in A and some were missing in B
[01:36] <dreammaster> But with all the game logic in the C++ classes, what is there in the assets folders but the AVIs, wave files, and the occasional image? :)
[01:36] <tsoliman> actually I take that back .. mac doesn't contain any video files not in GOG - just that GOG contains stuff not in mac
[01:36] <snover> are the other .st files different too?
[01:37] <dreammaster> Good question. If they are, it would indicate that the wave files were re-encoded as well
[01:39] <tsoliman> snover: the ST files are all identical
[01:39] <dreammaster> Excellent. That's one less thing to worry about then, all the sound effect and voice were left unchanged
[01:40] --> Polynomial-C joined #scummvm.
[01:42] <-- Lightkey left irc: Ping timeout: 255 seconds
[01:42] <-- Stormkeeper left irc: Quit: KVIrc 4.9.2 Aria http://www.kvirc.net/
[01:43] <tsoliman> ok this is cool - the experiment of replacing the mac AVIs with GOG ones worked flawlessly
[01:43] <tsoliman> so this means it should only be the AVI files that are different
[01:44] <dreammaster> Promising for supporting it then, in theory
[01:46] <tsoliman> how do you skip the doorbot spiel again?
[01:47] <dafioram> you can only skip deskbot or you can use ctrl+c and go where ever you want
[01:47] --> Stormkeeper joined #scummvm.
[01:54] --> Lightkey joined #scummvm.
[02:03] --> GitHub12 joined #scummvm.
[02:03] <GitHub12> [scummvm] dreammaster pushed 2 new commits to master: https://git.io/v5d6H
[02:03] <GitHub12> scummvm/master 6cac39b Paul Gilbert: TITANIC: Add differing response Id range checks for German
[02:03] <GitHub12> scummvm/master 787392f Paul Gilbert: DEVTOOLS: Add German Id mappings for Doorbot in create_titanic
[02:03] GitHub12 (GitHub12@192.30.252.35) left #scummvm.
[02:46] <dafioram> snover: is it possible to open resource.001 with sci companion? or some other way of inspecting resource.001?
[02:46] <snover> dafioram: which game?
[02:46] <dafioram> sci quest, a fan game
[02:46] <snover> i would assume sci companion will do such a thing, otherwise i am sure SV will
[02:47] <dafioram> SV?
[02:47] <snover> sci viewer. didnt i send this to you?
[02:47] <dafioram> companion didn't show it
[02:47] <dafioram> it did let me look at resource.map
[02:47] <snover> RESOURCE.MAP is the index into some RESOURCE.00x
[02:48] <dafioram> i dont think i got the viewer
[02:48] <snover> huh. i wonder who else i sent it too.
[02:54] <Lightkey> L0ngcat is looong
[02:54] <snover> i thought thats the helpful person it may have been.
[02:54] <snover> didnt want to raise them
[02:54] <snover> but now look what youve done.
[02:54] <-- dreammaster left irc:
[03:00] <-- fydo left irc: Remote host closed the connection
[03:01] --> kurtwr joined #scummvm.
[03:02] <-- kurtwr2 left irc: Ping timeout: 240 seconds
[03:04] <Lightkey> nyan?
[03:26] <dafioram> snover: I was looking for a bug that maybe easier than #9752 and I think #9977 maybe the one. Does that look like a good one or do you have a better one in mind?
[03:26] <snover> dafioram: let me look
[03:30] <snover> dafioram: #9977 or #10034 seem like good choices
[03:30] <snover> or #9790
[03:30] <snover> or #9824
[03:31] <snover> the other ones that arent caused by MIDI needing rewriting or pathfinding needing rewriting are look like they could easily be kind of shitty to figure out
[03:31] <dafioram> yes
[03:32] <snover> i already tried to figure out #9846 at one point and i couldnt do it
[03:32] <dafioram> 10034 is pretty much for all file name and author name in the phant2 computer, any chance that will conflict with other patches?
[03:33] <dafioram> like the one you patched
[03:33] <snover> i suspect its just patching the default y-coordinate of those objects, so unlikely it would conflict with something else. definitely unlikely to conflict with the patch i did, which just swapped the order of some calls
[03:33] <dafioram> yes
[03:34] <dafioram> okay let me do 10034
[03:36] <snover> object patching can be a pain in the butt since the script patcher wasnt really designed for it so its pretty hacky
[03:36] <snover> but maybe youll get lucky and the y-coord is actually set in a script
[03:36] <snover> since the computer seems to use all dynamic objects, i am cautiously optimistic that this is the case
[03:38] <dafioram> i will check it out, tomorrow
[03:38] <snover> ok
[03:38] <snover> gonna find more titanic bugs now instead? ;)
[03:39] <dafioram> I think titanic is significantly more bug free than the original game
[03:39] <snover> i imagine so!
[03:39] <snover> dreammaster has done a lot of hard work to get it there.
[03:39] <dafioram> he kept fixing them so I kept testing
[03:41] <dafioram> any unknown bugs would have to be simple graphic glitches or some ultra complicated logic bug
[03:41] <snover> or a memory bug :) i need to get back to it soon.
[03:42] <dafioram> titanic doesn't use much ram I think some1s computer would have to be really limited for that to be a major issue
[03:42] <dafioram> I think the cpu is more limiting for low end machines
[03:43] <snover> any memory leak is a bad memory leak, but more importantly, any use-after-free is a potential crash
[03:43] <snover> and any stack or heap overflow is a potential data corruption
[03:43] <snover> its really sad when those memory bugs end up with a corrupt save game
[03:44] <dafioram> i haven't run into that problem
[03:45] <dafioram> we've found some issues with the original save game saving bad data at spots
[03:45] <dafioram> but we've fixed that
[03:46] <snover> i am sure Bethesda devs never experienced corrupt save games during testing, and yet :)
[03:47] <snover> anyway, i am not saying thats going to happen, im just saying that *can* happen and ive caught a few different memory corruption issues so far
[03:48] <dafioram> what was getting corrupted?
[03:49] <snover> in one case, when you upgraded to first class, the wrong variable was used for array access when replacing the colour of text in the PET so it overflowed
[03:49] <snover> i cant remember if it was a read or write overflow
[03:50] <dafioram> did it pick some wrong color?
[03:50] <snover> i dont know what the outcome was, it triggered an ASan dump
[03:50] <snover> so i just fixed it
[03:51] <dafioram> well anytime you want to spend improving titanic will be for the better and 10113 is without an owner...
[03:52] <snover> but do you want titanic on an rpi3 more than you want to play lighthouse?
[03:53] <dafioram> you can't do that
[03:53] --> Begasus joined #scummvm.
[03:54] <dafioram> i want lighthouse running in titanic on an rpi3
[03:56] <snover> :)
[03:56] <snover> its on one of those other CDs at the start of the game
[03:57] --> Begas_VBox joined #scummvm.
[03:57] <dafioram> For KQ7, I don't have 1.51 nor german version, but I will try to reproduce 9790, soon.
[03:59] <dafioram> It looks like 9746, might be not too bad also
[04:00] <snover> i have not been able to reproduce 9746 since the first day i reproduced it
[04:00] <snover> i doubt that thats because the bug was fixed
[04:01] <snover> i think its more like your crasher bug where if you do things in exactly the wrong way an unrelated earlier wrong thing makes things explode
[04:01] <dafioram> random bugs are the worst
[04:02] <snover> they really are.
[04:03] <dafioram> for #9846 could you see what path the code takes after clicking the map button and inject that after the code after when you click on the car?
