ScummVM IRC Log - 04 Apr 2020

[Back to Index]

[00:29] Action: chrisf just watches this project from the sidelines, but that revert... yup :P
[00:57] <-- JohnnyonFlame left irc: Read error: Connection reset by peer
[01:10] --> JohnnyonFlame joined #scummvm.
[01:14] <grogbot> <Thunderforge> @Henke37 I think the concern of adding games that wont run is that there are plenty of games that wont do that. There are currently over 270 games on the compatibility list (which is why I want to organize it). Adding even more that arent supported makes it even harder to navigate.
[01:15] <grogbot> <Thunderforge> @sev Maybe we should talking about this directly for now while we work out details so that we dont spam everyone. People can still comment via GitHub if they want to provide feedback to the data reorganization.
[01:16] <-- SylvainTV left irc: Read error: Connection reset by peer
[01:16] --> dreammaster joined #scummvm.
[01:16] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[04:06] <grogbot> <pixelmischief> I have finished "The Beast Within, A Gabriel Knight Adventure" using SCUMMVM. Does anyone want me to update some compatibility list somewhere, or provide some notes on caveats concerning setup?
[04:15] <-- dreammaster left irc:
[04:17] <grogbot> <pixelmischief> OK. I'll keep an eye out, in case.
[04:21] <grogbot> <Trembyle> It looks like you've raised bug reports before. Have you seen any additional issues?
[04:23] <grogbot> <pixelmischief> Nope. Problem was an unpatched set of assets. This results in crashes at Chapter 4. An official patch resolves the problem, but invalidates saved games from the unpatched set. Users should be warned to be sure they use patched assets from the beginning or they will have to restart after completing over 2/3 of the game.
[04:25] <grogbot> <Trembyle> I liked sev's suggestion in that ticket. Maybe it should be reopened to report on the issue. Automatic patching would be nice.
[04:28] <grogbot> <pixelmischief> I can imagine that solution going in a number of directions. Are you thinking about simply storing a URL to hosted files? Or were you thinking something more like the patches being included with the supported game and applied at the "Add Game" stage?
[04:29] <grogbot> <Trembyle> I'm just a user, not a developer, but yeah I could see both of those solutions working.
[04:30] <grogbot> <pixelmischief> Maybe a separate project, the SCUMMVM Game Patcher that keeps a database of CRC's for the "best" versions of each file for a game and a hosted store of free, supported versions of those files.
[04:31] <grogbot> <pixelmischief> Hmmmm....
[05:36] <-- Scummette left irc: Ping timeout: 265 seconds
[05:36] --> Scummette joined #scummvm.
[07:05] <-- cd left irc: Quit: cd
[07:14] --> ldevulder_ joined #scummvm.
[07:17] <-- ldevulder left irc: Ping timeout: 264 seconds
[07:22] --> ny00123 joined #scummvm.
[07:35] <Scummette> [scummvm] mduggan pushed 1 new commits to master: https://git.io/Jvbgd
[08:13] <-- Lightkey left irc: Ping timeout: 252 seconds
[08:22] <-- BeefEats left irc: Read error: Connection reset by peer
[08:27] --> Lightkey joined #scummvm.
[08:29] <ScummBot> Port build status changed with 426af9fe: Failure: master-amigaos4, master-psp2full, master-ios, master-debian-x86, master-mingw-w32, master-ps3, master-mingw-w64-cplusplus11, master-osx_intel, master-debian-x86-clang, master-mingw-w64, master-debian-x86_64, master-ios7
[08:32] <grogbot> <Henke37> i'm not so worried about "won't run at all". it's the gray area between that i care about
[08:39] <grogbot> <mduggan> wtf why does xcode build that fine on my local
[08:41] <grogbot> <mduggan> oh, I forgot to update my own unit tests.. what a newbie error 😦
[08:44] <Scummette> scummvm/master 426af9f mduggan: ULTIMA8: Rename DataSource methods to match Common Streams
[08:44] <Scummette> [scummvm] mduggan pushed 1 new commits to master: https://git.io/JvbaU
[08:44] <Scummette> scummvm/master 4161cab mduggan: ULTIMA8: Update unit tests for function renames
[08:47] --> ajax16384 joined #scummvm.
