ScummVM IRC Log - 08 Nov 2017

[Back to Index]

[00:04] <criezy> OK, that's enough supernova for today.
[00:04] <snover> are you sure??
[00:07] <criezy> Yes, it's getting late. And I have already watched it explode 5 times today.
[00:09] --> Lightkey joined #scummvm.
[00:11] <criezy> If you want to watch it as well there is a let's play on youtube: https://www.youtube.com/watch?v=47P6dQLm4Q0&t=326s#t=05m15s
[00:11] <criezy> The alternative is to build the engine and play half of the game to that point.
[00:12] <snover> the video also has sound effects
[00:12] <snover> i did not know that a supernova sounded like that
[00:13] <criezy> That's because that's a German one. American ones probably sound different.
[00:16] <snover> space colonialism is the worst.
[00:26] --> dreammaster joined #scummvm.
[00:26] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[00:56] Nick change: Stormkeeper -> Storm-AFK
[01:07] <criezy> It looks like the GitHub bot has fallen asleep.
[01:07] <criezy> Or it doesn't like Xeen.
[01:09] <Lightkey> nowhere to be xeen
[01:11] <dreammaster> It should be Xeen but not heard
[01:12] <Lightkey> Farbenspiel!
[01:12] <Lightkey> wooooo~

[01:35] <-- criezy left irc: Quit: criezy
[01:41] --> criezy joined #scummvm.
[01:41] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[01:41] <-- criezy left irc: Client Quit
[01:48] <-- SylvainTV left irc: Read error: Connection reset by peer
[02:01] <-- dreammaster left irc:
[02:05] <-- Dominus left irc: Ping timeout: 260 seconds
[02:06] --> Dominus joined #scummvm.
[02:07] <-- ccawley2011 left irc: Ping timeout: 260 seconds
[02:42] <snover> so are there platforms today that will actually fail on unaligned pointers? i ask because cryo engine, lure engine, and sky engine all do this, and clang reports warnings about it, which i have been ignoring but am now not ignoring.
[03:02] <snover> some nebulous words exist on the internet about old ARM doing bad things. and not always is the code using the safe macros, some places dereference these unaligned pointers. so its probably not fully portable and will break somewhere sometime.
[03:15] <-- P2E left irc: Ping timeout: 240 seconds
[03:22] --> P2E joined #scummvm.
[03:33] <snover> how about now, githubbot?!
[03:33] <-- _sev left irc: Quit: Leaving
[03:34] <snover> curious.
[03:38] <wjp> unaligned reads have definitely caused problems for us on some platforms in the past, but my brain is too dead to recall much atm
[03:42] <wjp> what kind of warnings does clang produce?
[03:44] <snover> for lure/sky, its scummvm/engines/lure/res.cpp:238:6: Taking address of packed member 'leftOffset' of class or structure 'Lure::HotspotAnimResource' may result in an unaligned pointer value
[03:44] <snover> for cryo, its Pointer not aligned at address 0x1027E9A9E (__ZZN4Cryo8EdenGame14specialObjectsEPNS_7perso_tEcE21kSpecialObjectActions + 398 from eden.o)
[03:47] <snover> for cryo i dont think there is any reason at all to be packing this struct
[03:49] <snover> the cryo thing was done in 7bf6e1e
[03:50] <wjp> odd
[03:50] <snover> im recompiling now to see if i get a warning or not so i can know what this mysterious silenced warning is
[03:50] <wjp> makes no sense
[03:51] <wjp> struct packing should have zero impact here
[03:52] <snover> incidentally having you around in the evening is blowing my mind right now :)
[03:53] <snover> its not used to this at all.
[03:53] <wjp> hey, I'm always around in the evening ;-)
[03:53] <wjp> but my mind is also not quite adjusted to this yet
[03:54] <-- P2E left irc: Read error: Connection reset by peer
[03:54] --> P2E joined #scummvm.
[03:55] <wjp> could it be a warning about suboptimal member ordering in the struct?
[03:56] <wjp> (not that that matters for arrays...)
[03:56] <snover> that seems plausible, though i would imagine that warning to be everywhere
[03:56] <wjp> and that
[03:56] <wjp> that array could maybe also use some consts
[03:57] <snover> i mean, how can i say no to a suggestion for more consts ;)
[03:58] <snover> this misalignment is also happening more insidiously in EdenGame::incPhase
[03:58] <snover> since phase_t is also being packed
[03:58] <snover> a ton of stuff in defs.h is packed
[03:59] --> _sev joined #scummvm.
