ScummVM IRC Log - 14 Sep 2017

[Back to Index]

[00:10] --> Strangerke_ joined #scummvm.
[00:12] <-- Strangerke left irc: Ping timeout: 240 seconds
[00:12] Nick change: Strangerke_ -> Strangerke
[00:33] <-- Mia left irc: Read error: Connection reset by peer
[01:02] --> Poly-C joined #scummvm.
[01:04] <snover> dreammaster: fwiw the cursor that bugs me the most is the click interaction icon
[01:04] <-- Polynomial-C left irc: Ping timeout: 240 seconds
[01:04] <snover> since the & light&beams& are blended
[01:05] <-- Littleboy left irc: Read error: Connection reset by peer
[01:05] --> Littleboy joined #scummvm.
[01:05] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[01:08] <dreammaster> The starburst one? Yeh, me too, of all of them
[01:09] <-- Dominus left irc: Ping timeout: 260 seconds
[01:10] --> Dominus joined #scummvm.
[01:10] <-- ccawley2011 left irc: Quit: Page closed
[01:17] <snover> im more familiar with cursorman now since i had to do work on it for https://github.com/scummvm/scummvm/pull/1009 so let me peek again to see what can be done
[01:17] <dreammaster> I just feel really uncomfortable with the idea of trying to hack in manually drawing cursors within just the Titanic engine. If it's to be done, it's better the time be spent properly adding support in the core, so other games in the future can benefit from it if needed
[01:18] <dreammaster> Like with my GUI double quoted parameters pull request. I should have just done that in the first place, rather than hacking in code to my debug commands to try and do search and replaces for spaces and # characters
[01:18] <dreammaster> Go for it :)
[01:23] <dafioram> snover: want to give me some hints on doing https://bugs.scummvm.org/ticket/9752 ? Like where I should start looking and such?
[01:23] <snover> dafioram: you bet!
[01:24] <dafioram> you mention two ways of doing it would one be better?

[01:25] <snover> when possible, my default preference is to patch into the game scripts since that reduces the amount of conditional branching and hacking of the engine, though in this case its hard for me to say which one of these options is overwhelmingly better
[01:26] <snover> and it may not be possible to find enough room in the game script to make the call
[01:27] <snover> on the script-side, as far as i know game saves always go through Game::save and Game::restore
[01:28] <dafioram> if this happens in multiple games will the script solution solve multiple games?
[01:28] <snover> though in the case of shivers 1, i suspect youd be needing to patch either the event handler that goes into the game menu, or if youre lucky, some other function that it calls to switch into the menu room
[01:29] <dafioram> so for lighthouse this will need to be done again, separately?
[01:29] <snover> if youre really lucky, the event handler calls to some method that is shared across games
[01:29] <dafioram> i c
[01:29] <snover> like some hypothetical Game::openMenu
[01:29] <dafioram> lets see what I can dig up
[01:29] <snover> cool. this research will really make the difference in terms of deciding which approach to take.
[01:30] <snover> for some of the games like QFG4, the fix may be as simple as reordering the save function so that it closes the control panel before calling the kernel to save the game instead of after
[01:31] <snover> since some games like SQ6 also show a control panel, but their screenshots turn out OK because the control panel gets closed first
[01:32] <dafioram> so could there be an issue of the game saving a picture twice if it already saves it and If I add another way for it to save?
[01:32] <snover> the original engine had no mechanism for saving thumbs with saves, thats entirely a ScummVM invention
[01:33] <dafioram> but its already saving an image somehow
[01:34] <snover> yes, the game script will callk to kSaveGame (kSaveGame32 in ScummVM) when it wants the kernel to save the game, and then that calls to gamestate_save, which calls to Graphics::saveThumbnail, which uses the OSystem frame buffer for the thumbnail
[01:37] <snover> which& i actually suppose means some of these cases might be solved if we were to call g_sci->_gfxFrameout->frameOut(true) before trying to grab the thumbnail to make sure that OSystems frame buffer has been updated
[01:39] <snover> though due to forced vsync, GfxFrameout wont allow more than one call to update the system buffer per frame, so that could require an additional step of actually using the engines own buffer instead
[01:40] <snover> everything is complicated, whee.
