[Back to Index]

[01:05] <-- Joefish left irc: Ping timeout: 240 seconds
[01:09] --> James|GlideM joined #scummvm.
[01:09] #scummvm: mode change '+v James|GlideM' by ChanServ!ChanServ@services.
[01:09] <-- Mia left irc: Read error: Connection reset by peer
[01:14] --> DominusExult joined #scummvm.
[01:15] <-- Dominus left irc: Ping timeout: 260 seconds
[01:15] Nick change: DominusExult -> Dominus
[01:30] <-- t0by left irc: Quit: t0by
[04:02] --> GitHub90 joined #scummvm.
[04:02] <GitHub90> [scummvm] csnover pushed 1 new commit to master: https://git.io/v92Ol
[04:02] <GitHub90> scummvm/master 8d94a04 Colin Snover: SCI32: Disable game script video benchmarking...
[04:02] GitHub90 (GitHub90@192.30.252.42) left #scummvm.
[05:06] <-- LittleToonCat left irc: Remote host closed the connection
[05:13] <-- Uhfgood left irc: Quit: Leaving
[05:45] --> James|GlideM1 joined #scummvm.
[05:47] <-- James|GlideM left irc: Ping timeout: 268 seconds
[06:08] <-- Lightkey left irc: Ping timeout: 260 seconds
[06:21] --> Lightkey joined #scummvm.
[06:21] --> uruk-hai joined #scummvm.
[06:21] #scummvm: mode change '+o uruk-hai' by ChanServ!ChanServ@services.
[06:26] --> m_kiewitz joined #scummvm.
[06:26] #scummvm: mode change '+o m_kiewitz' by ChanServ!ChanServ@services.
[06:56] --> Simei joined #scummvm.
[07:04] <Simei> Hello, everyone. My proposal for projet SLUDGE engine was accepted and I'm quite thrilled. :D Thank you for your trust!
[07:05] --> James|GlideM joined #scummvm.
[07:05] #scummvm: mode change '+v James|GlideM' by ChanServ!ChanServ@services.
[07:05] <Simei> It will be a great experience to work with you on this project. :D
[07:09] <-- James|GlideM1 left irc: Ping timeout: 240 seconds
[07:14] --> Joefish joined #scummvm.
[07:20] <Strangerke|work> hi Simei, and welcome aboard =)
[07:20] <Strangerke|work> (and hi guys)
[07:22] <Simei> :)
[07:24] <Strangerke|work> Simei: Pure curiosity: do you speak French?
[07:27] <Harekiet> ip says yes
[07:27] <Simei> Yes, I speak French. :)
[07:29] <Strangerke|work> oki doki :)
[07:30] <Strangerke|work> Harekiet: you could live and study in France and speak another language... I guess
[07:30] <Harekiet> maybe maybe
[07:30] <Strangerke|work> with Erasmus exchange program, typically
[07:30] <Harekiet> you'd want to know some french at least :)
[07:31] <Strangerke|work> I could happily live in Netherlands without speaking a single word of dutch :P
[07:32] <Harekiet> yeh we have enough of those already!
[07:33] <Simei> Yes, that's true. Students from exchange programs don't speak much french at first, but it will be quite difficult to live in France.
[07:33] <Simei> A lot of french people speak English with heavy accent, haha
[07:34] <Harekiet> at least more and more are willing to speak english :)
[07:35] <Simei> Yes, that's also true. :)
[07:37] Action: Endy waves
[07:37] <Strangerke|work> What's the problem with French people speaking Engrish with a strong accent? :P
[07:37] <Strangerke|work> heya Endy :)
[07:37] <criezy> Don't trust IPs. Mine probably claim that I speak English... :P
[07:37] <Endy> Ho Strangerke|work :)
[07:38] <criezy> Congratulation Simei and welcome onboard.
[07:39] <Harekiet> criezy it's claiming you should speak french belgian....
[07:40] <Simei> criezy Thank you. :)
[07:40] <Simei> A lot of French people try to pronounce English words with French pronunciation rules. That causes problems, haha.
[07:41] <Harekiet> I like the way it sounds :)
[07:41] <Harekiet> better than the dutch speaking english
[07:42] <criezy> Joefish, Simei: we will send you an email later today or tomorrow with some information on how to get started.
[07:44] <Simei> Ok, I'll keep an eye on emails :)
[07:45] <Strangerke|work> I didn't manage to use a computer at home yesterday, I'll fix the studnt channel rights this evening hopefully
[07:46] --> GitHub116 joined #scummvm.
[07:46] <GitHub116> [scummvm-web] criezy pushed 1 new commit to master: https://github.com/scummvm/scummvm-web/commit/f3b9fb7d6030b73f323f5a1c263c89035d643833
[07:46] <GitHub116> scummvm-web/master f3b9fb7 Thierry Crozat: WEB: Announce GSoC projects
[07:46] GitHub116 (GitHub116@192.30.252.45) left #scummvm.
[07:48] <-- TMM left irc: Ping timeout: 268 seconds
[07:51] Action: Endy toasts to an awesome GSoC team - both student participents and project mentors :)
[07:52] <Strangerke|work> :)
[07:53] <Joefish> morning everyone \o/
[07:54] <Strangerke|work> hey Joefish
[07:55] <Joefish> Strangerke|work: hey. so you will be available only at office hours for now?
