ScummVM IRC Log - 15 Aug 2019

[Back to Index]

[00:09] <-- Polemic_ left irc: Read error: Connection reset by peer
[00:13] --> Polemic joined #scummvm.
[00:26] --> BeefEats joined #scummvm.
[00:42] <Grogbot> <rsn8887> @waltervn : Why are you saying the new stretch mode is the only one that starts up with black borders, when centered and integral also do?
[00:42] <Grogbot> <rsn8887> Also the argument how aspect correction is applied is not quite correct.
[00:42] <Grogbot> <rsn8887> Just disable the AR correction checkbox it is not needed if you use the new stretch mode.
[00:43] <Grogbot> <rsn8887> IN fact I consider the AR correction checkbox historical cruft that should ideally be eradicated.
[00:43] <Grogbot> <rsn8887> It used a slow and ugly software scaler that is not needed on most platforms.
[00:44] <Grogbot> <rsn8887> It increases the resolution of the drawing surface for no reason other than to support age-old SDL1 ports.
[00:45] <Grogbot> <rsn8887> Also a program like ScummVM should never change window size by itself. It should adapt to the window size the user sets IMO.
[00:46] <Grogbot> <rsn8887> Window size is not up to a program it is up to the user using the resize buttons.
[00:47] <Grogbot> <rsn8887> Aspect ratio correction tickbox is done via a ton of hacks all over the graphics managers using real2aspect and 200to240 scaler calls littered everywhere.
[00:48] <Grogbot> <rsn8887> That's my opinion.
[00:49] <Grogbot> <rsn8887> The rationale for the new stretch mode is simply to allow 4:3 aspect ratio to be enforced via hardware instead of software scaling, not using any scaler.
[00:51] <Grogbot> <rsn8887> This is faster, allows the application of scalers correctly, and works with all games.
[00:55] <Grogbot> <rsn8887> @criezy : If you change how the AR tickmark to do hardware AR correction in SDL2, using activeArea, then indeed the new stretch mode would probably become obsolete. Setting "Fit to window" and ticking the AR checkbox will then do the same. As "Fit to window (4:3)" now. However I fail to see the need for the tickbox, since we have the new stretch mode now?
[00:56] <Grogbot> <rsn8887> So basically, you might as well remove the tickbox for SDL2 based systems instead of changing it to hardware scaling, since the new stretch mode already does that.
[00:58] <Grogbot> <rsn8887> Using the tickbox with SDL2 currently, there are also a lot of hardcoded checks for game resolutions. For example AR correction is not applied with the tickbox when Dreamweb is loaded. In OpenGL, Dreamweb is AR adjusted with the tickbox. None of any of those checks are necessary IMO.
[01:00] <Grogbot> <rsn8887> In other words, "AR correction" was a confusing misnomer. It should be "Force 4:3 aspect".
[01:01] <Grogbot> <rsn8887> It doesn't make any sense to me to try to figure out the "correct" aspect for all the games and then apply a "correction". The old monitors could be adjusted to all kinds of aspect ratios. There's vast disagreement for each game what the correct AR should be. So we shouldn't pretend that we know.
[01:02] <Grogbot> <rsn8887> We should just offer users the choice of aspect ratio. E.g. 4:3, Square or original, maybe 16:9 and 16:10.
[01:03] <Grogbot> <rsn8887> And the idea of not allowing some games like Dreamweb to be "corrected" is also a frustrating experience. Maybe I want to play that game in 4:3 because that's how my Amiga monitor was setup back in the day. yes the circle in the first image looks stretched then, but so what?
[01:04] <Grogbot> <rsn8887> In some games like monkey island, the moon is stretched in one AR setting, and the windows in the church are stretched in another AR setting.
[01:06] <Grogbot> <rsn8887> Sorry but I just got back from work and saw the discussion I missed above.
[01:07] <Grogbot> <rsn8887> I hope you consider my viewpoint. I think we all agree that using hardware scaling to do the aspect stuff, when available, is the correct way to do it. The only question is how to present it in the interface. I don't really care how it can be enabled from a UI perspective, as long as it allows users to force 4:3 on any game they might want, and as long as hardware scaling is used for that.