[04:04] <snover> the problem is that clicking on the car works fine in lo-res mode but they stick an extra view for the car in hi-res mode
[04:04] <snover> and that view is blocking the car feature from responding to clicks for some reason
[04:04] <dafioram> im sure they made up for it with easter eggs
[04:05] <snover> i dont know that pq4 has *any* easter eggs
[04:06] <snover> apparently they had time for a funny message and some topical high score names in the mini-games at Short Stop and thats it
[04:07] <dafioram> could you make the vpl be the low res car and the pl be the hi res car?
[04:08] <snover> the visual plane list is the list of stuff that is rendered to the output buffer already, and the regular plane list is the list of stuff that is going to be rendered to the output buffer the next time GfxFrameout::frameOut is called.
[04:09] <snover> when that call happens then the two are the same.
[04:11] <dafioram> its a minor bug anyways
[04:12] <snover> most of them are now, thankfully! except for the pathfinding ones.
[04:12] <snover> and some of the midi ones are pretty bad too.
[04:13] <dafioram> the pathfinding ones arn't really gaming limiting though? they break immersion a bit?
[04:13] <snover> well you did get stuck in a locked room once&
[04:14] <dafioram> it was a lot more than once
[04:14] <snover> the pathfinding bugs are bad enough in qfg4 that its a game that wont get any release until the pathfinding is fixed
[04:16] <dafioram> are the sci games using the same pathfinding algos?
[04:16] <snover> yes, though i am sure with modifications over time
[04:16] <snover> sierra patented their pathfinding
[04:17] <snover> so at the time when freesci was written it was still under patent
[04:17] <snover> (obligatory software patents are dumb here)
[04:19] <dafioram> yup
[04:19] <snover> so scummvm inherited a subtly incompatible pathfinding system by necessity
[04:19] <dafioram> any chance we could individual make workarounds for the worse pathfinding issues and put off the rewrite till later?
[04:19] <snover> no, thats what was done up to this point
[04:19] <snover> that path has reached its termius
[04:20] <snover> terminus*
[04:21] <snover> same thing with the midi code. it needs to be reverse-engineered from the original interpreters to fix bugs caused by the guesswork and sierras slightly, er, non-standard Standard MIDI File parser
[04:22] <dafioram> the pathfinding rewrite could break previous games that have already done well in testing. From a scummvm release user standpoint this would extend the time between the next release with only a minor improvement in pathfinding from their view.
[04:22] <snover> its not getting done this release cycle.
[04:25] <dafioram> so for 10034 I would essentially be centering the text vertically within the Author and filename spaces
[04:25] <snover> i guess so, it is your report :)
[04:27] <dafioram> alright bed time
[04:28] <-- dafioram left irc: Quit: Leaving
[05:21] --> GitHub80 joined #scummvm.
[05:21] <GitHub80> [scummvm] Schallaven opened pull request #1021: Corrected spelling in xeen/map.cpp (master...patch-1) https://git.io/v5dD9
[05:21] GitHub80 (GitHub80@192.30.252.34) left #scummvm.
[05:32] Nick change: Stormkeeper -> Storm-AFK
[06:25] <-- Begas_VBox left irc: Quit: Vision[0.10.1]: i've been blurred!
[06:42] --> macdude22 joined #scummvm.
[06:54] --> m_kiewitz joined #scummvm.
[06:54] #scummvm: mode change '+o m_kiewitz' by ChanServ!ChanServ@services.
[08:19] --> Henke37 joined #scummvm.
[08:54] <-- LittleToonCat left irc: Remote host closed the connection
[10:52] --> girafe joined #scummvm.
[11:14] --> l0ftyWhizZ joined #scummvm.
[11:14] <l0ftyWhizZ> Hey guys! Hope everyone's having a good day! :D
[11:15] <l0ftyWhizZ> I just wanted to ask whether a NES engine has already been integrated or not.
[11:23] <Henke37> no, I don't think there will ever be a nes emulator included in scummvm
[11:23] <Henke37> try bizhawk instead
[11:26] <l0ftyWhizZ> Why is that so? Is it because Scumm focusses on Point-Click style games only?
[11:30] <somaen> Fairly out-of-scope yes
[11:31] <somaen> We don't in general emulate hardware platforms
[11:38] --> waltervn joined #scummvm.
[11:38] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[12:19] --> dafioram joined #scummvm.
[12:29] --> ajax16384 joined #scummvm.
[12:29] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[12:43] --> GitHub28 joined #scummvm.
[12:43] <GitHub28> [scummvm] m-kiewitz pushed 1 new commit to master: https://git.io/v5Fe3
[12:43] <GitHub28> scummvm/master 294503f Martin Kiewitz: SCI32: GK1CD: script for vine swing issue bug #9820...
[12:43] GitHub28 (GitHub28@192.30.252.42) left #scummvm.
[12:57] <-- Cheeseness left irc: Quit: Leaving.
[13:07] --> Strangerke_ joined #scummvm.
[13:09] <-- Strangerke left irc: Ping timeout: 240 seconds
[13:09] Nick change: Strangerke_ -> Strangerke
[13:38] --> jamm joined #scummvm.
[13:38] <-- jamm left irc: Changing host
[13:38] --> jamm joined #scummvm.
[13:48] <dafioram> m_kiewitz: what are your thoughts on #10034?
[13:51] <m_kiewitz> dafioram: is it only a problem for the "From" and "Subject" fields?
[13:52] <m_kiewitz> I guess changing coordinates of those shouldn't be that difficult
[13:52] <dafioram> no also File/Note, but that is very similar
[13:53] <dafioram> If I do the easy change of shifting all text fields up a bit to make the words with low handing letters like "g" then it will make currently centered words be uncentered
[13:54] <dafioram> so I am wondering if its worth doing it
[13:54] <m_kiewitz> right now File
[13:54] <m_kiewitz> and Note are stuck on the bottom
[13:54] <dafioram> ?
[13:54] <m_kiewitz> I guess changing those to be a bit higher would look fine regardless
[13:55] <m_kiewitz> https://bugs.scummvm.org/attachment/ticket/10034/scummvm00005.png
[13:55] <m_kiewitz> I doubt it would look bad in case we placed "Foreign Documentation on Product" 2 or 3 pixels upwards
[13:56] <dafioram> yes that case would look good, but what would other cases that are already centered, like the other attachment in that report?
[13:57] <dafioram> Career Opportunities would now be above vertical center
[13:57] <m_kiewitz> I wouldn't change the font itself
[13:58] <dafioram> the text y position?
[13:58] <m_kiewitz> that specific position
[13:59] <dafioram> what do you mean the font?
[13:59] <m_kiewitz> I wouldn't change the placement of all the text
[13:59] <dafioram> just ones with low handing letters in it?