[03:59] <-- _sev left irc: Changing host
[03:59] --> _sev joined #scummvm.
[03:59] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[04:02] <snover> several of these packed structs hold pointers so there is no way they could be safely loaded from data
[04:02] <snover> or safely used as an overlay onto data
[04:03] <snover> unless they overlay data only for the first fields prior to the pointer
[04:03] <wjp> could just be a leftover I suppose
[04:04] <wjp> but I haven't looked at these others
[04:04] <snover> im looking to see where the initial perso_t come from
[04:04] <snover> since there is some comment that they are loaded from cryo.dat
[04:05] <snover> i doubt phase_t needs packing, at least
[04:06] <snover> they look to be loaded by data stream so probably none of this needs packing
[04:08] <snover> the packing of defs.h comes from the initial commit for this engine, so i think you are probably right about it being a leftover
[04:09] <snover> phew. i am having horrible packet loss to vm2 right now.
[04:10] <snover> seems like a problem of the NOC, latency goes way up on the edge router
[04:11] <snover> guess ill look for lost eden bugs later!
[04:14] <wjp> bedtime for me; good night
[04:14] <snover> good night!
[04:18] <-- |Cable| left irc: Read error: Connection reset by peer
[04:18] --> |Cable| joined #scummvm.
[04:21] <snover> Strangerke: if you can say what warning you were trying to fix by packing in 7bf6e1e, please let me know
[04:26] <-- |Cable| left irc: Ping timeout: 248 seconds
[04:39] --> |Cable| joined #scummvm.
[04:52] <-- |Cable| left irc: Read error: Connection reset by peer
[04:52] --> |Cable| joined #scummvm.
[04:56] <-- |Cable| left irc: Remote host closed the connection
[04:58] --> |Cable| joined #scummvm.
[04:59] <snover> the missing bot is starting to bug me
[05:00] <-- |Cable| left irc: Remote host closed the connection
[05:03] --> |Cable| joined #scummvm.
[05:30] <snover> i guess taking the address of these packed members is going to turn out OK as long as the structs themselves are aligned since all of the members are 16-bit integers, so theres no chance of a misaligned read
[06:05] <-- user9 left irc: Read error: Connection reset by peer
[06:05] --> user9 joined #scummvm.
[06:25] <-- |Cable| left irc: Ping timeout: 248 seconds
[06:37] --> |Cable| joined #scummvm.
[07:33] <-- LittleToonCat left irc: Remote host closed the connection
[07:47] --> waltervn joined #scummvm.
[07:47] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[07:54] <-- _sev left irc: Quit: This computer has gone to sleep
[07:56] --> _sev joined #scummvm.
[07:56] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[08:00] <-- _sev left irc: Client Quit
[08:40] --> f2k joined #scummvm.
[09:17] <-- TMM left irc: Quit: Ex-Chat
[09:22] <-- rootfather left irc:
[09:42] gus (~quassel@2400:6180:0:d0::f76:7001) got netsplit.
[09:42] rsn8887 (Elite20518@gateway/shell/elitebnc/x-ubcaojefbpqjfgci) got netsplit.
[09:46] super_fox2[m] <-- (superfox2m@gateway/shell/matrix.org/x-ydtfktrykwtxfiyd) left irc: Ping timeout: 248 seconds
[09:46] tomazcuber[m] <-- (tomazcuber@gateway/shell/matrix.org/x-ixpnnhnstajnnrjj) left irc: Ping timeout: 248 seconds
[09:46] weezywinez[m] <-- (weezywinez@gateway/shell/matrix.org/x-wffabvdwvbuudumu) left irc: Ping timeout: 246 seconds
[09:46] Coldwine[m] <-- (coldwinema@gateway/shell/matrix.org/x-fupmhfnqqnvpqvqs) left irc: Ping timeout: 264 seconds
[09:46] trojan[m] <-- (trojanmatr@gateway/shell/matrix.org/x-izafmiujqgvvwkbq) left irc: Ping timeout: 255 seconds
[09:46] kusushi[m] <-- (kusushimat@gateway/shell/matrix.org/x-uridifxcfeptkuzp) left irc: Ping timeout: 255 seconds
[09:47] gus (~quassel@2400:6180:0:d0::f76:7001) returned to #scummvm.