[01:10] <Grogbot> <rsn8887> I added the "Fit to window (4:3)" stretch mode because it is completely optional (doesn't hurt anybody to have another choice), allows hardware scaling with minimal disruption of existing code, e.g. without touching all the AR correction code embedded in the graphics manager currently, fixes problems with the AR tickbox on Vita, and because honestly, no improvement on the AR correction (tickbox) code has been made in years.
[01:18] <Grogbot> <SupSuper> @rsn8887 i think they just wants the window to also resize to the ratio, instead of leaving black borders
[01:20] <Grogbot> <SupSuper> the only problem with something like fixed ratio in scummvm is there are a lot more ratios in old games than 4:3 and 16:9, so having an option for all of them is unfeasible
[01:20] <Grogbot> <rsn8887> That would be perfectly fine with me of course, although I personally don't like windows changing size on me like that, but that's fine
[01:20] <Grogbot> <SupSuper> but knowing what is the "correct" ratio for every game, and letting the user override all of them if preferrable, is also hard
[01:21] <Grogbot> <rsn8887> For the window size it might be enough to just trigger a window resize event of whatever using SDL2. But none of the other stretch modes do that either.
[01:22] <Grogbot> <rsn8887> In fact, the whole idea of stretch modes I thought was that you could just change the window size however you want with the bottom right corner button in windows, and then the ScummVM display automatically adapts.
[01:23] <Grogbot> <SupSuper> well it's hard to resize the window yourself to an exact size
[01:23] <Grogbot> <rsn8887> Ok fine, I don't see the problem of having black borders honestly?
[01:24] <Grogbot> <rsn8887> When the user first selects the mode, the window is changed to exactly fit the game.
[01:24] <Grogbot> <rsn8887> But then, when the user later decides to resize the window, it is allowed to change it to any size, and then black borders are displayed.
[01:25] <Grogbot> <rsn8887> Since it is a "Fit to window" mode, it must be allowed for the user to resize the window.
[01:25] <Grogbot> <rsn8887> It might also be possible with SDL2 to allow resizing of the window while the aspect ratio is maintained idk.
[01:26] <Grogbot> <rsn8887> I have never dealt with those things much, because PSP/Vita/Switch don't have that kind of functionality where the user can click the bottom right corner of a window to resize it.
[01:28] <Grogbot> <rsn8887> I guess it is time to delete a pointless rant.
[01:34] <Grogbot> <rsn8887> Ok I tried to defuse the rant, and delete some of my ramblings.
[01:35] <-- SylvainTV left irc: Read error: Connection reset by peer
[01:38] --> DominusExult joined #scummvm.
[01:38] <-- DominusExult left irc: Changing host
[01:38] --> DominusExult joined #scummvm.
[01:41] <-- Dominus left irc: Ping timeout: 245 seconds
[01:41] Nick change: DominusExult -> Dominus
[02:00] <-- Polemic left irc: Quit: -a- Connection Timed Out
[02:00] --> Polemic joined #scummvm.
[02:22] <-- timofonic left irc: Ping timeout: 272 seconds
[02:25] <-- nutron|w left irc: Ping timeout: 248 seconds
[02:33] <-- Polemic left irc: Read error: Connection reset by peer
[02:34] --> Polemic_ joined #scummvm.
[02:40] --> nutron|w joined #scummvm.
[03:03] <-- JohnnyonFlame left irc: Read error: Connection reset by peer
[04:16] <-- Drenn left irc: Ping timeout: 272 seconds
[04:47] <-- BeefEats left irc: Read error: Connection reset by peer
[04:48] --> Begasus joined #scummvm.
[04:52] --> Begas_VBox joined #scummvm.
[05:03] <-- SupSuper left irc: Quit: Rip
[05:46] <-- Begas_VBox left irc: Read error: No route to host
[05:47] --> Begasus_ joined #scummvm.
[05:48] <-- Begasus left irc: Ping timeout: 252 seconds
[05:51] --> Begas_VBox joined #scummvm.