[13:59] <m_kiewitz> just for example the File + Note text
[14:00] <dafioram> I was just proposing the File + Note text not the document text or any of that
[14:00] <m_kiewitz> File + Note should look fine in any case when placed 2 or 3 pixels upwards
[14:00] <m_kiewitz> that's what I'm talking about
[14:00] <dafioram> alright sounds good
[14:05] <-- Begasus left irc: Ping timeout: 240 seconds
[14:11] <m_kiewitz> dafioram: I need to look into dual mode (speech + subtitles at the same time) anyway
[14:11] <dafioram> i was hoping I could do this one
[14:11] <m_kiewitz> there is a patch, but it's basically all the scripts of the game and it also forces the game to have those 2 all the time and it also breaks saved game compatibility
[14:11] <m_kiewitz> so I'm looking into doing that properly
[14:12] <dafioram> I can do another one if it makes sense for you to multitask #10034 at the same time
[14:12] <m_kiewitz> ah wait, I think I looked into Gabriel Knight 2 dual mode
[14:13] <m_kiewitz> anyway for one of those there is already a "patch", but as i said - it's all scripts from the game and it breaks saved games
[14:13] <dafioram> i dont understand
[14:14] <m_kiewitz> Various Sierra games do not support speech + text at the same time
[14:14] <m_kiewitz> Gabriel Knight 2 is one of them
[14:14] <dafioram> yes
[14:14] <dafioram> ok
[14:14] <m_kiewitz> and some fans created a "patch", but that patch consists of every single script of the game, which is definitely not needed
[14:15] <m_kiewitz> and the modified scripts also cause saved games not work anymore with the patched version
[14:15] <m_kiewitz> Where can I use the computer in Phantasmagoria 2?
[14:15] <m_kiewitz> I don't have any saved games for that game
[14:16] <dafioram> grab the one from #10039
[14:17] --> Begasus joined #scummvm.
[14:19] <m_kiewitz> ah nice, thx
[14:28] <m_kiewitz> dafioram: can't use it, I got a different game version
[14:28] <m_kiewitz> also audio seems to be quite broken atm
[14:28] <dafioram> how so?
[14:28] <m_kiewitz> Phantasmagoria 2 main menu music is stuttering
[14:28] <m_kiewitz> and I can't even exit ScummVM
[14:29] <m_kiewitz> hmm weird, now it works. It seems trying to restore a saved game from a different game version makes the engine go crazy
[14:29] <m_kiewitz> well, i will just play to that point
[14:30] --> dreammaster joined #scummvm.
[14:30] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[14:34] --> GitHub58 joined #scummvm.
[14:34] <GitHub58> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5FTq
[14:34] <GitHub58> scummvm/master e1a84e2 Paul Gilbert: DEVTOOLS: Add other German bot sentence mappings to create_titanic
[14:34] GitHub58 (GitHub58@192.30.252.34) left #scummvm.
[15:05] --> Begas_VBox joined #scummvm.
[15:21] --> GitHub161 joined #scummvm.
[15:21] <GitHub161> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5FIK
[15:21] <GitHub161> scummvm/master fe83d2b Paul Gilbert: TITANIC: Fixes for sounds and Doorbot during German prologue
[15:21] GitHub161 (GitHub161@192.30.252.42) left #scummvm.
[15:27] <snover> dafioram: the movement of the lines needs to be just like one pixel, in those mockups the text is moved up too much
[15:27] <dafioram> i think m_kiewitz might be looking into it
[15:28] <m_kiewitz> yes, I just reached the computer
[15:31] <m_kiewitz> 2 or 3 pixels, moving it just by 1 still makes part of font go outside of the box
[15:38] --> criezy joined #scummvm.
[15:38] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[15:42] --> Littleboy joined #scummvm.
[15:42] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[15:43] --> GitHub19 joined #scummvm.
[15:43] <GitHub19> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5FLQ
[15:43] <GitHub19> scummvm/master 6ef5948 Paul Gilbert: TITANIC: German sound fixes for rest of prologue
[15:43] GitHub19 (GitHub19@192.30.252.35) left #scummvm.
[15:52] --> ny00123 joined #scummvm.
[15:55] <-- _sev_ left irc: Quit: This computer has gone to sleep
[15:56] --> _sev_ joined #scummvm.
[15:56] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[15:58] <-- dreammaster left irc:
[15:59] <dafioram> snover: any hints on how phant1 sets the cursor color? In the game it turns red when you have an action
[16:02] <snover> the call to the kernel to change the cursor is kSetCursor, and often the games put some different cursor objects on various VM globals and switch between those. the kernel call for the game to learn if a mouse is on a view/feature is kIsOnMe, which is usually invoked indirectly by a system script during the games main event loop
[16:04] <snover> i *think* that the list of objects scanned for hit detection is in the cast global (global 5)
[16:04] <snover> you can get a list of what all the globals are by just using `vv g`
[16:04] <dafioram> does it make sense that I don't see a plan object entry for the cursor?
[16:05] <snover> yes, the cursor is not drawn by the game
[16:06] <snover> in windows ssci the cursor is drawn by windows, in dos ssci the kernel draws it directly to vram because it responds via interrupt so cannot participate in the normal rendering loop
[16:08] <-- Littleboy left irc: Read error: Connection reset by peer
[16:08] --> Littleboy joined #scummvm.
[16:08] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[16:12] <dafioram> should snk kSetCursor work?
[16:12] --> K4T joined #scummvm.
[16:13] <snover> i mean, yes? i dont quite understand the context of that question :)
[16:15] <dafioram> I'm tring to see how/when the cursor color is getting set so I am trying to break around when it does that
[16:15] <snover> oh, i wonder if snk doesnt have the error detection of bpk
[16:15] <snover> it should be snk SetCursor
[16:16] <dafioram> yes
[16:16] <snover> or bpk SetCursor, depending upon if your goal is to break once or break every time
[16:18] <m_kiewitz> dafioram: this is okay?
[16:18] <m_kiewitz> https://i.imgur.com/qjjNowh.png
[16:19] <dafioram> can do do up 1 or 2 more? for that "g"
[16:23] <m_kiewitz> https://i.imgur.com/NrTNowF.png
[16:23] <m_kiewitz> I wouldn't want to put it any higher than that
[16:23] <snover> the two boxes have different heights
[16:23] <m_kiewitz> that's now -3
[16:23] <snover> so the top one needs to move more than the bottom one for them to appear visually correct
[16:23] <snover> if that is even possible&
[16:23] <m_kiewitz> it is
[16:24] <dafioram> it looks good
[16:24] <snover> i think the bottom one was correct before and the top one is correct now
[16:24] <m_kiewitz> yes, thats what i wanted to do just now
[16:24] <snover> therell still probably be problems with the descender in the lower box, but thats phant2 for you :)
[16:24] <m_kiewitz> very weird though with that height difference
[16:26] <m_kiewitz> https://i.imgur.com/lF2Blx3.png
[16:26] <m_kiewitz> done
[16:26] <dafioram> you shifted canada down?
[16:27] <m_kiewitz> yes, intentionally
[16:28] <m_kiewitz> the lower box (note) is 16 pixels, the upper box (file) is 17 pixels
[16:28] <dafioram> how does Emplyees/ArnoldB.doc look?
[16:29] <dafioram> and Email from TrevorB: Happy Monday!
[16:29] <m_kiewitz> https://i.imgur.com/IOZRdn0.png
[16:30] <m_kiewitz> https://i.imgur.com/YNtfJeg.png
[16:31] <dafioram> should that last one have the bottom shifted up one?
[16:31] <snover> love that mail interface consistency
[16:32] <dafioram> there a couple other emails with g/y in the subject line that probably still hang down
[16:32] <m_kiewitz> im not fixing that too :p
[16:33] <dafioram> how did you implement the fix so far?
[16:33] <dafioram> at a high level
[16:33] <m_kiewitz> well if I go one pixel up for subject/note it won't be consistent anymore, because of the pixel differences
[16:33] <m_kiewitz> via heap
[16:33] <m_kiewitz> heap patch
[16:34] <m_kiewitz> I could go up 1 pixel once more for both
[16:34] <m_kiewitz> but idk how that will look
[16:34] <dafioram> what looks worse hanging letters or non-vertically centered text?