[01:43] <snover> dreammaster: is there something in the engine already that will give me an RGBA8888 surface?
[01:45] <snover> womp-womp. i forgot to update my titanic.dat.
[01:45] <snover> also, devtools/create_titanic/create_titanic_dat.cpp:944:54: warning: unknown escape sequence '\X' [-Wunknown-escape-sequence] i think you meant \x
[01:48] <-- Lightkey left irc: Ping timeout: 252 seconds
[01:50] <dreammaster> Ah. Thanks
[01:53] --> GitHub49 joined #scummvm.
[01:53] <GitHub49> [scummvm] dreammaster pushed 1 new commit to master: https://git.io/v59qb
[01:53] <GitHub49> scummvm/master 49d2dd5 Paul Gilbert: DEVTOOLS: Fix escape sequence in create_titanic
[01:53] GitHub49 (GitHub49@192.30.252.41) left #scummvm.
[01:53] Nick change: Stormkeeper -> Storm-AFK
[01:57] <dreammaster> Given it's just a single character, it's not worth regenerating the titanic.dat file. That can wait until such time that the German version is ever properly supported
[01:57] <snover> dreammaster: how about that question about the RGBA8888 surface, is there an existing utility function to do this or do i need to read the transparent surface and the pixel data surface and build the pixels for the RGBA8888 surface myself?
[02:01] <dreammaster> You mean for the cursors?
[02:01] <snover> yeah
[02:01] --> Lightkey joined #scummvm.
[02:02] <dreammaster> As per the bug, I have a branch ready where the cursors are built up and passed to CursorMan as rgba
[02:02] <snover> oh, great. for some reason, i didnt figure out that that is what was going on.
[02:02] <dreammaster> The "titanic_cursors" branch of my github.com/dreammaster/scummvm.git repository
[02:03] <snover> as per usual i just got confused and misunderstood what was written
[02:03] <dreammaster> Yeh, the previous code kept the pixels and transparency surface separate, and manually built up a cursor using the key color for any sufficiently transparent pixel. The branch simply rips that out, and forms an RGBA surface with both the pixels and alpha information combined.
[02:04] <dreammaster> At least, I think so. I didn't get around to doing dummy draws or anything to test the resultant cursors, but they should be fine
[02:11] <snover> well, uh. this code seems to be introducing a use-after-free :)
[02:12] <snover> upon `delete surface` in mouse_cursor.cpp:107, AVISurface dtor tries to use _videoSurface->_flipVertically but _videoSurface is already gone
[02:16] <dreammaster> Whoopsy :)
[02:20] <snover> youre downconverting the cursors to rgb565 in setCursor. remove that and send surf directly and the alpha cursors work fine :)
[02:22] <snover> oh, oops
[02:22] <snover> i need that ifdef.
[02:23] <dreammaster> Oh, I'd forgotten about that. I used that to verify at least that the cursors were correctly formed.
[02:23] <snover> hurrah, you already fixed it. :)
[02:23] <snover> just need to fix that use-after-free too and everything seems to be a-ok!
[02:24] <snover> https://zetafleet.com/i/59b9e871ee6f9.jpg
[02:24] <dreammaster> Just taking care of that now. I'm changing OSMovie to new/delete, that'll allow me to free it before I free the surface
[02:26] <dafioram> Do you have a reservation?
[02:26] <dreammaster> skip
[02:27] <snover> i feel like someone has stolen *my* brain tonight.
[02:29] <dreammaster> I feel likewise to a degree. Each of my first three attempts/experiments to simplify the audio buffer code kept resulting in no sound at all being played. It may be more effective to work on it on the weekend when I've got the day to properly fiddle around with it. And look into the actual music differences
[02:31] <dreammaster> Looks good. I've committed in the fix, as well as removing the IFDEF block
[02:32] <snover> im usually happy to take a second look at whatever, as long as you can deal with my brain dun broke sometimes :)
[02:33] <dreammaster> I think I can deal with that. :) Plus, unlike the whole Indeo transparencies, I still consider this a non-essential nice to have
[02:33] <snover> well it works!