[07:55] <criezy> Good morning
[07:55] <Joefish> hi criezy :)
[07:56] <Strangerke|work> no, I shouldn't have issues now... hopefully
[08:25] --> Mia joined #scummvm.
[08:44] <_sev> hi Joefish
[08:44] <_sev> hi Simei
[08:50] <Joefish> hey sev
[08:50] --> criezy|Work joined #scummvm.
[08:51] #scummvm: mode change '+o criezy|Work' by ChanServ!ChanServ@services.
[09:01] --> TMM joined #scummvm.
[09:01] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[09:04] --> t0by joined #scummvm.
[09:04] #scummvm: mode change '+v t0by' by ChanServ!ChanServ@services.
[09:11] <uruk-hai> hey all :)
[09:13] <t0by> Heyyyyy, uruk-hai:
[09:13] <t0by> (sans : )
[09:13] <t0by> What's cooking?
[09:14] <t0by> Simei: hi and welcome!
[09:14] <uruk-hai> trying to wrap my head around MS's bot framework and trying to understand why our customers think that chatbots are such a great thing
[09:14] <uruk-hai> but it seems i am too oldschool for this kind of stuff :)
[09:14] <Strangerke|work> uruk-hai: yes, you're old. tsk.
[09:15] <uruk-hai> :( :D
[09:15] <t0by> To be fair the customers also probably dim as soup and chasing some fad.
[09:15] <t0by> But I can confirm, you're decrepit.
[09:15] <t0by> That's the most important bit.
[09:16] <uruk-hai> i know i could rely on you guys :D
[09:16] <uruk-hai> knew*
[09:16] <t0by> You're welcome.
[09:17] <uruk-hai> and what's up on that side of the internet? :)
[09:17] <t0by> Strangerke|work: fwiw I found out that my ancestors on my maternal grandfather's side are in all likelihood from the Calais area.
[09:17] <t0by> (Since the surname is ubiquitous in there and virtually non-existent elsewhere)
[09:21] <Strangerke|work> i have ancestors from a bit everywhere, so it doesn't surprise me that much. i'm pretty sure I have italian blood in my veins too, because of the roman presence in France 2 thousands of years ago
[09:36] --> omer_mor_ joined #scummvm.
[09:38] --> ST joined #scummvm.
[09:38] #scummvm: mode change '+o ST' by ChanServ!ChanServ@services.
[09:38] --> ny00123 joined #scummvm.
[09:39] <-- omer_mor left irc: Ping timeout: 255 seconds
[09:52] <-- Simei left irc: Ping timeout: 260 seconds
[10:21] --> omer_mor joined #scummvm.
[10:23] <-- omer_mor_ left irc: Ping timeout: 255 seconds
[10:40] <-- user9 left irc: Ping timeout: 260 seconds
[10:40] <-- uruk-hai left irc: Ping timeout: 255 seconds
[10:48] --> Axy joined #scummvm.
[10:48] <-- Axy left irc: Changing host
[10:48] --> Axy joined #scummvm.
[10:51] <-- Mia left irc: Ping timeout: 240 seconds
[10:59] --> uruk-hai joined #scummvm.
[10:59] #scummvm: mode change '+o uruk-hai' by ChanServ!ChanServ@services.
[11:14] <-- James|GlideM left irc: Ping timeout: 240 seconds
[11:14] --> Tomaz^W joined #scummvm.
[11:29] --> James|GlideM joined #scummvm.
[11:29] #scummvm: mode change '+v James|GlideM' by ChanServ!ChanServ@services.
[12:04] --> James|GlideM1 joined #scummvm.
[12:07] <-- James|GlideM left irc: Ping timeout: 255 seconds
[12:24] --> user9 joined #scummvm.
[12:42] --> Simei joined #scummvm.
[12:52] --> Tomaz^W2 joined #scummvm.
[12:53] --> Mia joined #scummvm.
[12:53] <-- Mia left irc: Changing host
[12:53] --> Mia joined #scummvm.
[13:02] Tomaz^W (~tompsson@84.216.7.40) got netsplit.
[13:02] Axy (~Mia@unaffiliated/mia) got netsplit.
[13:02] bgK (~bgk@2001:41d0:2:599c::2a60:8434) got netsplit.
[13:02] |Cable| (~cable@33.138.117.91.dynamic.reverse-mundo-r.com) got netsplit.
[13:08] |Cable| (~cable@33.138.117.91.dynamic.reverse-mundo-r.com) returned to #scummvm.
[13:13] bgK (~bgk@2001:41d0:2:599c::2a60:8434) got lost in the net-split.
[13:13] Axy (~Mia@unaffiliated/mia) got lost in the net-split.
[13:13] Tomaz^W (~tompsson@84.216.7.40) got lost in the net-split.
[13:13] --> bgK joined #scummvm.
[13:13] #scummvm: mode change '+o bgK ' by orwell.freenode.net
[13:25] bgK (~bgk@2001:41d0:2:599c::2a60:8434) got netsplit.