[16:35] <m_kiewitz> at least it doesn't overlap anymore
[16:35] <m_kiewitz> to properly fix it one would have to change the background image
[16:41] <m_kiewitz> weird, it doesn't let me go any higher
[16:43] <-- K4T left irc:
[16:46] <-- l0ftyWhizZ left irc: Ping timeout: 260 seconds
[16:46] <m_kiewitz> ah, it seems there is another coordinate for the email stuff
[16:47] <m_kiewitz> this is a mock-up how it looks now, after going another pixel up for both fields
[16:47] <m_kiewitz> https://i.imgur.com/JKr1CUa.png
[16:48] <dafioram> what happened to the note text field?
[16:48] <m_kiewitz> I added a "g"
[16:48] <m_kiewitz> using paint shop
[16:48] <m_kiewitz> because there is no "g" anywhere (outside of emails) that I can access at that moment
[16:49] <dafioram> i think that looks good
[16:52] --> l0ftyWhizZ joined #scummvm.
[16:54] --> LittleToonCat joined #scummvm.
[16:57] <snover> m_kiewitz: did the view data you sent me for the pq4 audio mode toggle include a cel header or was it just the uncompressed pixels?
[16:58] <m_kiewitz> uncompressed bitmap data
[16:59] <snover> what are the dimensions?
[16:59] <snover> same as the existing ones in view 10982 loop 15?
[17:00] <snover> though actually i do need the dimensions& the cels in view 10982 loop 15 have different widths&
[17:03] <m_kiewitz> yes
[17:03] <m_kiewitz> moment, will check
[17:03] <m_kiewitz> width is 85
[17:04] <snover> ok, thanks!
[17:05] <m_kiewitz> dafioram: https://i.imgur.com/430Zqxt.png
[17:05] <m_kiewitz> I even fixed the inconsistent horizontal text placement for emails
[17:05] <m_kiewitz> hopefully there is no very long subject
[17:06] --> AndChat|594561 joined #scummvm.
[17:07] <dafioram> looks good
[17:08] <-- l0ftyWhizZ left irc: Ping timeout: 240 seconds
[17:08] <-- AndChat|594561 left irc: Read error: Connection reset by peer
[17:08] --> l0ftyWhizZ joined #scummvm.
[17:10] --> AndChat|594561 joined #scummvm.
[17:10] <-- l0ftyWhizZ left irc: Read error: Connection reset by peer
[17:10] --> l0ftyWhizZ joined #scummvm.
[17:12] --> GitHub177 joined #scummvm.
[17:12] <GitHub177> [scummvm] m-kiewitz pushed 1 new commit to master: https://git.io/v5F3I
[17:12] <GitHub177> scummvm/master 4a9098d Martin Kiewitz: SCI32: Heap patch for Phantasmagoria 2 to fix DocuStore placements...
[17:12] GitHub177 (GitHub177@192.30.252.35) left #scummvm.
[17:12] <m_kiewitz> text in the Japanese Phantasmagoria 2 is fully broken atm
[17:14] <snover> i would imagine so, the normal interpreter doesnt support wide characters
[17:14] <-- AndChat|594561 left irc: Ping timeout: 246 seconds
[17:15] <m_kiewitz> WARNING: font.7007 is missing glyph 143!
[17:15] <m_kiewitz> etc.
[17:15] <dafioram> snover: is there a bpw for kernel functions?
[17:15] <snover> no, that doesnt make sense, kernel calls arent objects
[17:17] <L0ngcat> snover: what does the Hi-quality video option do? GK2 looks great, I am just wandering what exactly this option does?
[17:17] <L0ngcat> *wondering
[17:17] <snover> L0ngcat: linear interpolation in rgb colour space
[17:17] <snover> rgb24*
[17:17] <L0ngcat> right
[17:18] <L0ngcat> sweet
[17:18] <L0ngcat> it looks great
[17:19] <L0ngcat> I've played through Phantasmagoria 1 & 2, Thorins Quest, King's Quest 7, Larry 7 and now GK2 on my macbook 12
[17:20] <L0ngcat> is this a vmd only thing or can it be applies to other games?
[17:20] <snover> which other games where?
[17:20] <L0ngcat> other game ssupported by scummvm i.e. other video formats
[17:20] <L0ngcat> -s
[17:21] <snover> its up to the engine to decide whether it can switch video modes
[17:21] <m_kiewitz> omg, what's going on
[17:21] <m_kiewitz> http://www.ebay.com/itm/Mixed-Up-Mother-Goose-Sierra-Roberta-Williams-IBM-PC-Big-Box-Floppies-/263199162409
[17:21] <L0ngcat> so it is possible?
[17:21] <m_kiewitz> $46 for a child's game and the auction is not even done
[17:22] <L0ngcat> my GK never looked this good
[17:22] <snover> L0ngcat: anything is possible, but theres no automatic process by which everything can be instantly updated
[17:22] <L0ngcat> or sounded this good
[17:22] <L0ngcat> this should be implemented as a standard feature
[17:22] <L0ngcat> m_kiewitz: lol. welcome to the world of capitalism
[17:23] <L0ngcat> the buyer sets the price ;)
[17:23] <m_kiewitz> well jokes on them, I got my AGI version for a few bucks only :p
[17:23] <L0ngcat> bah, agi is good for historical reason. I always pick VGA if it's an option
[17:24] <m_kiewitz> no no, the actual rare versions are SCI1 EGA
[17:24] <m_kiewitz> those were special releases with just EGA graphics, dithered down from VGA graphics
[17:24] <L0ngcat> I bet
[17:24] <m_kiewitz> I own 2 of them. Never managed to get any more
[17:24] <L0ngcat> I also imagine there were fewer copys released in EGA than in VGA versions of Sierra games?
[17:24] <m_kiewitz> at least I should own now a full collection of all Japanese SCI games
[17:25] <m_kiewitz> in Japan you can get those for relatively cheap, I don't want to know what those crazy collectors would pay
[17:25] <m_kiewitz> yes, of course
[17:25] <m_kiewitz> but it's really crazy collectors only
[17:25] <m_kiewitz> no one in their right mind would want to play the EGA versions
[17:26] <m_kiewitz> so the crazy prices are really paid for bad graphics
[17:26] <L0ngcat> I had aspirations of owning an all versions LucasArts adventure collection once. But I gave up. Just to expensive to get mint conditions
[17:26] <m_kiewitz> especially the FM-Towns ones :p
[17:26] <m_kiewitz> those aren't even cheap within Japan
[17:26] <L0ngcat> did you ever find a copy of Cranston Manor?
[17:27] <m_kiewitz> I think that's the one, yes
[17:27] <L0ngcat> yeah
[17:27] <m_kiewitz> ah yes
[17:27] <m_kiewitz> floppy only
[17:27] <m_kiewitz> we actually got 2 now
[17:27] <L0ngcat> fm-town is crazy expensive
[17:27] <m_kiewitz> 1 set is at sev
[17:27] <L0ngcat> nice
[17:27] <L0ngcat> good to hear
[17:27] <m_kiewitz> the other set is at dreammaster, who will send it to me at some point with all the other stuff lol
[17:28] <m_kiewitz> well floppies only was $65 for me
[17:28] <m_kiewitz> pretty crazy
[17:28] <L0ngcat> are all hi-res adventures fully supported now?
[17:28] <m_kiewitz> oh yeah BUT I may have gotten an actual master floppy
[17:28] <L0ngcat> jebus
[17:28] <L0ngcat> the show-case one?
[17:28] <m_kiewitz> no idea, not working on those. walter is the man to ask that
[17:28] <L0ngcat> nice!