[02:34] <dreammaster> That reminds me, I'll take you up on sending me a copy of your IDB for Indeo. Just for reference in case it's ever needed. Thanks :)
[02:34] <snover> no problem, one moment.
[02:38] <dreammaster> Who knows, the set of IDBs we have for for the various games may one day be a treasure trove for future generation researchers investigating the operation of early era computer games :)
[02:40] <snover> dreammaster: so are you waiting to land in the fixed cursor code in master for anything in particular?
[02:43] <dreammaster> Oh.. your question puzzled me for a moment, because it doesn't actually work yet. Then I realised I've only ever tested it in SDL, not OpenGL. Seems like it works fine in OpenGL. Yay for small blessings. But in SDL mode they're totally screwed up
[02:44] <dreammaster> Hence the need for core SDLGraphicsManager (I think it was) fixes
[02:44] <snover> are you using sdl1?
[02:44] <dreammaster> SDL2
[02:44] <snover> im rebuilding off my working branch to see if i inadvertently fixed whatever was broken
[02:44] <dreammaster> The backend code doesn't even allow for pixel formats greater than the bit depth of the system surface, from memory.
[02:45] <snover> that was a badly worded sentence.
[02:45] <snover> i merged your code onto my working branch and built from that
[02:45] <snover> now i am building your branch directly
[02:45] <dreammaster> There looks like there was some code started to support them, but it was unfinished
[02:45] <snover> that screenshot is using the sdlsurface graphics manager.
[02:45] <dreammaster> If your code does actually fix the problem accidentally, than double huzzah :)
[02:50] <snover> or im just really crazy and shouldnt be saying anything tonight. lets see where we end up.
[02:51] <snover> i switched from hq2x to opengl to check so i feel like im pretty sure it was already using the surface renderer
[02:52] <snover> but i didnt really do much to the surface cursor rendering. at one point i considered bypassing some software blit code but i dont think i actually did bypass it&
[02:52] <dreammaster> Well, if you still are on the OpenGL renderer, that would explain why it looked okay
[02:52] <dreammaster> I hadn't actually realised OpenGL worked okay, since I normally use the SDL renderer
[02:53] <snover> normally i use the surface renderer, i switched because it was working and i wanted to check that it worked in opengl too
[02:53] <snover> but i had been testing the cursor hiding in the other backends code i was working on earlier so i might have just been mistaken about what my default setting was at the time
[02:55] <dreammaster> Well, it's getting kind of late for me. We can pick this up again tomorrow night. You can let me know how you went
[02:55] <snover> ok. good night!
[02:55] <dreammaster> Goodnight
[02:55] <-- dreammaster left irc:
[02:56] <snover> turns out i was just using the wrong renderer before. whoops.
[02:56] <snover> thats fine. i can fix this.
[03:08] --> GitHub36 joined #scummvm.
[03:08] <GitHub36> [scummvm] stevenhoefel pushed 1 new commit to master: https://git.io/v593i
[03:08] <GitHub36> scummvm/master b674c57 stevenhoefel: DIRECTOR: Initial work for D5 RTE Cast Member loading.
[03:08] GitHub36 (GitHub36@192.30.252.35) left #scummvm.
[03:14] <-- _sev left irc: Read error: Connection reset by peer
[03:16] --> _sev joined #scummvm.
[03:16] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[03:23] <-- |Cable| left irc: Remote host closed the connection
[03:24] <-- dafioram left irc: Quit: Leaving
[03:31] <-- _sev left irc: Read error: Connection reset by peer
[03:32] --> _sev joined #scummvm.
[03:32] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[03:54] --> GitHub199 joined #scummvm.
[03:54] <GitHub199> [scummvm] stevenhoefel pushed 1 new commit to master: https://git.io/v59Gu
[03:54] <GitHub199> scummvm/master 3feaaff stevenhoefel: DIRECTOR: Fix Warning.
[03:54] GitHub199 (GitHub199@192.30.252.35) left #scummvm.
[04:06] <ScummBot> Port build status changed with b674c571: Failure: master-openpandora, master-amigaos4, master-osx_intel, master-caanoo, master-gp2x, master-gp2xwiz
[04:19] <ScummBot> Port build status changed with 3feaaffd: Success: master-openpandora, master-amigaos4, master-osx_intel, master-caanoo, master-gp2x, master-gp2xwiz
[04:20] <snover> wat.