[13:26] bgK (~bgk@2001:41d0:2:599c::2a60:8434) returned to #scummvm.
[13:26] #scummvm: mode change '+o bgK' by ChanServ!ChanServ@services.
[13:31] --> jamm joined #scummvm.
[13:37] <-- ny00123 left irc: Quit: Leaving
[13:56] --> waltervn joined #scummvm.
[13:56] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[14:37] --> James|GlideM joined #scummvm.
[14:37] #scummvm: mode change '+v James|GlideM' by ChanServ!ChanServ@services.
[14:41] <-- James|GlideM1 left irc: Ping timeout: 240 seconds
[14:45] --> James|GlideM1 joined #scummvm.
[14:49] <-- James|GlideM left irc: Ping timeout: 240 seconds
[14:54] --> Henke37 joined #scummvm.
[15:01] --> rootfather joined #scummvm.
[15:01] #scummvm: mode change '+o rootfather' by ChanServ!ChanServ@services.
[15:03] carloratm (carloratm@137.ip-92-222-83.eu) left #scummvm.
[15:15] --> Begasus joined #scummvm.
[15:21] --> Begas_VBox joined #scummvm.
[15:23] <-- TMM left irc: Ping timeout: 240 seconds
[15:26] <-- vpelletier left irc: Quit: leaving
[15:30] --> Uhfgood joined #scummvm.
[15:34] <-- Simei left irc: Quit: Page closed
[15:50] <m_kiewitz> :(
[15:50] <m_kiewitz> he wanted to be back in 2 weeks
[16:01] --> James|GlideM joined #scummvm.
[16:01] #scummvm: mode change '+v James|GlideM' by ChanServ!ChanServ@services.
[16:02] <-- uruk-hai left irc: Ping timeout: 240 seconds
[16:04] <-- James|GlideM1 left irc: Ping timeout: 268 seconds
[16:11] <t0by> it's two weeks and *12 hours already*!!!!111
[16:20] <m_kiewitz> 13 hours! :P
[16:21] --> WooShell joined #scummvm.
[16:26] --> LittleToonCat joined #scummvm.
[16:37] <WooShell> meow =^.^=
[16:39] <-- Begasus left irc: Ping timeout: 240 seconds
[16:40] <-- Begas_VBox left irc: Ping timeout: 260 seconds
[16:43] <Strangerke|work> m_kiewitz: He'll be back soon. At worst monday.
[16:44] --> Begasus joined #scummvm.
[16:45] --> emilknievel joined #scummvm.
[16:45] <-- Strangerke|work left irc: Quit: Page closed
[16:46] <m_kiewitz> Strangerke|work: horray
[16:46] <m_kiewitz> damn
[16:47] --> Begas_VBox joined #scummvm.
[16:49] --> frankyboy_ joined #scummvm.
[16:56] --> girafe joined #scummvm.
[17:00] <snover> does anyone know if there is away to annotate translations in source code instead of in the weblate interface?
[17:09] <snover> (im considering committing some code that uses a translatable string that is copied into a fixed-size buffer, so would like to specify the maximum string length)
[17:10] <-- James|GlideM left irc: Read error: Connection reset by peer
[17:15] <bgK> snover: that seems to work: https://github.com/scummvm/scummvm/blob/master/engines/mohawk/dialogs.cpp#L180
[17:15] <bgK> not sure it's the best way
[17:25] --> vinayakvivek joined #scummvm.
[17:25] <tsoliman> snover: do you think the SCI32 savegame change is ready to land? https://github.com/tsoliman/scummvm/commit/7adc368fd52447f67d15bfd80c48e6a45c3b0e0b
[17:31] --> Farmboy0 joined #scummvm.
[17:31] <-- Farmboy0 left irc: Changing host
[17:31] --> Farmboy0 joined #scummvm.
[17:33] <snover> tsoliman: probably, but i have been searching to try to figure out why loading untitled saves isnt allowed, so that if someone has made an untitled game, it can still be loaded
[17:34] <snover> unless there is some good reason for this to not be allowed, which i dont know of yet (for sci32, at least)
[17:36] <snover> im being irrationally annoyed by this :)
[17:36] <tsoliman> I understand - I will push it to master and we can always revisit that point
[17:37] --> GitHub175 joined #scummvm.
[17:37] <GitHub175> [scummvm] tsoliman closed pull request #945: SCI32: Set a savegame description if none was entered (master...sci32-savegame-timestamps) https://git.io/v94TB
[17:37] GitHub175 (GitHub175@192.30.252.42) left #scummvm.
[17:38] --> GitHub58 joined #scummvm.
[17:38] <GitHub58> [scummvm] tsoliman pushed 1 new commit to master: https://git.io/v9aaf
[17:38] <GitHub58> scummvm/master f30f34c Tarek Soliman: SCI32: Set a savegame description if none was entered...
[17:38] GitHub58 (GitHub58@192.30.252.34) left #scummvm.