[17:28] <m_kiewitz> http://thumbs.ebaystatic.com/d/l275/pict/253034062414_1.jpg
[17:29] <L0ngcat> waltervn: are alle hi-res adventures fully supported now?
[17:29] <m_kiewitz> who knows if it's even readable
[17:29] <m_kiewitz> at least the one that sev got was
[17:29] <L0ngcat> yeah, I remember you posting about it
[17:29] <L0ngcat> good going getting it
[17:29] <L0ngcat> a real treasure
[17:30] <L0ngcat> how do you read it to a modern system? usb? or real hardware?
[17:30] <L0ngcat> ring looks worn, though
[17:30] <L0ngcat> hope it works for you
[17:30] <m_kiewitz> no idea, sev will do that
[17:31] <m_kiewitz> i only do regular DOS games and such
[17:31] --> fydo joined #scummvm.
[17:31] <L0ngcat> :)
[17:31] <m_kiewitz> I even got Castle of Dr. Brain Japanese also waiting to get shipped to sev, because I can't read that one
[17:31] <L0ngcat> I wish I had all my floppies from the eighties still.. so many great systems
[17:31] <m_kiewitz> even getting a regular 5.25" drive is a pain now
[17:32] <m_kiewitz> If I would have known this, I would have bought 100 of them.
[17:32] <L0ngcat> jupp
[17:32] <L0ngcat> around 1995 would have been a good time to harvest vintage equipment and games
[17:33] <L0ngcat> I'm still crying I didn't
[17:33] <m_kiewitz> NES and SNES games too
[17:33] <m_kiewitz> what's really weird is that you can Famicom and Super-Famicom games for really cheap in Japan
[17:33] <m_kiewitz> *get
[17:33] <L0ngcat> mint copies, all mt-32 modules, floppy drives, memorobilia ..
[17:34] <L0ngcat> probably, bc there are a lot of them
[17:34] <L0ngcat> nintendo always sucked at releasing in the west
[17:35] <L0ngcat> not that I really care about nintendo
[17:35] <m_kiewitz> I would assume that there should be more western NES + SNES carts than Famicom + Super-Famicom carts :p
[17:35] <waltervn> L0ngcat: they're all working, but not officially supported yet. No one has tested them yet.
[17:35] <L0ngcat> few/to none adventure games
[17:35] <L0ngcat> but I like the graphics
[17:35] <L0ngcat> super Star Wars was pretty neat
[17:35] <waltervn> L0ngcat: and when I say "all", I mean Apple II versions only
[17:35] <m_kiewitz> waltervn: I guess no one owns them heh (except for the few that are on the Roberta Williams collection, and even that is quite rare)
[17:36] <L0ngcat> waltervn: great, thanx for the feedback! (and the hard work)
[17:36] <waltervn> maybe I should reach out to the people that tested the last batch
[17:36] <snover> it will be interesting to see how much longer floppy data remains viable
[17:36] <L0ngcat> m_kiewitz: is Cranston Manor not in the Roberta Williams collection? (I have it lying around here, somewhere)
[17:37] <m_kiewitz> nope
[17:37] <snover> (in terms of being able to still retrieve a signal off the media)
[17:37] <-- balrog left irc: Ping timeout: 240 seconds
[17:37] <L0ngcat> aww
[17:37] <L0ngcat> my ScummVm Collection is so screwed
[17:37] <m_kiewitz> Mission: Asteroid, Mystery House, Wizard and the Princess, Time Zone, Dark Crystal
[17:37] <m_kiewitz> that's it
[17:37] <L0ngcat> how much for a copy?
[17:37] <m_kiewitz> Roberta Williams?
[17:38] <L0ngcat> Cranston Manor?
[17:38] <m_kiewitz> a boxed one?
[17:38] <L0ngcat> I have the RW collection
[17:38] <m_kiewitz> my floppy was $65 as I said
[17:38] <L0ngcat> a working one
[17:38] <m_kiewitz> yes, I got 2 now
[17:38] <m_kiewitz> Had it without box for ages
[17:39] <snover> those apple images were on the KQ Collection 1997 too
[17:39] <L0ngcat> I do not care about thephysical media, only the data
[17:39] <m_kiewitz> and I just got lucky and got one CIB mint for $28
[17:39] <m_kiewitz> well $65 as I said
[17:39] <m_kiewitz> and that was lucky
[17:39] <m_kiewitz> I got a website watcher, that checks ebay for certain items every few minutes
[17:39] <L0ngcat> snover: cranston mano? I have that collection as well
[17:39] <snover> no
[17:39] <snover> the five m_kiewitz said
[17:39] <m_kiewitz> that way I find it earlier than most people, which is why I was able to get it
[17:40] <m_kiewitz> cranston manor is not on that collection
[17:40] <waltervn> Dark Crystal is missing in some sets
[17:40] <L0ngcat> ah
[17:40] <m_kiewitz> ah, it is? wow
[17:40] <m_kiewitz> need to put that on the wiki
[17:40] <L0ngcat> god, that is expensive for a game that I am probably never going to play
[17:40] <m_kiewitz> well that was actually cheap
[17:40] <L0ngcat> oh well
[17:40] <m_kiewitz> normally it's $300 or $400
[17:40] <m_kiewitz> lol
[17:40] <L0ngcat> crap
[17:40] <m_kiewitz> as I said - I got very lucky
[17:41] <m_kiewitz> especially with the master disc
[17:41] <L0ngcat> how much will you sell it for? ;)
[17:41] <m_kiewitz> lol
[17:41] <m_kiewitz> never
[17:41] <L0ngcat> bah
[17:41] <waltervn> snover: so it's 4 games on the KQ collection
[17:41] <L0ngcat> see?
[17:41] <m_kiewitz> :p
[17:41] <L0ngcat> I'm screwed
[17:41] <L0ngcat> ScummVM
[17:42] <L0ngcat> should be banned from supporting such rare games. Getting Bambou nearly killed me
[17:42] <snover> ok. i guess i cant count :)
[17:42] <m_kiewitz> yes, KQ collection seems to miss out on Dark Crystal
[17:42] <waltervn> Roberta Williams Anthology is the only one with Dark Crystal
[17:42] <m_kiewitz> No idea why they didn't put everything on a CD
[17:42] <L0ngcat> at least I have that
[17:42] <m_kiewitz> or well, nowadays, why not sell them on GOG or Steam
[17:42] <waltervn> for Dark Crystal it's probably a licensing issue
[17:42] <m_kiewitz> absolutely idiotic
[17:42] <L0ngcat> ask Activision
[17:42] <m_kiewitz> yeah right, movie
[17:42] <waltervn> I don't think Sierra owns that game
[17:43] <m_kiewitz> I almost missed the steelbook for Dark Crystal
[17:43] <L0ngcat> are they not owning the ip?
[17:43] <m_kiewitz> Dark Crystal is a movie
[17:43] <m_kiewitz> and possibly also a book, idk
[17:43] <waltervn> my impression is that the actual game is Hensons as well
[17:43] <L0ngcat> Dark Crystal was one of the Disney licensed games
[17:43] <L0ngcat> from the movie, yes
[17:43] <m_kiewitz> ah right, I saw Donald Duck for $80 or so
[17:43] <L0ngcat> pretty good movie, actually
[17:44] <L0ngcat> m_kiewitz: steelbook?
[17:44] <m_kiewitz> for the movie
[17:44] <m_kiewitz> embossed even
[17:44] <L0ngcat> aluminium case?