[04:22] <snover> good news! sdl renderer will now work with 32bpp cursors.
[04:23] <snover> well, i didnt bother to go so far as to allow them to be scaled, since i dont think any of the scalers can deal with >16bpp data. but this should be fine.
[04:48] <-- bgK left irc: Ping timeout: 255 seconds
[04:49] --> bgK joined #scummvm.
[04:49] #scummvm: mode change '+o bgK' by ChanServ!ChanServ@services.
[05:30] --> GitHub18 joined #scummvm.
[05:30] <GitHub18> [scummvm] csnover pushed 3 new commits to master: https://git.io/v59co
[05:30] <GitHub18> scummvm/master 3eb12dc Colin Snover: SCI32: Remove ADGF_CD from games that have no non-CD version...
[05:30] <GitHub18> scummvm/master 421a2de Colin Snover: SCI32: Drop platform from game ID for games that are Windows-only
[05:30] <GitHub18> scummvm/master 8842097 Colin Snover: SCI32: Fix PQ:SWAT platform
[05:30] GitHub18 (GitHub18@192.30.252.42) left #scummvm.
[05:34] --> waltervn joined #scummvm.
[05:34] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[06:01] --> DJWillis joined #scummvm.
[06:01] #scummvm: mode change '+o DJWillis' by ChanServ!ChanServ@services.
[06:55] <-- mankeli left irc: Quit: Reconnecting
[06:55] --> mankeli joined #scummvm.
[07:07] <-- kurtwr left irc: Read error: Connection reset by peer
[07:16] --> Strangerke|work joined #scummvm.
[07:16] <Strangerke|work> hi guys
[07:26] <madmoose> Morning all
[07:31] <Strangerke|work> hi madmoose
[07:38] <-- cpasjuste left irc: Quit: ZNC 1.6.3+deb1 - http://znc.in
[07:53] --> Henke37 joined #scummvm.
[08:02] <-- TMM left irc: Quit: Ex-Chat
[08:16] --> jamm joined #scummvm.
[08:29] <-- LittleToonCat left irc: Remote host closed the connection
[08:33] --> _sev|work joined #scummvm.
[08:33] <-- _sev|work left irc: Changing host
[08:33] --> _sev|work joined #scummvm.
[08:33] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[08:49] --> TMM joined #scummvm.
[08:49] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[08:53] --> ajax16384 joined #scummvm.
[08:53] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[09:08] --> |Cable| joined #scummvm.
[09:13] <-- |Cable| left irc: Remote host closed the connection
[09:22] --> user9 joined #scummvm.
[09:27] --> |Cable| joined #scummvm.
[10:08] --> ignalina joined #scummvm.
[11:15] --> te_lanus joined #scummvm.
[11:22] <-- te_lanus left irc: Quit: Leaving
[11:31] <-- _sev|work left irc: Quit: This computer has gone to sleep
[12:07] --> _sev|work joined #scummvm.
[12:07] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[12:18] --> cpasjuste joined #scummvm.
[12:48] --> ccawley2011 joined #scummvm.
[13:03] <-- _sev|work left irc: Quit: This computer has gone to sleep
[13:05] --> _sev|work joined #scummvm.
[13:05] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:06] <-- _sev|work left irc: Client Quit
[13:11] <-- ignalina left irc: Quit: ignalina
[13:20] --> ignalina joined #scummvm.
[13:25] --> _sev|work joined #scummvm.
[13:25] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:27] <-- _sev|work left irc: Client Quit
[13:40] --> exmensa joined #scummvm.
[13:42] --> _sev|work joined #scummvm.
[13:42] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:49] <-- _sev|work left irc: Quit: This computer has gone to sleep
[13:50] --> _sev|work joined #scummvm.
[13:50] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:54] <-- _sev|work left irc: Client Quit
[14:11] <-- ignalina left irc: Quit: ignalina
[14:28] --> ignalina joined #scummvm.
[14:33] <-- antlarr left irc: Ping timeout: 264 seconds
[14:35] --> antlarr joined #scummvm.