[17:39] <tsoliman> the thing that bothers me ever so slightly is that when saving a game with the timestamps (and I love the timestamp feature, don't get me wrong) and overwriting it, the timestamp doesn't get updated
[17:39] <tsoliman> I realize this isn't SCI-specific
[17:40] <tsoliman> but like sometimes I need a "quicksave" slot and I would like to leave it untitled and have it default to the timestamp the savegame was created
[17:41] <tsoliman> if the "default to timestamp" code is taken out of everything .. you end up with the GUI's default "(untitled savegame)"
[17:41] <tsoliman> I would like _that_ text to be like "(untitled savegame @ YYYYMMDD-HHMMSS)"
[17:41] <tsoliman> or something
[17:42] <snover> well, the timestamp *is* in the save game
[17:42] <tsoliman> right exactly so that's very doable
[17:42] <tsoliman> and it does show up in the "preview" area once you select a slot
[17:43] <tsoliman> what I would prefer is if it showed up in the list of saves
[17:43] <tsoliman> so that way you can have 3 different "quicksave" slots
[17:44] <tsoliman> I guess this is more relevant to SCI than say SCUMM just because of the nature of the games and the need for saving often .. but yeah this is a system-wide thing not SCI-specific
[18:04] <criezy|Work> snover: as bgK said, use a I18N comment just before the string you want to annotate
[18:04] <criezy|Work> This is documented here: http://wiki.scummvm.org/index.php/Supporting_GUI_Translation
[18:05] <-- emilknievel left irc: Quit: My MacBook has gone to sleep. ZZZzzz&
[18:08] <snover> criezy|Work: thanks. is this some custom functionality? i searched and searched online for weblate "source string comment" and couldnt find anything about it
[18:12] <criezy|Work> Weblate is only a frontend to help us work on the translation. The source for weblates are po files generated by the GNU gettext tools.
[18:12] <criezy|Work> So those comments are parsed by gettext.
[18:13] <criezy|Work> And we tell gettext to look for the I18N comments:
[18:13] <criezy|Work> https://github.com/scummvm/scummvm/blob/master/po/module.mk#L8
[18:15] --> Strangerke joined #scummvm.
[18:15] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[18:20] <snover> aha! i understand and will hopefully remember this from now on. thanks :)
[18:21] <-- jamm left irc: Quit: Leaving
[18:23] <snover> tsoliman: so after describing what you actually want (and what sounds like a better general GUI change), it seems like that patch might not be the right solution after all :( :)
[18:33] <tsoliman> snover: no .. this patch just makes SCI32 consistent with SCI and most other engines
[18:33] <Strangerke> /query chanserv
[18:33] <Strangerke> access #scummvm list
[18:33] <Strangerke> yeah, fail.
[18:35] --> digitall joined #scummvm.
[18:35] <-- digitall left irc: Changing host
[18:35] --> digitall joined #scummvm.
[18:35] #scummvm: mode change '+o digitall' by ChanServ!ChanServ@services.
[18:35] <Strangerke> hi digitall :)
[18:35] <digitall> Evening.
[18:40] <Lightkey> interesting.. the programmer of Mission Supernova otherwise only worked on MAG!!!, which I added a review for as my first contribution on MobyGames
[18:49] <logix> Lightkey: also "andromeda's erbe"
[18:50] <logix> (without the apostrophe, sorry...)
[18:52] <Lightkey> the credits for that is a little bit lacking, only mentioning his brother
[19:21] --> ajax16384 joined #scummvm.
[19:21] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[19:22] --> emilknievel joined #scummvm.
[19:23] <-- Strangerke left irc: Ping timeout: 268 seconds
[19:26] <-- emilknievel left irc: Ping timeout: 240 seconds
[19:41] --> Strangerke joined #scummvm.
[19:41] #scummvm: mode change '+o Strangerke' by ChanServ!ChanServ@services.
[19:42] <-- criezy|Work left irc: Quit: Page closed
[20:10] -ChanServ:#scummvm- _sev set flags +AOort on T0by
[20:10] #scummvm: mode change '+o t0by' by ChanServ!ChanServ@services.
[20:12] -ChanServ:#scummvm- _sev set flags +Viv on Joefish
[20:12] #scummvm: mode change '+v Joefish' by ChanServ!ChanServ@services.
[20:13] --> omer_mor_ joined #scummvm.
[20:15] <-- omer_mor left irc: Ping timeout: 255 seconds
[20:17] --> TMM joined #scummvm.
[20:17] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[20:18] <tsoliman> someone is having frame skipping in GK2 in ScummVM but not dosbox - https://www.reddit.com/r/adventuregames/comments/695qtp/scummvm_blinding_you_with_science/dh6c4x6/
[20:19] <tsoliman> I am not experiencing the same - I get some tearing
[20:19] <tsoliman> bug report or forum post?
[20:23] <m_kiewitz> tsoliman: you use OpenGL?
[20:23] <tsoliman> no - I am trying now with OpenGL
[20:24] <tsoliman> to see if I don't get the tearing
[20:24] <m_kiewitz> OpenGL depends on driver
[20:24] <m_kiewitz> some drivers do v'sync, which would then cause frame skipping
[20:24] <m_kiewitz> others don't and thus would then cause tearing
[20:25] <m_kiewitz> ah it's for videos
[20:25] <snover> theres not enough information in that thread for me to do anything
[20:25] <m_kiewitz> do you get tearing for videos only or also for gameplay?