[06:13] <-- Begas_VBox left irc: Ping timeout: 245 seconds
[06:16] <-- Begasus_ left irc: Ping timeout: 252 seconds
[06:16] <-- eriktorbjorn left irc: Quit: Leaving.
[06:19] --> Begasus joined #scummvm.
[06:22] --> Begas_VBox joined #scummvm.
[06:41] --> beresky joined #scummvm.
[06:47] <-- Begas_VBox left irc: Ping timeout: 245 seconds
[06:48] <-- Begasus left irc: Ping timeout: 252 seconds
[06:49] --> Begasus joined #scummvm.
[06:53] --> Begas_VBox joined #scummvm.
[07:20] <Grogbot> <waltervn> @rsn8887 my main "complaint" is with the borders. Let's say a game uses a 640x400 frame buffer and I use the 4:3 option. IMHO there's no valid reason to use a 640x400 window size and then draw a 4:3 thing inside of that with borders
[07:22] <Grogbot> <waltervn> going one step further, it would be nice if the window aspect ratio was maintained when resizing it. No idea how hard that would be to implement though. I'd be happy with no having the borders at launch in any case
[07:22] <Grogbot> <waltervn> I don't think I'll be using this mode in windowed mode, in its current form
[07:26] <-- cd left irc: Quit: cd
[07:32] <LePhilousophe> criezy: yep forum guy is right, copy paste error :/
[07:32] <LePhilousophe> I push
[07:32] <Scummette> [scummvm] lephilousophe pushed 1 new commits to master: https://git.io/fjdm1
[07:32] <Scummette> scummvm/master 6761ef6 lephilousophe: CRYOMNI3D: Fix Spanish compressed file size
[07:37] <LePhilousophe> criezy: should I add PROGRAM.Z to fallback entries? I am not sure it's a good idea because the name is really generic but without that when the file is unknown, the detection fallbacks to first entry with a PROGRAM.Z file which is versailles-win-fr
[07:43] <-- ink0gnito left irc: Ping timeout: 248 seconds
[08:04] --> jamm joined #scummvm.
[08:18] <-- jamm left irc: Read error: Connection reset by peer
[08:20] --> jamm joined #scummvm.
[08:56] <-- yuv422 left irc: Remote host closed the connection
[08:57] --> yuv422 joined #scummvm.
[08:58] <Scummette> [scummvm] bluegr pushed 2 new commits to master: https://git.io/fjdYj
[08:58] <Scummette> scummvm/master 384715d bluegr: STARTREK: Add a console
[08:58] <Scummette> scummvm/master 089c04f bluegr: STARTREK: Cleanup away mission chain loading code
[09:03] --> timofonic joined #scummvm.
[09:11] <Grogbot> <Henke37> and here i hoped that it'd use the chain loading system
[09:24] <Grogbot> <criezy> LePhilousophe: I am not sure. It seems to me that adding it as a fallback detection could lead to many incorrect detections, although I don't know if any of the other game we support does have such as file as well or not.
[09:25] <Grogbot> <criezy> Also it is my understanding that the fallback detected game might not work anyway depending on the way the font and sound files are named?
[09:33] --> vliaskov joined #scummvm.
[10:22] --> vyzigold joined #scummvm.
[10:25] --> SylvainTV joined #scummvm.
[10:25] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[10:39] <LePhilousophe> criezy: yep, you are right maybe we shouldn't have any fallback? or should we customize detection to determine the game flags and language?
[10:47] <LePhilousophe> hum I read again the fallback detection code, and I think I could do something
[10:55] <Grogbot> <criezy> I am guessing the engine could also found out at runtime when starting what files are present (e.g. check a couple of font and sound files)?
[10:55] <Grogbot> <criezy> And then we would not need to have that information in the detection table.
[11:44] --> ink0gnito joined #scummvm.
[11:47] <-- Begas_VBox left irc: Ping timeout: 248 seconds
[11:47] <-- Begasus left irc: Ping timeout: 252 seconds
[11:56] --> Begas_VBox joined #scummvm.
[12:13] <-- Begas_VBox left irc: Read error: No route to host
[12:17] --> Begas_VBox joined #scummvm.