[17:44] <m_kiewitz> https://www.youtube.com/watch?v=-arS-1FZdbA
[17:44] <L0ngcat> oh
[17:44] <L0ngcat> blu-ray
[17:44] <m_kiewitz> sold out in UK
[17:44] <L0ngcat> so futuristic
[17:44] <L0ngcat> I thought dvd
[17:45] <m_kiewitz> Germany had a few left and Amazon lowered the price to 8 EUR or so
[17:45] <waltervn> there's a web-based version of Dark Crystal at http://www.darkcrystal.com/play/index.php, which I think is further evidence that the game is fully owned by Henson
[17:45] <L0ngcat> nice box
[17:45] <L0ngcat> whut
[17:45] <L0ngcat> this is VGA graphics
[17:46] <L0ngcat> thanx!
[17:46] <L0ngcat> I am actually going to play the this
[17:46] <m_kiewitz> quite nice that they put it up on the web
[17:46] <waltervn> if you look at page 2 of the manual linked on that page, there's no (c) sierra, only henson
[17:47] <waltervn> it only says "Produced by Sierra On-Line"
[17:47] <L0ngcat> so they own the ip?
[17:47] <waltervn> Henson definitely owns the IP, but I believe they also own the game itself
[17:47] <waltervn> which is why it it is missing from the KQ collection
[17:47] <m_kiewitz> so maybe they put it on the RW by accident
[17:48] <waltervn> that did cross my mind as well, but it's speculation :P
[17:57] --> balrog joined #scummvm.
[18:04] <-- Littleboy left irc: Read error: Connection reset by peer
[18:12] <L0ngcat> funfunfun
[18:14] --> Strangerke_ joined #scummvm.
[18:16] <-- Strangerke left irc: Ping timeout: 240 seconds
[18:16] Nick change: Strangerke_ -> Strangerke
[18:18] <snover> m_kiewitz: can the speech+subtitles toggle cel be cel 3 of the loop instead of cel 0 of a separate loop?
[18:19] <m_kiewitz> in theory yes, but when you select a button it sets cel 1 and afterwards it goes back to cel 0
[18:19] <m_kiewitz> so effectively no
[18:19] <m_kiewitz> but you may also create a new view altogether
[18:19] <m_kiewitz> would have to adjust the code a bit then though
[18:19] <snover> creating a new view would be much easier than trying to add a new loop to an existing view.
[18:20] <m_kiewitz> then do that
[18:20] <snover> ok
[18:20] <m_kiewitz> will adjust the code after the injection code is in there
[18:20] <m_kiewitz> right now it would probably error() out, idk
[18:20] <m_kiewitz> no idea if you need to create a cel 1 as well
[18:21] <m_kiewitz> you can just point to the same data
[18:24] <m_kiewitz> if you decompressed the data and put it into tables like the SCI16 view class, it would be trivial to inject raw bitmap data
[18:25] <snover> im not going to do that, thanks though.
[18:29] <dafioram> snover: what does it mean for the selector to be bad for disasmed code?
[18:30] <snover> it means that number is definitely not a selector :)
[18:31] <dafioram> so that doesn't crash the game?
[18:32] <dafioram> theres 8 bad selector calls in this disasmed function from the base game
[18:33] <snover> the disassembler does not have the knowledge to know if a number is a number or a selector. writing the selectors from the selector vocab next to numbers in the disassembly is just a convenience for the person reading the disassembly. it doesnt mean anything is thinking or using them as selectors.
[18:34] <dafioram> alright so its an object
[18:35] <snover> if a game were to try to use a number as a selector which was not a selector, it would indeed crash the game.
[18:36] <-- freanux left irc: Quit: Leaving Circada 0.4.1
[18:43] <-- Begas_VBox left irc: Quit: Vision[0.10.1]: i've been blurred!
[18:44] <-- Begasus left irc: Quit: Ex-Chat
[18:54] --> Farmboy0 joined #scummvm.
[18:54] <-- Farmboy0 left irc: Changing host
[18:54] --> Farmboy0 joined #scummvm.
[18:58] <criezy> snover: before starting looking at the code I am playing with your SDL PR.
[18:58] <criezy> I see we can now resize the window with SDL graphics mode. But is that on purpose that is behaves differently from OpenGL mode?
[18:58] --> freanux joined #scummvm.
[18:58] <criezy> With OpenGL it keeps the correct aspect ratio and adds black bars. WIth SDL it stretches the picture.
[18:59] <snover> criezy: its not very intentional that the GUI isnt full window size, i just didnt have motivation to figure out how to make that happen for the surface renderer
[19:00] <snover> and being able to resize is better than not, even though the launcher GUI doesnt look perfect
[19:00] <criezy> I was actually thinking of in game. But indeed now that you wrote it I just noticed that the GUI has blck bars with the surface renderer.
[19:01] <snover> wait, the game area is being stretched?
[19:01] <criezy> Yes
[19:01] <criezy> With the surface renderer.
[19:01] <criezy> With OpenGL it has black bars.
[19:01] <snover> they use the same resizing code so im not sure offhand how that could even happen
[19:02] <snover> what game are you testing with?
[19:02] <criezy> Dreamweb.
[19:02] <criezy> I just noticed that with Drascula I do get black bars.
[19:03] <snover> are you using your aspect ratio calculation patch at the same time?
[19:03] <criezy> No. I am only using your PR with no other modification.
[19:03] <snover> let me make a build with dreamweb engine enabled, one second.
[19:04] <criezy> And now I am getting black bars with dreamweb as well and the surface renderer....
[19:05] <criezy> I think there is something a bit strange with the way it handles aspect ratio maybe.
[19:05] <snover> macos 10.12 and sdl2, aspect ratio correction disabled, the pc cd version of dreamweb?
[19:05] <criezy> Yes, Yes, AR enabled, Yes
[19:06] <criezy> No it is actually correct when I resize.
[19:06] <criezy> I was foole by the fact it has black bars by default when we get to the game proper now.
[19:06] <snover> haha. ok. phew!
[19:06] <criezy> Because it no longer resize the window after the intro
[19:06] <snover> giving me a heart attack here :)
[19:08] <snover> nice, right? no more having to re-resize at the start
[19:10] <criezy> It's nice to have the filter option has well, because resizing a x2 window gives something a bit ugly in the GUI otherwise :P
[19:11] <criezy> *as well
[19:12] <criezy> Is there a way, such as a keyboard shortcut, to get back to the default window size?
[19:13] <snover> i was thinking of maybe using ctrl+option+0 for that, right now you can ctrl+option+plus and then ctrl+option+minus
[19:14] <criezy> Ha right. Those same shortcuts I have always have issue with...
[19:14] <criezy> I can use the crtl+option+minus OK, but crtl+option+plus doesn't work, because to get the plus key I need to use the shift key as well.
[19:15] <criezy> And for the window size it doesn't go full circle apparently. So I can only reduce the window size but not increase it :P
[19:15] <snover> ctrl
[19:15] <snover> oops
[19:15] <snover> ctrl+option+equals does not work on your system?
[19:16] <criezy> Oh it does!
[19:16] <criezy> I though I tried it before...
[19:16] <snover> sorry, i wasnt thinking when i said plus :)
[19:16] <criezy> That doesn't play very well with the GUI options though...
[19:18] <snover> i didnt change any of these keyboard shortcuts so the weirdness with them is not the fault of these backend changes, i dont *think*.
[19:20] <criezy> I would not be so sure, part of it might be related to the changes.
[19:20] <criezy> The part where it doesn't resize the window now when you change from normal to x2 or x3 or the other way around.