[20:26] <tsoliman> openGL doesn't seem to have tearing
[20:26] <tsoliman> just videos - gameplay doesn't seem to have rapid horizontal motion
[20:26] <tsoliman> so if I am getting tearing there, I am not seeing it
[20:27] <snover> marienplatz scrolls the entire game window
[20:28] <m_kiewitz> Do we default to OpenGL?
[20:28] <tsoliman> good call - I haven't started my scummvm playthrough of it yet - just intro videos and such
[20:28] <m_kiewitz> we don't right?
[20:28] <tsoliman> we don't
[20:28] <tsoliman> we default (I think) to 2x SDL
[20:29] <m_kiewitz> my Japanese Phantasmagoria 2 has even worse issues, the intro video freezes (video only)
[20:29] <m_kiewitz> tsoliman: do you own kq7?
[20:29] <m_kiewitz> if so, can you please try that one too and check if the videos have the same issue?
[20:29] <m_kiewitz> and do you use SDL2 or 1?
[20:30] <tsoliman> yes I do .. SDL2 .. HQ3x
[20:30] <tsoliman> macOS .. I have a discrete card but I force the intel card
[20:32] <m_kiewitz> i don't remember getting any tearing ever on my notebook and it's really slow. I'm using SDL1 though.
[20:32] <m_kiewitz> can you try what happens when using SDL1 please?
[20:32] <tsoliman> the new macOS hates SDL1
[20:32] <tsoliman> we're talking chug-chug city
[20:32] <m_kiewitz> urgh
[20:33] <m_kiewitz> you have another computer?
[20:33] <tsoliman> yes a windows 7 one
[20:33] digitall (digitall@unaffiliated/digitall) left #scummvm.
[20:33] <tsoliman> my linux laptop is broken :(
[20:33] <tsoliman> no dev environment on windows though
[20:33] <m_kiewitz> can you try SDL2 on that one and if you get tearing, try SDL1 too?
[20:33] <m_kiewitz> just grab a daily build
[20:33] <m_kiewitz> i can send you a SDL1 win32 build
[20:34] <m_kiewitz> i wonder abou the video, it shouldn't take up much of the CPU
[20:34] <m_kiewitz> *about
[20:35] <snover> on marienplatz i see a small amount of tearing.
[20:36] --> vpelletier joined #scummvm.
[20:37] <tsoliman> the intro video in KQ7 is a tiny box
[20:37] <tsoliman> it also seems to be like 5 fps or so
[20:38] <m_kiewitz> but no tearing in that case? or at least no visible tearing?
[20:38] <m_kiewitz> it's AVI and I think frame rate is quite low
[20:38] <tsoliman> not that I can see
[20:38] <criezy> Speaking of KQ7 videos, there is a question on the forum: http://forums.scummvm.org/viewtopic.php?t=14322&sid=a2b02ea6cee084fd5602879c412e0d99
[20:38] <criezy> "any way to use the avi videos in KQVII instead of rbt? The framerate is better in the avi videos."
[20:38] <snover> the dos-version uses robot, not avi
[20:38] <m_kiewitz> ah yes
[20:39] <snover> the way to use the higher framerate video is to run the windows version& theyre exactly the same except for those videos
[20:39] <m_kiewitz> my localized kq7 has .avis only
[20:39] <snover> there was no dos interpreter until 2.00
[20:39] <m_kiewitz> (and is windows 3.11 only :P)
[20:41] <m_kiewitz> snover: you get no tearing and also no frame skip?
[20:41] <m_kiewitz> you are using mac, right?
[20:41] <snover> yes
[20:41] <snover> i dont perceive any frameskip
[20:41] <snover> doesnt mean there isnt any& im going to instrument this to be sure
[20:41] <tsoliman> yeah I don't get frameskip either
[20:42] <m_kiewitz> you get tearing though. when you have v'sync enabled it will frame skip
[20:43] <m_kiewitz> on consoles companies often disable v'sync to get a "better" frame rate, but then you get tearing
[20:43] <m_kiewitz> i'm really sensitive to it, can't stand it
[20:43] <tsoliman> I can't stand tearing normally
[20:43] <tsoliman> I will force vsync on in windows usually
[20:44] <tsoliman> so with HQ3x on a mac, the CPU usage is like < 18% on the most-used core .. so it isn't pegging
[20:45] <snover> hqx should not be relevant for hires sci32 games, they are initialised to normal1x
[20:45] <tsoliman> this is on windowed mode anyway .. so I can see htop's output
[20:45] <tsoliman> I can check with fullscreen .. but I'll have to ssh in or something
[20:46] <m_kiewitz> maybe it's even a SDL2 issue
[20:46] <m_kiewitz> i don't remember getting any tearing ever. And I also can't remember getting frame skip either unless something was completely broken (and there is nothing completely broken lol)
[20:48] <m_kiewitz> tsoliman: you could maybe also try king's quest 6?!