[12:37] <-- jamm left irc: Ping timeout: 246 seconds
[12:40] <-- Begas_VBox left irc: Read error: Connection reset by peer
[12:44] --> Begas_VBox joined #scummvm.
[13:04] <-- Begas_VBox left irc: Ping timeout: 244 seconds
[13:37] <-- Polemic_ left irc: Quit: -a- Connection Timed Out
[13:43] --> Polemic joined #scummvm.
[13:52] <-- Polemic left irc: Read error: Connection timed out
[14:07] --> Polemic joined #scummvm.
[14:24] <-- beresky left irc: Read error: Connection reset by peer
[14:24] --> beresky joined #scummvm.
[14:42] --> Drenn joined #scummvm.
[14:47] <-- vyzigold left irc: Ping timeout: 245 seconds
[15:21] --> Begas_VBox joined #scummvm.
[15:29] <Scummette> [scummvm] digitall pushed 1 new commits to master: https://git.io/fjdWH
[15:29] <Scummette> scummvm/master 54c465e digitall: BACKENDS: Fix GCC Warnings in Surface SDL Graphics
[16:05] <-- ink0gnito left irc: K-Lined
[16:05] <Grogbot> <rsn8887> What about mouseCurState
[16:06] <Grogbot> <rsn8887> Shouldnt it be zeroed too?
[16:16] <Grogbot> <rsn8887> According to buildbot, this move lua... commit seems to have broken ds builds. https://github.com/scummvm/scummvm/commit/c1f029c6dcd606b3dbe8e3952ac55f89968ab6c4
[16:17] <Grogbot> <rsn8887> https://buildbot.scummvm.org/builders/master-ds/builds/1768
[16:17] <Grogbot> <rsn8887> See here ^
[16:20] --> ny00123 joined #scummvm.
[16:22] <Grogbot> <rsn8887> Is the buildbot not reporting into this channel anymore?
[16:23] <Grogbot> <rsn8887> I didnt see a build failure message from the buildbot here re ds, even though it hasnt not compiled correctly for a while
[16:23] <LePhilousophe> rsn8887: last ScummBot message is
[16:23] <LePhilousophe> 18:12:54 < ScummBot> Port build status changed with e609d024: Success: master-openpandora, master-psp, master-gamecube, master-ps2, master-pspfull, master-gcw0, master-gp2xwiz, master-dc-serial, master-dingux, master-wii, master-webos, master-dc, master-caanoo
[16:23] <LePhilousophe> IIRC it was after the Lua move
[16:24] <Grogbot> <rsn8887> DS build should have reported as failing
[16:24] <LePhilousophe> it has been reported failing before I suppose
[16:24] <LePhilousophe> but I don't have the backlog anymore
[16:25] <LePhilousophe> and noone thought that ScummBot didn't say all ports build fine now
[16:25] <Grogbot> <rsn8887> But the buildbot webpage shows a successful build just before the lua move commit
[16:25] <LePhilousophe> so master-dc was still failing
[16:25] <Grogbot> <rootfather> The failure was reported here last tuesday according to the discord logs
[16:25] <LePhilousophe> humpf
[16:25] <LePhilousophe> it says success in what I pasted
[16:25] <Grogbot> <rsn8887> ds
[16:26] <Grogbot> <rsn8887> Not dc 😉
[16:26] <LePhilousophe> oh great so it's not told
[16:26] <LePhilousophe> so ds has been announced as failed with all others
[16:26] <Grogbot> <rsn8887> Heres the history https://buildbot.scummvm.org/builders/master-ds
[16:26] <LePhilousophe> and never announced as success after that
[16:26] <LePhilousophe> and noone noticed that ScummBot didn't tell All ports build fine
[16:26] <Grogbot> <rsn8887> I think the clue is the missing all ports successfully good job type msg
[16:26] <Grogbot> <rsn8887> Yep my saying
[16:28] <Grogbot> <rsn8887> Also I dont understand the missing init of mouseCurState in @digitall Last commit but that might just be my failure to see something.