[09:47] rsn8887 (Elite20518@gateway/shell/elitebnc/x-ubcaojefbpqjfgci) returned to #scummvm.
[09:51] --> _sev|work joined #scummvm.
[09:51] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[09:53] <-- _sev|work left irc: Client Quit
[09:54] --> kurtwr2 joined #scummvm.
[09:54] <-- kurtwr left irc: Ping timeout: 260 seconds
[09:59] tomazcuber[m] --> (tomazcuber@gateway/shell/matrix.org/x-vgkppemskgznthbo) joined #scummvm.
[10:30] trojan[m] --> (trojanmatr@gateway/shell/matrix.org/x-cxqbgdntrrkgtffg) joined #scummvm.
[10:30] weezywinez[m] --> (weezywinez@gateway/shell/matrix.org/x-atzltzsvspjilara) joined #scummvm.
[10:30] super_fox2[m] --> (superfox2m@gateway/shell/matrix.org/x-uyoyoqulossliwju) joined #scummvm.
[10:30] kusushi[m] --> (kusushimat@gateway/shell/matrix.org/x-yydnughruqkecqww) joined #scummvm.
[10:30] Coldwine[m] --> (coldwinema@gateway/shell/matrix.org/x-fiswdqauuwyhxoex) joined #scummvm.
[10:57] --> ajax16384 joined #scummvm.
[10:57] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[11:02] --> _sev|work joined #scummvm.
[11:02] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[11:23] --> Tea joined #scummvm.
[11:31] --> TMM joined #scummvm.
[11:31] <-- TMM left irc: Changing host
[11:31] --> TMM joined #scummvm.
[11:31] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[11:54] <-- TMM left irc: Quit: Ex-Chat
[11:57] <-- _sev|work left irc: Read error: Connection reset by peer
[11:57] --> _sev|work_ joined #scummvm.
[11:57] #scummvm: mode change '+o _sev|work_' by ChanServ!ChanServ@services.
[12:14] Action: waltervn wonders if he should try to find some last-minute ADL testers
[12:19] <-- user9 left irc: Ping timeout: 240 seconds
[12:20] <waltervn> I really should implement some kind of debug function that just runs through a list of commands to automate the testing
[12:24] <logix> ADL is text-only, so you mean something like a list of commands that solve the game and then you check at the end if the game ending screen is shown?
[12:24] <waltervn> yeah, or maybe grab screenshots after every command and compare that
[12:24] <logix> right, sure
[12:26] <logix> I have kind of a deja vu, just for graphical adventures - either something like this has come up a few times here in the channel or I just thought that it something like that might be a good idea
[12:26] <logix> s/ it / /
[12:26] <logix> I suppose the former, but I'm not entirely certain
[12:27] <logix> so basically a "move mouse to position x,y", "click" etc.
[12:27] <logix> a bot that plays through the game according to a fixed script if you will
[12:36] <-- ST left irc: Ping timeout: 240 seconds
[12:38] --> ST joined #scummvm.
[12:38] #scummvm: mode change '+o ST' by ChanServ!ChanServ@services.
[12:58] <-- _sev|work_ left irc: Quit: This computer has gone to sleep
[13:01] <waltervn> logix: we have that, it's called the "event recorder"
[13:01] <waltervn> unfortunately it needs some work
[13:02] <waltervn> http://wiki.scummvm.org/index.php/Event_Recorder
[13:02] --> _sev|work joined #scummvm.
[13:02] <-- _sev|work left irc: Changing host
[13:02] --> _sev|work joined #scummvm.
[13:02] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:03] <-- Mia left irc: Read error: Connection reset by peer
[13:05] --> Mia joined #scummvm.
[13:05] <-- Mia left irc: Changing host
[13:05] --> Mia joined #scummvm.
[13:15] <-- _sev|work left irc: Quit: This computer has gone to sleep
[13:34] --> _sev|work joined #scummvm.
[13:34] <-- _sev|work left irc: Changing host
[13:34] --> _sev|work joined #scummvm.
[13:34] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[13:36] --> jamm joined #scummvm.
[13:36] <-- jamm left irc: Changing host
[13:36] --> jamm joined #scummvm.