[20:48] <m_kiewitz> no idea how bad the frame rate for the kq6 videos are
[20:48] <m_kiewitz> or maybe you could try gabriel knight 1
[20:48] <m_kiewitz> that one also has videos and i think frame rate shouldn't be that bad
[20:49] <tsoliman> yeah CPU is a total red herring here .. it uses less CPU playing videos than it does while idling on the main menu in GK2
[20:49] <tsoliman> (we're talking 35% vs 10%)
[20:49] <m_kiewitz> yes, that's what i expected
[20:50] <m_kiewitz> i can try it on my Wii too
[20:51] <tsoliman> oh wow .. KQ6 is tearing on the intro
[20:51] <tsoliman> I didn't expect that
[20:52] <tsoliman> there's a camera pan from above the chandelier that makes it clear
[20:52] <m_kiewitz> o_O
[20:53] <tsoliman> (this is on mac as well still)
[20:53] <m_kiewitz> win32 sdl1 build just went out
[20:53] <tsoliman> got it
[20:53] <tsoliman> I am gonna switch to the windows box now
[20:54] <m_kiewitz> i never had tearing / frame skip in the kq6 intro ever
[20:56] <m_kiewitz> tsoliman: please also download the last official scummvm release for mac and try that one too
[20:56] <tsoliman> does it matter if I pick "dos" vs "windows"? I've been picking "dos"
[20:57] <m_kiewitz> it shouldn't
[20:58] <snover> i have been wondering lately about how frequently sci32 graphics code has to call to updateScreen and whether that is not working appropriately
[20:59] <tsoliman> sdl.dll is missing - I think I better grab a daily build a replace the exe with the one you sent
[20:59] <m_kiewitz> snover: that could be a problem, yes
[21:00] <m_kiewitz> it shouldn't be called too often
[21:00] <m_kiewitz> tsoliman: will send one right now
[21:01] <tsoliman> m_kiewitz: don't worry it works
[21:01] <m_kiewitz> kk
[21:01] <tsoliman> I put the exe in the same dir and called it scummvm-sdl1.exe
[21:01] <tsoliman> so now scummvm.exe is the normal (presumably sdl2) daily build
[21:02] <tsoliman> and scummvm-sdl1.exe is the one you sent
[21:02] <tsoliman> and they both start
[21:03] <m_kiewitz> i'm not absolutely sure that the daily build is sdl2
[21:03] <m_kiewitz> taking a look
[21:04] <tsoliman> something is wrong
[21:04] <tsoliman> I keep installing the daily build
[21:04] <tsoliman> and when I run it .. it says oct 2016
[21:04] <m_kiewitz> daily build still seems to be SDL1?!
[21:05] <snover> which daily build? there are 3 of them
[21:05] <snover> for windows
[21:05] <tsoliman> the big button one
[21:06] <tsoliman> windows installer
[21:06] <snover> what is the url?
[21:06] <m_kiewitz> daily build is dated 5.5.2017
[21:06] <m_kiewitz> tsoliman: go here http://buildbot.scummvm.org/snapshots/master/mingw-w32-master-latest.zip
[21:07] <tsoliman> http://scummvm.org/frs/scummvm/1.9.0/scummvm-1.9.0.2-win32.exe
[21:07] Action: tsoliman is an idiot
[21:07] <m_kiewitz> lol
[21:07] <tsoliman> I was downloading stable
[21:08] <tsoliman> in my defense I've been building my own for a few years now
[21:09] <t0by> lol
[21:10] <criezy> I wanted to see what the KQ6 intro looked like with my PPC build on my old intel core duo mac...
[21:10] <criezy> I don't even get to the intro: it hit the asert at resource_audio.cpp:417 :(
[21:11] <-- ajax16384 left irc: Read error: Connection reset by peer
[21:11] <criezy> The exact same version (the GoG one) works well with the intel version of ScummVM though.
[21:12] Action: t0by squints
[21:12] <t0by> Oh
[21:13] <m_kiewitz> criezy: possibly the same assert that i was getting in another game :/
[21:13] <tsoliman> m_kiewitz: no tearing in kq6 in windows
[21:13] <m_kiewitz> well that's SDL1
[21:13] <tsoliman> daily build yeah
[21:14] <m_kiewitz> anyone has SDL2 build on win32?
[21:14] <m_kiewitz> for me SDL2 building never works out
[21:14] <tsoliman> the only reason I use SDL2 on the mac was because of problems in both ScummVM and Exult that were resolved by either SDL2 or OpenGL
[21:15] <snover> kirbens daily builds come with sdl2
[21:15] <snover> http://scummvm.org/downloads/scummvm-win32.exe
[21:16] <tsoliman> ok grabbing that
[21:16] <m_kiewitz> im so excited :P
[21:17] <tsoliman> I wish we put an SDL feature flag on builds so in the about box we can see SDL1 vs SDL2 on an exe
[21:17] <-- Begas_VBox left irc: Quit: Vision[0.9.8]: i've been blurred!