[08:47] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[08:52] <ScummBot> Port build status changed with 4161cab9: Success: master-amigaos4, master-psp2full, master-ios, master-debian-x86, master-mingw-w32, master-ps3, master-mingw-w64-cplusplus11, master-osx_intel, master-debian-x86-clang, master-mingw-w64, master-debian-x86_64, master-ios7
[09:54] <-- girafe left irc: Read error: No route to host
[10:00] <-- _sev left irc: Quit: This computer has gone to sleep
[10:05] --> girafe joined #scummvm.
[10:08] --> waltervn joined #scummvm.
[10:08] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[10:11] --> _sev joined #scummvm.
[10:11] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[10:25] <-- Scummette left irc: Ping timeout: 272 seconds
[10:25] --> Scummette joined #scummvm.
[11:01] --> criezy joined #scummvm.
[11:01] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[11:17] <Scummette> [scummvm] lephilousophe opened pull request #2178: ANDROID: Multiple fixes for the new Gradle based build (master...fix-android) https://git.io/Jvbwu
[11:18] <LePhilousophe> @antoniou79: no, but you now get a PR to review which fixes it :)
[11:18] <LePhilousophe> @peterkohaut: if you can look at it too, as you did the Gradle thing, that would be great
[11:19] <grogbot> <antoniou79> nice. That's even better.
[11:21] <LePhilousophe> the fix is not really one though
[11:21] <LePhilousophe> the real fix would be to implement SAF in backends/fs
[11:22] <LePhilousophe> I didn't see any other way to access external data properly
[11:23] --> DominusExult joined #scummvm.
[11:23] <-- DominusExult left irc: Changing host
[11:23] --> DominusExult joined #scummvm.
[11:24] <LePhilousophe> in Android 11 it seems there will be a way to access data using paths but I wonder how it will work for Android 10 if we target an API level greater than 29
[11:26] <-- Dominus left irc: Ping timeout: 260 seconds
[11:26] Nick change: DominusExult -> Dominus
[11:28] <grogbot> <xesf> Is there a wiki guide on how to deal with game metadata? (which is not part of the original game assets)
[11:31] <grogbot> <antoniou79> LePhilousophe: Yes, we are aware of the Android 10 issue and have discussed a few solutions. I believe Peter is working on it.
[11:32] <grogbot> <antoniou79> We'll see about the Android 11 stuff when it comes to that. (I hate all these restrictions by the way)
[11:33] <grogbot> <Bigmanjapan> Hello. In ScummVM's Blade Runner build a player has to press Enter button for each text line to be skipped. Is this an intentional change to the original build of the game? The original have the Enter button skip all the lines if you simply hold it down.
[11:36] <LePhilousophe> @antoniou79: oh ok, if Peter is working on it I won't try to code something
[11:36] <LePhilousophe> @xesf: what do you mean by metadata?
[11:37] <Scummette> [scummvm] stevenhoefel opened pull request #2179: DEVTOOLS: Add ignore flag for byte packing (master...msvc-2019-fix) https://git.io/Jvbwj
[11:39] <grogbot> <xesf> I am testing few things for The X-Files Game which uses a embedded database file that we cannot extract the data without the original source code (the database is serialised without schema definitions). So I need to create the data from the database into a separate file/files to be included as part of an engine
[11:39] <grogbot> <antoniou79> @Bigmanjapan that doesn't sound right. Did it skip all dialogue lines by holding Enter down?
[11:39] <grogbot> <antoniou79> I don't remember that, and I don't think it's in the manual either
[11:41] <grogbot> <Bigmanjapan> It would skip all the lines one by one in really fast manner.
[11:41] <grogbot> <Bigmanjapan> I speedrun the game, therefore I quickly noticed the change.
[11:42] <grogbot> <Bigmanjapan> I've tried running the game off PCem and via the compatability patch available on replaying.de they both have the fast skipping feature present.
[11:43] <LePhilousophe> @xesf: for binary data bundled in executable, you have to create a script/binary in devtools which will generate a data file placed in dists/engine-data
[11:44] <grogbot> <sev> @xesf why not implementing the reading from the database?