[19:20] <criezy> (change the raphics mode in the GUI dialog)
[19:20] <snover> you are talking about when you are at the GUI and not in-game?
[19:20] <criezy> Yes
[19:22] <criezy> So if you start in X2 by default, change it to x3, it now displays a x3 GUI but with the window still in 640x480. So the text is very small.
[19:22] <snover> hmm. my window size seems to change ok.
[19:23] <snover> what filter mode are you using? normal2x?
[19:23] <criezy> Yes. I am just playing with "normal", "2x" and "3x" for now.
[19:23] <snover> hmm. i randomly got it to not work once after pressing a bunch of buttons
[19:24] <snover> oh, that was when i switched aspect ratio off
[19:24] <snover> OH
[19:24] <snover> i see what youre talking about.
[19:24] <snover> when you are changing it from the UI and not the keyboard shortcuts.
[19:24] <snover> ok.
[19:24] <snover> yes.
[19:24] <snover> good bug!
[19:25] --> dreammaster joined #scummvm.
[19:25] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[19:25] <criezy> Yes, I am changing it with the Options dialog in the UI now.
[19:25] <criezy> With the shortcuts it seems to work properly.
[19:26] <snover> i guess not suppressing programmatic window size changes when the launcher is visible is probably the way to go
[19:27] <-- jamm left irc: Ping timeout: 246 seconds
[19:35] <snover> oh. that requires the backend refactoring so that property is available. heh heh.
[19:41] <snover> i tried just unlocking the window size when a setGraphicsMode call happens but that isnt great for games that swap between forceNormal1x and a scaler mode, like dreamweb
[19:43] <-- dafioram left irc: Quit: Leaving
[19:43] <snover> im not sure what the best solution there is since the first mode that dreamweb switches to is the 640x480 normal1x mode
[19:44] <snover> this is fine for 1x or 2x, but 3x is larger so will get downscaled
[19:45] <snover> starting at 1x has a similar problem, the 640x480 mode gets downscaled
[19:46] <criezy> I just realized that using dreamweb to compare the mouse behaviour (droping events and hiding the game cursor) might not have been a good idea since it is not impacted by the change :P
[19:46] <criezy> I am assuming it is because it doesn't use CursorMan.
[19:47] <snover> ah, yeah. it seems it needs the sci32 treatment to stop calling to hide the system cursor
[19:48] <snover> the backends would previously draw either the previous mouse cursor (surface), or garbage/crash (opengl), if you didnt provide a non-null cursor data to CursorMan
[19:48] <criezy> I have a nice bug now where clicking on any game in the list reset the selection to the first game :/
[19:48] <snover> i fixed that, but now the engines need to also be fixed to stop hiding the cursor when they dont use it
[19:48] <criezy> I can use the arrow though to select other games...
[19:48] <snover> modern gui?
[19:49] <criezy> Yes.
[19:49] <criezy> That happened after using Crtl+option+= to switch to x3.
[19:49] <criezy> But was fixes by switching back to x2.
[19:49] <criezy> And now when I switch to x3 I can't reproduce...
[19:51] <snover> i can reproduce it
[19:52] <snover> its an input bug
[19:52] <snover> if you hold down your mouse button and then hit ctrl+option+whatever it breaks
[19:53] <snover> i think the gui thinks you are holding down the whatever-key and so it is constantly selecting the nearest title starting with =
[19:54] <snover> so just keep your fist off the trackpad
[19:55] <criezy> I don't that's what happened with me.
[19:55] <criezy> I am not on a laptop and I don't have a trackpad.
[19:55] <snover> hm.
[19:56] <criezy> Clicking on the mouse button and using crtl+option+whatever requires some contortion and I don't think I would have done that without meaning to.
[19:56] <snover> yeah, i am using a separate mouse and my hand positioning was a little crazy
[19:56] <snover> but i guess if it can happen in that situation, it could happen in some other situation if the input was dropped or consumed in some unexpected manner
[19:57] <snover> oh, now i cant reproduce it with my contortion
[19:58] <snover> well, at least i think its probably due to the = key input being persisted for some reason
[19:59] <criezy> I think I have a way to reproduce: start from 2x. Go to GUI dialog change to 3x and click OK. Now use crtl+option+=
[20:03] <criezy> It also happens with master with the same steps.
[20:03] <criezy> So at least it is not related to your change ;)
[20:16] <-- Lightkey left irc: Ping timeout: 246 seconds
[20:20] --> AndChat|594561 joined #scummvm.
[20:20] <-- l0ftyWhizZ left irc: Read error: Connection reset by peer
[20:21] --> l0ftyWhizZ joined #scummvm.
[20:21] <-- AndChat|594561 left irc: Read error: Connection reset by peer
[20:25] --> GitHub136 joined #scummvm.
[20:25] <GitHub136> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5F4k
[20:25] <GitHub136> scummvm/master 481e18e Paul Gilbert: TITANIC: Create a separate g_language global in it's own file...
[20:25] GitHub136 (GitHub136@192.30.252.41) left #scummvm.
[20:27] --> K4T joined #scummvm.
[20:27] <-- K4T left irc: Client Quit
[20:28] <ScummBot> Port build status changed with 481e18ee: Failure: master-ios7, master-mingw-w32, master-openpandora, master-android_mips, master-android_x86, master-ios, master-wii, master-debian-x86, master-mingw-w64-cplusplus11, master-webos, master-osx_intel, master-psp2, master-gp2x, master-dingux, master-debian-x86-clang, master-gamecube, master-gp2xwiz, master-android_arm,
[20:28] <ScummBot> master-mingw-w64, master-deb
[20:28] <ScummBot> ian-x86_64
[20:29] --> AndChat|594561 joined #scummvm.
[20:30] --> Lightkey joined #scummvm.
[20:30] <dreammaster> Ny bad. Missed actually adding the new file
[20:30] <-- AndChat|594561 left irc: Read error: Connection reset by peer
[20:30] --> AndChat-594561 joined #scummvm.
[20:31] <Lightkey> eh.. what was that?
[20:31] <Lightkey> dreammaster: its*
[20:31] <criezy> That didn't stop some of the ports to compile successfully :P
[20:31] --> GitHub179 joined #scummvm.
[20:31] <GitHub179> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5F4C
[20:31] <GitHub179> scummvm/master 0b258ae Paul Gilbert: TITANIC: Add translation.h missing from prior commit
[20:31] GitHub179 (GitHub179@192.30.252.35) left #scummvm.
[20:32] <dreammaster> I added a new translation.h to basically hold a language global and a macro for specifying the english/german pairs of values
[20:32] <Lightkey> hm, more bugs in Darksaber2k's run of I Have No Brain, and I Must Stream that haven't been reported, when Ted opens the maid's room there is an annoying visual and audio glitch every time and when Gorrister pulls the lever after stopping the engine it still says that he has the animals' blood on his hands when taking the key, even though they are alive
[20:33] <-- l0ftyWhizZ left irc: Ping timeout: 240 seconds
[20:36] --> travis-ci joined #scummvm.
[20:36] <travis-ci> scummvm/scummvm#4435 (master - 481e18e : Paul Gilbert): The build was broken.
[20:36] <travis-ci> Change view : https://github.com/scummvm/scummvm/compare/4a9098d08368...481e18ee1d38
[20:36] <travis-ci> Build details : https://travis-ci.org/scummvm/scummvm/builds/276606776
[20:36] travis-ci (travis-ci@ec2-54-167-99-251.compute-1.amazonaws.com) left #scummvm.
[20:45] <-- P2E left irc: Read error: No route to host
[20:47] --> P2E joined #scummvm.