[21:18] <-- Begasus left irc: Quit: Ex-Chat
[21:18] <snover> anyway, i am not sure offhand how the tearing could be sci engines fault, since surfacesdlgraphicsmanager has its own back buffer, but im not confident in my understanding of OSystem at this point
[21:19] <m_kiewitz> i doubt it's sci engine's fault
[21:19] <m_kiewitz> and well at least for sdl1 win32 it works fine
[21:19] <m_kiewitz> so probably even sdl2 related
[21:19] <tsoliman> I don't see the tearing in KQ6 SDL2 .. trying GK2
[21:19] <tsoliman> I am not 100% sure
[21:20] <tsoliman> I wish I could "move the camera" :)
[21:21] <m_kiewitz> ah well besides calling updateScreen() too many times or something like that
[21:21] <m_kiewitz> although not sure if i would classify that as an engine fault
[21:21] <snover> during vmd playback it should be called only once per frame
[21:22] <m_kiewitz> and it tears in kq6 intro too, which definitely should not happen
[21:22] <m_kiewitz> snover: depends on the frame rate. if you call it not enough times per second, mouse may feel unresponsive. and the window may also be unresponsive
[21:22] <snover> tsoliman: you may try going to room 400 in the debugger
[21:24] <tsoliman> lol the clock is still ticking
[21:24] <tsoliman> thanks that was what I needed though
[21:24] <snover> hmmmm.
[21:24] <tsoliman> ok marienplatz is smooth in the kirben build
[21:25] <tsoliman> let me try back on the mac
[21:25] <t0by> I'm sorry
[21:25] <t0by> allow me
[21:26] <t0by> GK2 woooooooo
[21:26] <m_kiewitz> snover: you are on mac, right?
[21:26] <snover> m_kiewitz: yes
[21:26] <t0by> (it is not being said enough)
[21:26] <m_kiewitz> and you don't get tearing/frame skip?
[21:26] <m_kiewitz> you use SDL2 or 1?
[21:26] <tsoliman> marienplatz tearing on the mac on SDL2 for sure
[21:26] <snover> i do see marienplatz tearing
[21:27] <m_kiewitz> aah
[21:27] <m_kiewitz> snover, and you are on sdl2?
[21:27] <snover> yes
[21:27] <m_kiewitz> can you try sdl1?
[21:27] <m_kiewitz> or is sdl1 broken for you too?
[21:27] <snover> probably.
[21:27] <snover> well, it does not work very well, bt it is not totally broken
[21:28] <tsoliman> yeah SDL1 worked fine in Yosemite
[21:28] <tsoliman> I'll give it a try as well
[21:29] <-- vinayakvivek left irc: Quit: Connection closed for inactivity
[21:29] <snover> i *might* know what could be occurring during the VMD playback. earlier i was about to say i know there is only one updateScreen call because the cursor is hidden and the palettes dont change, but the palettes *do* change on some of these videos
[21:30] <snover> so frameout calls updateScreen once and then GfxPalette32::updateHardware is called a second time
[21:30] <snover> er
[21:30] <snover> GfxPalette32::updateHardware calls it a second time
[21:31] <m_kiewitz> 2 calls shouldn't be a problem
[21:31] <m_kiewitz> even if the frame rate was 30, and it surely isn't for those vmds
[21:32] <m_kiewitz> this could explain palette flashing
[21:32] <m_kiewitz> though
[21:32] <m_kiewitz> i saw that yesterday while trying around
[21:35] <tsoliman> m_kiewitz: sdl1 works fine :(
[21:35] <m_kiewitz> ha
[21:35] <m_kiewitz> gotcha
[21:35] <tsoliman> marienplatz doesn't tear with SDL1
[21:35] <tsoliman> on a mac
[21:36] <m_kiewitz> maybe we are simply supposed to call an API for that. Or maybe SDL2 is simply broken too in other ways :P
[21:36] <m_kiewitz> can you please ask the user on reddit if he/she uses mac / SDL2 too?
[21:36] <tsoliman> windows 10
[21:36] <tsoliman> daily build .. so SDL1
[21:36] <tsoliman> but I will double check
[21:37] <snover> i can verify that, but also verify that cursor movement with sdl1 manages to destroy the framerate
[21:37] <snover> i would guess that it has vsync on and is blocking
[21:40] <m_kiewitz> snover: and you get no frame skipping either?
[21:40] <m_kiewitz> (when you are not moving the mouse)
[21:40] <snover> no skipping during video playback that i can see
[21:40] <snover> definitely some unpleasantness with new palette frames
[21:41] <m_kiewitz> i still wonder why it would cause problems when moving the mouse? maybe the current code isn't collecting all mouse movements, but processes ever movement at once, then updates the screen
[21:41] <m_kiewitz> otherwise i don't understand why moving the mouse would cause problems
[21:41] <m_kiewitz> *every movement
[21:41] <m_kiewitz> at one time
[21:42] <m_kiewitz> if the code processes all mouse movement and then update screen, there definitely shouldn't be a problem with the framerate
[21:42] <tsoliman> those videos are compressed into oblivion .. I've noticed what I would call unpleasantness on what seem to be keyframes but in windows and mac alike .. I assumed it was just massive compression
[21:42] <m_kiewitz> *updates
[21:45] <snover> ssci used mouse interrupt instead of polling, so the cursor would still move even when the game was not receiving events; to try to improve this in scummvm i stuck a call to getSciEvent(SCI_EVENT_PEEK) in frameout, which is not the sole reason for the problem, but makes it much worse
[21:46] <m_kiewitz> but our SCI is using backend for mouse cursor movement, right?