[11:45] <grogbot> <antoniou79> @Bigmanjapan I don't think I've noticed anything about this behavior in the code, maybe some of the other developers did, @peterkohaut , @madmoose ?
[11:45] <grogbot> <sev> @xesf you're parsing it by some script, right? so why not incorporate that script? We put data in a separate file only if it is embedded in the .exe, and even in those cases we still try to read from the .exe unless it is not practical, e.g. would require hundreds of hard-coded offsets
[11:45] <grogbot> <antoniou79> Still the in-game Help menu, does not mention it
[11:45] <grogbot> <antoniou79> It does say that Enter skips current line of dialogue. Nothing about holding it down
[11:45] <grogbot> <Bigmanjapan> Not sure if it's OK to post it here, but here you can see how fast the text is skipped:
[11:45] <grogbot> <Bigmanjapan> https://youtu.be/1nt15f2p8q8?t=17
[11:45] <grogbot> <Bigmanjapan> I simply hold down the Enter button.
[11:46] <grogbot> <antoniou79> What OS and set up is this version of Blade Runner played at?
[11:47] <grogbot> <Bigmanjapan> Windows 7 64, original Blade Runner 4CD build installed via the compatability patch from replaying.de.
[11:48] <grogbot> <xesf> @sev No, the data is not parsed by a script. Is is manually generated. Reversing the whole database deserialisation will be a though one, probably harder than recreate the data inside it.
[11:49] <grogbot> <sev> what do you mean "manually generated"?
[11:49] <grogbot> <sev> and how big is the data?
[11:51] <Scummette> [scummvm] sev- pushed 2 new commits to master: https://git.io/Jvbrs
[11:51] <Scummette> scummvm/master 4829527 sev-: DIRECTOR: More robust preprocessing for `sound` command
[11:51] <Scummette> scummvm/master 5752cb1 sev-: CRYO: Use ScummVM packing headers instead of direct #pragmas
[11:51] <grogbot> <xesf> the binary it self is not that big 6MB, but there is quite a few data there
[11:52] <grogbot> <xesf> manually generate, I am creating a separate metadata file where I have a new format with the links between the video nodes
[11:52] <grogbot> <xesf> and their definitions
[11:55] <grogbot> <sev> you're making me confused
[11:55] <grogbot> <sev> where that DB data sits? in an .exe? in a separate file?
[11:57] <grogbot> <sev> @lephilousophe ping
[11:57] <_sev> LePhilousophe: ping
[11:58] -ChanServ:#scummvm- _sev set flags +AOViortv on lePhilousophe
[11:58] #scummvm: mode change '+o LePhilousophe' by ChanServ!ChanServ@services.
[11:58] <grogbot> <xesf> @sev its a separate file.
[11:59] <grogbot> <sev> 6M?
[11:59] <grogbot> <xesf> yes
[11:59] <grogbot> <sev> yikes
[11:59] --> ccawley2011 joined #scummvm.
[11:59] #scummvm: mode change '+o ccawley2011' by ChanServ!ChanServ@services.
[11:59] <grogbot> <xesf> that is included with the original game, but I am avoiding to trying to parse that file
[12:00] <grogbot> <sev> I am not sure that putting that big chunk of data under GPL is a great idea
[12:00] <grogbot> <xesf> The file is an embedded NeoAccess Database from NeoLogic Technology company
[12:00] <grogbot> <sev> again, what is the game and file name? perhaps I can take a look if I have a copy
[12:01] <grogbot> <xesf> its The X-Files Game, files XFILES.HDB and XFILES.GAM
[12:01] <grogbot> <sev> because you could also treat it as a usual binary blob, and use our ReadStream logic over it
[12:02] <grogbot> <sev> perhaps it contains indexes and btrees, and those could be silently ignored during the parsing
[12:02] <grogbot> <sev> @xesf this? https://www.mobygames.com/game/x-files-game
[12:02] <grogbot> <xesf> yes that one
[12:03] <grogbot> <xesf> I've tried in the past Hex Editing the file to see if I could understand the format, but was pretty hard
[12:04] <grogbot> <antoniou79> @Bigmanjapan you are right. I see it there. It could be a side effect of the original game polling the enter key too fast or intended. I will look into it. Could you create a bug report ticket on the scummvm bug tracker about it?