[13:37] --> TMM joined #scummvm.
[13:37] <-- TMM left irc: Changing host
[13:37] --> TMM joined #scummvm.
[13:37] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[13:38] --> user9 joined #scummvm.
[13:40] --> ccawley2011 joined #scummvm.
[13:45] <logix> waltervn: huh, cool, TIL
[13:46] <logix> I wonder if people are TASing games yet
[13:47] <-- _sev|work left irc: Quit: This computer has gone to sleep
[14:00] <-- Tomaz^ left irc:
[14:31] --> _sev|work joined #scummvm.
[14:31] <-- _sev|work left irc: Changing host
[14:31] --> _sev|work joined #scummvm.
[14:31] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[14:36] <-- Strangerke left irc: Ping timeout: 240 seconds
[14:43] <-- exmensa left irc: Ping timeout: 248 seconds
[15:42] <-- f2k left irc: Quit: Leaving
[15:46] <-- ajax16384 left irc: Quit: Leaving
[16:00] --> Henke37 joined #scummvm.
[16:05] <L0ngcat> WOW! Heart of China and Rise of the Dragon released on gog
[16:05] <L0ngcat> Dynamix ftw
[16:12] <waltervn> nice
[16:13] <-- TMM left irc: Quit: Ex-Chat
[16:15] --> vasc joined #scummvm.
[16:15] <vasc> hey, anyone knows how to blit a transparent 8bpp bitmap (with 0 as the transparent color) onto the screen?
[16:25] --> nast joined #scummvm.
[16:28] <-- jamm left irc: Ping timeout: 252 seconds
[16:33] <eriktorbjorn> L0ngcat: There's also this, but I don't know if it's available outside of Steam: http://store.steampowered.com/app/664850/8bit_Adventure_Anthology_Volume_I/
[16:44] <snover> vasc: theres no OSystem API for that, normally you would have a backbuffer and draw to that and then copy from the backbuffer to the screen with copyRectToScreen. the TransparentSurface class has binary blending code, so can help if you happen to have a game engine whose renderer is amenable to using such an API.
[16:54] --> LittleToonCat joined #scummvm.
[17:02] <-- _sev|work left irc: Quit: This computer has gone to sleep
[17:04] --> ajax16384 joined #scummvm.
[17:04] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[17:06] <vasc> yeah, i locked the screen surface and wrote to it direct. but should be slow.
[17:06] <vasc> the transparent surface seems interesting.
[17:06] <vasc> thx
[17:15] --> ny00123 joined #scummvm.
[17:18] --> Strangerke joined #scummvm.
[17:25] --> Tomaz^ joined #scummvm.
[17:28] --> girafe joined #scummvm.
[17:28] <snover> that EventRecorder conversation is topical, since I am going through tickets and someone says it just crashes. is there a chance that it is ever going to actually be done and working at this point?
[17:44] <waltervn> there's at least two new issues on top of the ones it had when it was shelved
[17:46] <waltervn> I don't remember the exact details, but there's an incompatibility with the SaveManager (is that the name?) refactoring done by LordHoto a few years ago
[17:46] <waltervn> I should probably just look in my logs instead of guessing, sorry, let me see...
[17:46] <snover> SaveFileManager, I guess
[17:47] <waltervn> there's also an issue with the cloud stuff, so if you record without cloud, then play back with cloud, you get sync errors. This one might be easy to fix, hopefully
[17:58] <waltervn> unfortunately I can't find anything in the logs... What I seem to recall is that there's multiple SaveFileManagers in the event recorder. At some point a file list cache of some sorts was added to the default implementation. As the event recorder saves a file with one instance and then tries to open it with the other, it fails because the loading instance is using an outdated cache
[17:59] <waltervn> I think the commit in question would be d6d63a1
[18:00] <snover> hmm.
[18:01] <snover> the caching behaviour is not great generally, when i am downloading save games from tickets to test i frequently end up needing to restart scummvm because it wont query the filesystem to see the newly placed saves. i wonder if the caching could be relaxed generally for systems that dont really need it, to at least have the event recorder work again on desktop systems.
[18:01] <snover> iirc the save caching was added for systems with very slow save memory like dreamcast
[18:02] <waltervn> I think for more complex engines the event recorder never really worked. I was able to use it to some extent in ADL
[18:03] <waltervn> still, for ADL I may prefer a simple textual input script rather than a complex (binary, I suppose) event recording
[18:15] <-- nast left irc: Quit: Leaving
[18:17] --> frankyboy_ joined #scummvm.