[20:48] <snover> dreammaster: i think i maybe axed this once before, is the build-by-default supposed to be true for titanic engine now that it is in testing?
[20:48] <snover> in engine.configure or whateverit is?
[20:48] <ScummBot> Port build status changed with 0b258ae1: Success: master-ios7, master-mingw-w32, master-openpandora, master-android_mips, master-android_x86, master-ios, master-wii, master-debian-x86, master-mingw-w64-cplusplus11, master-webos, master-osx_intel, master-psp2, master-gp2x, master-dingux, master-debian-x86-clang, master-gamecube, master-gp2xwiz, master-android_arm,
[20:48] <ScummBot> master-mingw-w64, master-deb
[20:48] <ScummBot> ian-x86_64
[20:49] <dreammaster> Actually, that's a good question. You'd think after all these engines I've worked on, I'd remember. But I keep forgetting. I *think* we switch it to build by default at the same time testing ends.
[20:50] <dreammaster> That and remove the ADGF_TESTING at the same time. I guess it could be done at any time now. We don't exactly have to wait until your cursor changes are merged in
[20:51] <dreammaster> Maybe someone else can chime in to reassure me?
[20:51] <-- waltervn left irc: Quit: Leaving
[20:52] --> GitHub45 joined #scummvm.
[20:52] <GitHub45> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5FBa
[20:52] <GitHub45> scummvm/master 42b24e3 Paul Gilbert: TITANIC: Add skeleton derived bot script classes for German version
[20:52] GitHub45 (GitHub45@192.30.252.40) left #scummvm.
[20:53] Nick change: Storm-AFK -> Stormkeeper
[20:53] <ScummBot> Port build status changed with 42b24e32: Failure: master-ios7, master-mingw-w32, master-openpandora, master-android_mips, master-android_x86, master-ios, master-wii, master-debian-x86, master-mingw-w64-cplusplus11, master-webos, master-osx_intel, master-psp2, master-gp2x, master-dingux, master-debian-x86-clang, master-gamecube, master-gp2xwiz, master-android_arm,
[20:53] <ScummBot> master-mingw-w64, master-deb
[20:53] <ScummBot> ian-x86_64
[20:54] <snover> well, you fixed it for one build ;)
[20:54] <dreammaster> Damn. What's wrong now!
[20:54] <snover> CaSe sENsiTiVity
[20:55] <dreammaster> Ah, yes. Hold on..
[20:55] <-- ny00123 left irc: Quit: Leaving
[20:56] <snover> i managed to screw that up once recently which makes think my mind is going
[20:56] --> GitHub168 joined #scummvm.
[20:56] <GitHub168> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5FBS
[20:56] <GitHub168> scummvm/master 0a8167b Paul Gilbert: TITANIC: Fix case sensitivity in header includes
[20:56] GitHub168 (GitHub168@192.30.252.42) left #scummvm.
[20:57] <snover> if it happens again ill be putting my scummvm checkout on a case-sensitive disk image :)
[20:58] <dreammaster> :)
[21:01] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[21:05] --> travis-ci joined #scummvm.
[21:05] <travis-ci> scummvm/scummvm#4437 (master - 42b24e3 : Paul Gilbert): The build was broken.
[21:05] <travis-ci> Change view : https://github.com/scummvm/scummvm/compare/0b258ae16f31...42b24e328afd
[21:05] <travis-ci> Build details : https://travis-ci.org/scummvm/scummvm/builds/276613158
[21:05] travis-ci (travis-ci@ec2-54-242-46-198.compute-1.amazonaws.com) left #scummvm.
[21:06] <ScummBot> Port build status changed with 0a8167be: Success: master-ios7, master-mingw-w32, master-openpandora, master-android_mips, master-android_x86, master-ios, master-wii, master-debian-x86, master-mingw-w64-cplusplus11, master-webos, master-osx_intel, master-psp2, master-gp2x, master-dingux, master-debian-x86-clang, master-gamecube, master-gp2xwiz, master-android_arm,
[21:06] <ScummBot> master-mingw-w64, master-deb
[21:06] <ScummBot> ian-x86_64
[21:07] <dreammaster> Good. Now that the skeleton classes are there, I can start fleshing them out with any methods that look different in the German version
[21:09] <-- ajax16384 left irc: Quit: Leaving
[21:18] <snover> i sure like Synalyze It!. i wish i had a suggestion for non-Mac users that need to reverse-engineer or reference data in binary files easily. its just great.
[21:21] <-- _sev_ left irc: Quit: This computer has gone to sleep
[21:23] --> _sev_ joined #scummvm.
[21:23] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[21:29] --> GitHub173 joined #scummvm.
[21:29] <GitHub173> [scummvm] dreammaster closed pull request #1021: Corrected spelling in xeen/map.cpp (master...patch-1) https://git.io/v5dD9
[21:29] GitHub173 (GitHub173@192.30.252.45) left #scummvm.
[21:29] --> GitHub190 joined #scummvm.
[21:29] <GitHub190> [scummvm] dreammaster pushed 2 new commits to master: https://git.io/v5F0l
[21:29] <GitHub190> scummvm/master a8215f8 Sven Kochmann: Corrected spelling...
[21:29] <GitHub190> scummvm/master 4bdaa8d Paul Gilbert: Merge pull request #1021 from Schallaven/patch-1...
[21:29] GitHub190 (GitHub190@192.30.252.41) left #scummvm.
[21:32] <-- Farmboy0 left irc: Remote host closed the connection
[21:44] <tsoliman> I just discovered that Japanese Police Quest 2 exists ... https://www.youtube.com/watch?v=w4RwlqaFq4M
[21:45] <tsoliman> there are no words to describe the weirdness - why??
[22:03] <m_kiewitz> I own 2 copies
[22:03] <m_kiewitz> I love it
[22:04] <m_kiewitz> (and ScummVM runs it just fine)
[22:13] <tsoliman> nice
[22:13] <tsoliman> are there more Japanese game versions that feature crazy hair?
[22:17] <tsoliman> https://kotaku.com/5812431/japan-what-the-hell-did-you-do-to-this-classic-pc-adventure-game
[22:17] <tsoliman> m_kiewitz: your video is in the article :)
[22:30] --> Strangerke_ joined #scummvm.
[22:32] <-- Strangerke left irc: Ping timeout: 240 seconds
[22:32] Nick change: Strangerke_ -> Strangerke
[22:40] <m_kiewitz> tsoliman: sadly there aren't any more crazy localizations
[22:40] <m_kiewitz> that's probably where plenty of hits came from
[22:40] <m_kiewitz> was already wondering why so many people watched that video
[22:41] <m_kiewitz> Sierra should have made more like those. They localized Quest for Glory too, sadly no crazy hair
[22:43] --> GitHub153 joined #scummvm.
[22:43] <GitHub153> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v5FgK
[22:43] <GitHub153> scummvm/master b033d1b Paul Gilbert: TITANIC: Added German specific TTnpcScript doSentenceEntry code...
[22:43] GitHub153 (GitHub153@192.30.252.45) left #scummvm.
[23:04] <-- m_kiewitz left irc: Quit: technology isn't intrinsically good or evil. It's how it's used. Like the Death Ray.
[23:08] --> Cheeseness joined #scummvm.
[23:09] <-- AndChat-594561 left irc: Quit: Bye
[23:31] <-- criezy left irc: Quit: criezy
[23:39] <-- girafe left irc: Quit: Leaving
[23:47] <-- stevenh left irc: Quit: Page closed
[00:00] --- Mon Sep 18 2017