[15:09] --> _sev|work joined #scummvm.
[15:09] <-- _sev|work left irc: Changing host
[15:09] --> _sev|work joined #scummvm.
[15:09] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[15:14] --> LittleToonCat joined #scummvm.
[15:20] --> Begasus joined #scummvm.
[15:21] <-- TMM left irc: Quit: Ex-Chat
[15:37] <-- user9 left irc: Remote host closed the connection
[15:42] <-- ignalina left irc: Quit: ignalina
[15:48] <-- Begasus left irc: Ping timeout: 255 seconds
[15:50] <-- noobineer left irc: Ping timeout: 248 seconds
[15:59] <-- _sev|work left irc: Quit: This computer has gone to sleep
[15:59] --> ignalina joined #scummvm.
[16:00] --> Begasus joined #scummvm.
[16:05] <-- ajax16384 left irc: Read error: Connection reset by peer
[16:12] <-- ignalina left irc: Quit: ignalina
[16:22] <-- jamm left irc: Ping timeout: 246 seconds
[16:25] <-- Begasus left irc: Ping timeout: 255 seconds
[16:37] --> Begasus joined #scummvm.
[16:42] --> ny00123 joined #scummvm.
[16:47] --> jamm joined #scummvm.
[17:02] <-- DJWillis left irc: Ping timeout: 252 seconds
[17:25] <-- waltervn left irc: Ping timeout: 240 seconds
[17:33] <-- macdude22 left irc: Read error: Connection reset by peer
[17:35] --> girafe joined #scummvm.
[17:47] --> m_kiewitz joined #scummvm.
[17:47] #scummvm: mode change '+o m_kiewitz' by ChanServ!ChanServ@services.
[17:49] <-- ny00123 left irc: Quit: Leaving
[17:49] <-- Begasus left irc: Quit: Ex-Chat
[17:51] --> ajax16384 joined #scummvm.
[17:51] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[17:52] --> ny00123 joined #scummvm.
[17:53] <-- ccawley2011 left irc: Quit: Page closed
[18:03] <-- |Cable| left irc: Ping timeout: 255 seconds
[18:15] --> waltervn joined #scummvm.
[18:15] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[18:16] --> |Cable| joined #scummvm.
[18:28] <-- |Cable| left irc: Ping timeout: 240 seconds
[18:41] --> |Cable| joined #scummvm.
[19:04] --> TMM joined #scummvm.
[19:04] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[19:59] --> criezy joined #scummvm.
[19:59] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[20:46] --> macdude22 joined #scummvm.
[20:52] <-- criezy left irc: Quit: criezy
[21:33] <-- waltervn left irc: Quit: Leaving
[21:35] <-- ajax16384 left irc: Read error: Connection reset by peer
[21:39] --> noobineer joined #scummvm.
[22:00] <m_kiewitz> PSA: PSYCHONAUTS *FOR* *FREE*
[22:00] <m_kiewitz> humble bundle
[22:02] <m_kiewitz> (sadly steam)
[22:02] <-- girafe left irc: Quit: Leaving
[22:11] <-- m_kiewitz left irc: Quit: technology isn't intrinsically good or evil. It's how it's used. Like the Death Ray.
[22:57] --> kurtwr joined #scummvm.
[23:09] <-- ny00123 left irc: Quit: Leaving
[23:14] <-- jamm left irc: Ping timeout: 264 seconds
[23:20] <exmensa> It looks like that includes a non-DRM download from HB directly :)
[23:27] <fydo> exmensa: oh nice
[23:27] <fydo> Linux too :D
[23:34] Action: Lightkey pities the fool who didn't get Humble Indie Bundle V
[23:34] <Lightkey> best bundle ever
[23:35] <snover> what was in that one?
[23:36] <Lightkey> https://en.wikipedia.org/wiki/List_of_Humble_Bundles#2012
[23:37] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[23:53] <snover> oh. turns out i have that one.
[23:53] <snover> hurrah!.
[23:53] <snover> but i also already had psychonauts before then. it is a great game.
[23:54] <snover> i hope the new one is going o-kay.
[00:00] --- Fri Sep 15 2017