[18:18] <bgK> IMO the event recorder needs more than a few quick fixes. The approach used, without engine interaction/feedback, shows its limitations very quickly. And there's also the fact that it hooks a bit everywhere in the event system, making other changes more difficult.
[18:23] <snover> if its so, should it maybe just be carved out of master and put into a feature branch if someone wants to reuse its bones in the future?
[18:27] <bgK> you have my vote
[18:40] <snover> madmoose peterkohaut: do either of you want to do anything with this at the moment? https://bugs.scummvm.org/ticket/9722
[18:47] <-- eriktorbjorn left irc: Ping timeout: 240 seconds
[18:52] --> ccawley2011_ joined #scummvm.
[18:52] <-- ccawley2011 left irc: Ping timeout: 260 seconds
[18:53] --> eriktorbjorn joined #scummvm.
[18:54] <snover> I am doing ticket reviews for the next release. Should I just be closing really old tickets about dead ports like WinCE?
[18:54] --> SylvainTV joined #scummvm.
[18:54] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[18:54] --> Sylvain joined #scummvm.
[18:55] <snover> like this one, https://bugs.scummvm.org/ticket/5752
[18:57] #scummvm: mode change '+o eriktorbjorn' by ChanServ!ChanServ@services.
[19:15] <-- SylvainTV left irc: Quit: User pushed the X - because it's Xtra, baby
[19:16] <-- Sylvain left irc: Quit: User pushed the X - because it's Xtra, baby
[19:16] --> SylvainTV joined #scummvm.
[19:16] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[19:18] <-- ccawley2011_ left irc: Ping timeout: 260 seconds
[19:25] --> ccawley2011 joined #scummvm.
[19:29] <snover> im taking the plunge to upgrade to xcode 9.1. i hope i am not making a huge mistake.
[19:42] <logix> good plunge, we're all counting on you
[19:58] --> ccawley2011_ joined #scummvm.
[20:00] <-- ccawley2011 left irc: Ping timeout: 260 seconds
[20:28] <madmoose> snover: Not at the moment, no, for me.
[20:50] --> criezy joined #scummvm.
[20:50] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[20:54] <-- ccawley2011_ left irc: Ping timeout: 260 seconds
[20:54] --> ccawley2011 joined #scummvm.
[21:09] <vasc> https://www.youtube.com/watch?v=0WbZRFXHb6Y
[21:09] <vasc> got the title screen for Rise of the Dragon (DGDS) kinda working.
[21:13] <logix> sweet, great job!
[21:18] <vasc> prolly need to get the font loader working next. to be able to play the dialog screens.
[21:28] <madmoose> vasc: Looks good! Nice progress.
[21:32] <Harekiet> Cool :)
[21:32] <-- ajax16384 left irc: Read error: Connection reset by peer
[21:37] --> nast joined #scummvm.
[21:47] <-- waltervn left irc: Quit: Leaving
[21:52] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[21:54] --> JohnJohn joined #scummvm.
[22:12] <-- frankyboy_ left irc: Remote host closed the connection
[22:14] JohnJohn (d45c7a42@gateway/web/freenode/ip.212.92.122.66) left #scummvm.
[22:24] <L0ngcat> vasc: excellent :)
[22:24] <snover> vasc: are you pushing code updates to a fork somewhere?
[22:25] <vasc> https://github.com/vcosta/scummvm
[22:25] <vasc> also has a wiki with some info.
[22:37] --> _sev joined #scummvm.
[22:37] <-- _sev left irc: Changing host
[22:37] --> _sev joined #scummvm.
[22:37] #scummvm: mode change '+o _sev' by ChanServ!ChanServ@services.
[22:41] <snover> nice, thank you for the link!
[22:48] <-- ny00123 left irc: Quit: Leaving
[22:48] <-- nast left irc: Quit: Leaving
[23:15] <-- criezy left irc: Quit: criezy
[23:30] <-- girafe left irc: Quit: Leaving
[23:54] <-- Lightkey left irc: Ping timeout: 258 seconds
[23:55] --> dreammaster joined #scummvm.
[23:55] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[00:00] --- Thu Nov 9 2017