[16:29] <Grogbot> <rsn8887> The ds compile error looks complicated some kind of thread related stuff
[16:30] <Grogbot> <rsn8887> Maybe a tool chain issue no clue
[16:31] --> Polemic_ joined #scummvm.
[16:32] <-- Polemic left irc: Ping timeout: 272 seconds
[16:33] <Grogbot> <ccawley2011> I'm fairly sure that the issue is that the main executable is too large now that Lua has been moved to Common.
[16:33] --> ajax16384 joined #scummvm.
[16:33] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[16:34] <Grogbot> <ccawley2011> The __sync_synchronize warning occurs with successful builds like this one: https://buildbot.scummvm.org/builders/master-ds/builds/1757/steps/compile/logs/stdio
[16:38] <-- vliaskov left irc: Remote host closed the connection
[16:39] <-- Polemic_ left irc: Read error: Connection reset by peer
[16:41] <Scummette> [scummvm] ccawley2011 pushed 1 new commits to master: https://git.io/fjd8v
[16:42] <Scummette> scummvm/master 1e6e6c1 ccawley2011: DS: Disable Bink and Lua
[16:43] --> Polemic joined #scummvm.
[16:56] --> vyzigold joined #scummvm.
[17:36] --> eriktorbjorn joined #scummvm.
[17:36] #scummvm: mode change '+o eriktorbjorn' by ChanServ!ChanServ@services.
[18:11] --> cd joined #scummvm.
[18:12] --> SupSuper joined #scummvm.
[18:40] <-- Lightkey left irc: Ping timeout: 264 seconds
[18:52] --> Lightkey joined #scummvm.
[19:00] <-- SupSuper left irc: Quit: Rip
[19:04] <-- Polemic left irc: Ping timeout: 245 seconds
[19:07] --> Polemic_ joined #scummvm.
[19:11] <-- Polemic_ left irc: Read error: Connection reset by peer
[19:11] <-- timofonic left irc: Ping timeout: 246 seconds
[19:13] --> Polemic joined #scummvm.
[19:14] --> timofonic joined #scummvm.
[19:19] --> BeefEats joined #scummvm.
[19:25] --> Polemic_ joined #scummvm.
[19:26] <-- Polemic left irc: Ping timeout: 250 seconds
[19:47] <-- timofonic left irc: Ping timeout: 245 seconds
[19:47] --> timofonic joined #scummvm.
[19:58] --> SupSuper joined #scummvm.
[20:00] --> criezy joined #scummvm.
[20:00] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[20:24] <-- Polemic_ left irc: Ping timeout: 264 seconds
[20:27] <-- BeefEats left irc: Ping timeout: 264 seconds
[20:27] --> BeefEats joined #scummvm.
[21:03] <-- ajax16384 left irc: Quit: Leaving
[21:06] <-- Begas_VBox left irc: Quit: Vision[0.10.3]: i've been blurred!
[21:07] <-- vyzigold left irc: Ping timeout: 244 seconds
[21:21] --> Polemic joined #scummvm.
[21:23] <-- ny00123 left irc: Quit: Leaving
[21:27] <Scummette> [scummvm] criezy pushed 1 new commits to master: https://git.io/fjdRy
[21:27] <Scummette> scummvm/master bf65647 criezy: CREDITS: Add credits for Supernova and update iOS and macOS credits
[21:55] <Scummette> [scummvm] antoniou79 pushed 2 new commits to master: https://git.io/fjd08
[21:55] <Scummette> scummvm/master a2df43b antoniou79: BLADERUNNER: Cleaner format in constants and init_script(partial)
[21:55] <Scummette> scummvm/master 9a76434 antoniou79: BLADERUNNER: Early restoration for Dino Klein
[22:24] <ScummBot> Port build status changed with 1e6e6c1e: Success: master-ds. Nice work, all ports built fine now
[22:30] <-- Polemic left irc: Read error: Connection reset by peer
[22:33] --> Polemic joined #scummvm.
[22:42] <-- criezy left irc: Quit: criezy
[23:10] <Grogbot> <rsn8887> You fixed it! Awesome!
[23:45] <-- SylvainTV left irc: Read error: Connection reset by peer
[00:00] --- Fri Aug 16 2019