[12:06] <grogbot> <sev> @xesf let me see
[12:06] <grogbot> <xesf> Cheers
[12:07] <grogbot> <Bigmanjapan> @antoniou79 Sure.
[12:07] <grogbot> <xesf> @sev the good news is that using movie player done as part of the Mohawk engine, I can play the Cinepak video entries, as well as the static mjpeg entries
[12:29] <-- yuv422 left irc: Remote host closed the connection
[12:29] --> yuv422 joined #scummvm.
[12:29] #scummvm: mode change '+o yuv422' by ChanServ!ChanServ@services.
[12:54] <LePhilousophe> _sev: pong
[12:54] <LePhilousophe> @sev: pong
[12:54] <_sev> LePhilousophe: thanks
[12:54] <_sev> it would be nice if we drop program.z as the detection target file
[12:55] <LePhilousophe> _sev: it is already kind of dropped
[12:55] <_sev> or at least add to it another one, so make that entry consisting of 2 files
[12:55] <_sev> the file name is too generic
[12:55] <LePhilousophe> yes it's because I fixed this in master and didn't want to backport new detection code in stable
[12:56] <LePhilousophe> it depends on a patch in the generic detector so I wouldn't like to backport it even today
[12:56] <_sev> I see in master 12 detection entries are still using it
[12:57] <LePhilousophe> there is a macro which associate it with some data files
[12:57] <_sev> ah, you're right. I missed that
[12:58] <_sev> cool! that's all I wanted to say
[12:58] <LePhilousophe> good so you confirm that detection shouldn't trigger on program.z only anymore
[12:59] <LePhilousophe> that's what I intended to do
[12:59] <_sev> right
[14:02] <grogbot> <peterkohaut> @Bigmanjapan, @antoniou79 i dont remember this behavior but i can check disassembly again,, but yes, opening a bug report would be best
[14:09] <-- Scummette left irc: Ping timeout: 265 seconds
[14:10] --> Scummette joined #scummvm.
[14:18] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://git.io/Jvbil
[14:20] --> SylvainTV joined #scummvm.
[14:20] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[14:23] <Scummette> scummvm/master d47b2fb sev-: DIRECTOR: Do not double delete MacArchive
[14:23] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://git.io/Jvbio
[14:23] <Scummette> scummvm/master 2cbb611 sev-: GRAPHICS: MACGUI: Fix double freeing of menus
[15:44] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://git.io/Jvb1w
[15:44] <Scummette> scummvm/master 4fd5e2e sev-: DIRECTOR: Persist screen between movies
[16:25] --> cd joined #scummvm.
[17:10] <grogbot> <Thunderforge> https://bugs.scummvm.org/ticket/2477
[17:10] <grogbot> <Thunderforge> > Setting as pending with resolution outdated. Should be closed in a month if no one has responded with a replication case by then.
[17:10] <grogbot> <Thunderforge> 17 months ago... 😄
[17:10] <grogbot> <Thunderforge> I think it's time to close it!
[17:19] <-- ccawley2011 left irc: Ping timeout: 272 seconds
[18:25] <-- Scummette left irc: Ping timeout: 272 seconds
[18:25] --> Scummette joined #scummvm.
[18:26] --> BeefEats joined #scummvm.
[19:06] <grogbot> <SupSuper> any msvc users here?
[19:10] <grogbot> <Henke37> 🙋
[19:10] <grogbot> <SupSuper> how can i put out a msvc update and tell every msvc user to actually get it 😛
[19:28] --> ccawley2011 joined #scummvm.
[19:28] #scummvm: mode change '+o ccawley2011' by ChanServ!ChanServ@services.
[20:24] <grogbot> <Strangerke> change something in the code that breaks if the new create_project isn't used
[20:24] <grogbot> <Strangerke> that's typically the only thing that made me change to a new create_project since it exists
[21:30] <grogbot> <NeoDement> haha
[21:30] <grogbot> <NeoDement> why does ctrl z close scummvm instantly
[21:30] <grogbot> <NeoDement> accidentally just found that
[21:33] <-- ajax16384 left irc: Quit: Leaving
[21:40] <Strangerke> ctrl-Z?