[21:46] <m_kiewitz> at least sci16 does
[21:46] <m_kiewitz> so i wonder how that is handled
[21:46] <m_kiewitz> ah wait, i think you changed something for the mouse cursor limit feature in sci32?
[21:47] <tsoliman> snover: something I noticed in SDL1/OpenGL vs SDL2 .. is the fullscreen animation .. there's 3 ways of doing "fullscreen" on mac .. do you know what I am talking about?
[21:47] <m_kiewitz> the one where the game can tell the engine to not allow the mouse cursor outside of a given rect
[21:48] <m_kiewitz> but wait, kq6 has tearing too
[21:49] <m_kiewitz> and sci16 doesn't handle the cursor internally. it simply tells scummvm about the cursor.
[21:49] <snover> i think there is more than one problem here :)
[21:49] <snover> m_kiewitz: yeah, i did. sci16 calls to EventManager::updateScreen every time EventManager::getSciEvent is called; sci32 does something& slightly different& that i am trying to find right now
[21:50] <snover> here we are.
[21:50] <m_kiewitz> yes, but i wonder what scummvm does internally for mouse cursor movement
[21:50] <m_kiewitz> or is that mouse cursor movement issue with sdl1 exclusive to sci32?
[21:50] <snover> in EventManager::getScummVMEvent, GfxCursor32::deviceMoved is called which restricts the mouse position and then moves the cursor and renders it
[21:51] <snover> and by renders it, i mean, ends up calling OSystem::updateScreen
[21:51] <m_kiewitz> yes, but what do you do when there are multiple mouse movement events?
[21:51] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[21:51] <m_kiewitz> do you process all at once?
[21:51] <m_kiewitz> or just once and then update the cursor and update the screen?
[21:51] <m_kiewitz> s/once/one
[21:53] <snover> getScummVMEvent already forwards past all mousemove events, SCI32 just receives the last mouse position
[21:53] <snover> GfxCursor32* (sorry, my brain is starting to crumble)
[21:53] <m_kiewitz> hmm, what is going on then?
[21:53] <m_kiewitz> why is sdl1 going crazy?
[21:53] <snover> too many updateScreen calls probably
[21:53] <m_kiewitz> but you seem to call it twice only per frame?
[21:54] <snover> i think thats enough to get things stuck if the call is blocked until vsync
[21:54] <snover> mousemovement, updateScreen (till vsync), frameout, updateScreen (till next vsync)
[21:55] <snover> but, thats speculative, i havent verified this is what SDL is doing at all yet
[21:55] <m_kiewitz> sure, but you should get 60 fps
[21:55] <snover> tsoliman: sorry i did not mean to ignore your question. there are 2 ways of fullscreen in sdl2
[21:55] <m_kiewitz> so even if you cause it to do 2 v'syncs, that would still allow for 30 fps
[21:56] <m_kiewitz> i don't know how bad it is, but 30 fps shouldn't feel horrible for mouse movement
[21:59] <snover> i can cause the same thing to happen in sdl2 by changing the SDL_CreateRenderer call to include SDL_RENDERER_PRESENTVSYNC flag
[21:59] <snover> which also fixes the screen tearing
[22:00] <m_kiewitz> ...but?
[22:01] <m_kiewitz> :P
[22:03] <snover> i dont know that there is a but
[22:04] <snover> i think the scrolling is already not happening at 60fps because it is synchronised to the movement of ego
[22:04] <snover> which seems to be 3 frames per cel
[22:06] <snover> anyway, i will work on some solutions to reduce the number of times the screen is updated later today
[22:06] <snover> i need to go away for a while now
[22:06] <m_kiewitz> kk bye
[22:13] vpelletier (vincent@p6205-ipngn5801hodogaya.kanagawa.ocn.ne.jp) left #scummvm.
[22:13] <-- Gentle left irc: Ping timeout: 258 seconds
[22:21] --> macdude22 joined #scummvm.
[22:27] --> Gentle joined #scummvm.
[22:43] <-- frankyboy_ left irc: Remote host closed the connection
[22:45] <-- girafe left irc: Read error: Connection reset by peer
[23:00] <tsoliman> SDL_RENDERER_PRESENTVSYNC does indeed work :)
[23:02] <tsoliman> I wonder if passing in SDL_RENDERER_SOFTWARE will make scummvm not want to switch to the discrete card
[23:03] --> emilknievel joined #scummvm.
[23:06] <tsoliman> nope - it still prefers discrete
[23:06] <-- emilknievel left irc: Client Quit
[23:08] <-- Mia left irc: Read error: Connection reset by peer
[23:37] <-- waltervn left irc: Quit: Leaving
[23:45] <-- Farmboy0 left irc: Remote host closed the connection
[23:47] <-- WooShell left irc: Quit: If you understand or if you don't, if you believe or if you doubt - There's a universal justice, and the eyes of truth are always watching you.
[00:00] --- Sat May 6 2017