[21:42] <grogbot> <NeoDement> yeah, i was trying to undo but i was tabbed into the wrong window
[21:43] <Scummette> [scummvm] rvanlaar opened pull request #2180: DIRECTOR: Adds debug option to automate tests. (master...rvanlaar/debug) https://git.io/JvbFP
[21:57] <grogbot> <antoniou79> is that on Linux?
[21:58] <grogbot> <SupSuper> can confirm it works on windows
[21:58] <grogbot> <SupSuper> the ultimate undo 😄
[22:07] <grogbot> <antoniou79> This does not work on Linux. Ctrl+Q quits, but that's expected
[22:09] <grogbot> <ccawley2011> The default key combination for exiting ScummVM is different on different platforms: https://github.com/scummvm/scummvm/blob/master/backends/events/default/default-events.cpp#L362
[22:20] <Scummette> [scummvm] sev- pushed 3 new commits to master: https://git.io/JvbbS
[22:31] <Scummette> scummvm/master 54fc3b5 sev-: DIRECTOR: LINGO: Remove _twoWordBuiltins as an obsolete concept
[22:31] <Scummette> scummvm/master e819543 sev-: DIRECTOR: LINGO: Remove special treatment of 'sound' in lexer. Now we do it with preprocessor
[22:31] <Scummette> scummvm/master 825b714 sev-: DIRECTOR: LINGO: Merged sound-* with b_sound()
[22:31] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://git.io/JvbNT
[22:31] <Scummette> scummvm/master f2a0032 sev-: DIRECTOR: Do not automatically delete archive resources, as we do it manually
[22:34] <grogbot> <sev> in fact @ccawley2011 that makes not a good impression
[22:34] <grogbot> <sev> @NeoDement it is covered in the readme: https://github.com/scummvm/scummvm#54-hotkeys
[22:35] <grogbot> <NeoDement> it is!
[22:35] <grogbot> Command sent by sev
[22:35] <grogbot> ...but who is reading those...
[22:35] <grogbot> <NeoDement> just surprised me when i found it by mistake lol
[22:37] <grogbot> <NeoDement> I've found a minor obscure bug that I'll mention here in case anyone finds it interesting
[22:38] <grogbot> <NeoDement> If you Look At an item in Day of the Tentacle with a description text that goes on for at least 2 sentences (such as the textbook), then rapidly right click an item without a description set for that character (such as the bucket of soapy water), it prints parts of the "bucket of soapy water" string to the text above the character
[22:38] <grogbot> <NeoDement> it does NOT do that in dosbox
[22:40] <grogbot> <NeoDement> in fact, it doesn't even do it in scummvm today, it did yesterday lol
[22:40] <grogbot> <NeoDement> maybe it has to do with how many strings are loaded in or something
[22:41] <grogbot> <SupSuper> @sev the readme makes it look like Windows was added later
[22:41] <grogbot> <NeoDement> https://cdn.discordapp.com/attachments/581224061091446795/696127865833390181/unknown.png
[22:42] <grogbot> <NeoDement> nah wait there it goes
[22:46] <grogbot> <NeoDement> in fact, it does do it in dosbox with a bit more testing
[22:47] <grogbot> <NeoDement> so i guess it's a game bug, not a vm bug at all
[23:07] <-- ny00123 left irc: Quit: Leaving
[23:17] <-- ccawley2011 left irc: Quit: Leaving
[23:24] <-- criezy left irc: Quit: criezy
[23:43] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://git.io/JvbxW
[23:43] <Scummette> scummvm/master 51fbba2 sev-: DIRECTOR: Fix warning
[23:49] <-- waltervn left irc: Quit: Leaving
[23:54] <Scummette> [scummvm] sev- pushed 1 new commits to master: https://git.io/Jvbx2
[23:54] <Scummette> scummvm/master b217fff sev-: DIRECTOR: Plug memory leak in MacArchive
[00:00] --- Sun Apr 5 2020