[Back to Index]

  
[00:13] <-- SylvainTV left irc: Read error: Connection reset by peer
[00:14] --> dreammaster joined #scummvm.
[00:14] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[00:25] <tsoliman> do you think adding "SDL1" or "SDL2" to the about box feature list (when compiled with SDL) is a good idea?
[00:27] <snover> the entire version from SDL_GetVersion, perhaps
[00:27] <snover> of course, who wants to bet this is an SDL2 invention&
[00:28] <tsoliman> :D
[00:28] <tsoliman> I thought about that but then what happens when it is dynamically linked? does it have to match the exact version?
[00:29] <snover> that is the reason for using SDL_GetVersion and not SDL_VERSION
[00:29] <snover> but, good news, wed have to report SDL1 as just being SDL1 in any case since that function does not exist in SDL1
[00:30] <tsoliman> ah
[00:30] <tsoliman> so resolve it at runtime
[00:37] <-- Lightkey left irc: Ping timeout: 255 seconds
[00:50] --> Lightkey joined #scummvm.
[01:16] <-- ccawley2011 left irc: Quit: Page closed
[01:30] --> DominusExult joined #scummvm.
[01:34] <-- Dominus left irc: Ping timeout: 260 seconds
[01:34] Nick change: DominusExult -> Dominus
[01:39] <tsoliman> snover: you said in your email to make sure getWidth() and getOverlayWidth() return the correct values .. I don't understand what these are supposed to be .. I am reading Doxygen off of the wiki
[01:40] <snover> doxygen off the wiki?
[01:40] <tsoliman> http://doxygen.scummvm.org/d9/df4/classOSystem.html#a09ffdad22d90db55c78774c75cbac711
[01:40] <tsoliman> (I clicked the link in the wiki)
[01:40] <-- dhewg left irc: Ping timeout: 240 seconds
[01:41] <tsoliman> I don't understand (in the context of SDL let's say) what is the overlay vs the "virtual screen"
[01:41] <snover> the virtual screen is the hardware screen that the game uses
[01:41] <snover> the overlay is the launcher
[01:48] --> dhewg joined #scummvm.
[01:48] #scummvm: mode change '+o dhewg' by ChanServ!ChanServ@services.
[02:16] <-- ced117_ left irc: Ping timeout: 240 seconds

[02:30] --> ced117 joined #scummvm.
[02:52] <-- dreammaster left irc: Ping timeout: 240 seconds
[03:32] <-- Cheeseness left irc: Quit: Leaving.
[03:37] --> Cheeseness joined #scummvm.
[05:13] <tsoliman> snover: before your refactor, SurfaceSdlGraphicsManager::notifyResize did nothing unless SDL_VERSION_ATLEAST(2, 0, 0) ... now it calls handleResize() in SDL1 .. that's the source of the regression I think
[05:17] <snover> tsoliman: opengl in sdl1 needs that
[05:17] <tsoliman> before and after the refactor I get the weird SDL_VIDEORESIZE with the full screen resolution, not matching the actual fullscreen window mode size .. I think it is part of the modified SDL for MAEMO (I dunno) .. but before the refactor it was a NOOP ...
[05:17] <snover> oh
[05:17] <snover> sorry
[05:17] <snover> surfacesdl
[05:18] <snover> pretty sure its still needed for games that switch resolutions regardless
[05:18] <snover> the windowedgraphicsmanager cant do its job if it doesnt know the size of the window
[05:19] <tsoliman> regardless of why that event is received, it wasn't processed in SDL1 and the current processing of it is bad .. it should be a full resolution change with SDL_SetVideoMode according to the SDL doc but I don't think that's relevant for scummvm since ScummVM SDL1 windows aren't resizable
[05:20] <snover> ok, well, talk to me about it tomorrow/later. i need to go to sleep.
[05:20] <snover> we can figure something out.
[05:23] <-- tsoliman left irc: Ping timeout: 240 seconds
[05:35] --> tsoliman joined #scummvm.
[05:35] #scummvm: mode change '+o tsoliman' by ChanServ!ChanServ@services.
[05:35] <tsoliman> internet outage
[05:37] <tsoliman> internet's back
[05:39] <tsoliman> this is the text I wrote during the outage (znc thinks it is sent but logs.scummvm.org doesn't think so .. so here they are FWIW):
[05:39] <tsoliman> > but it doesn't switch the resolution though
[05:39] <tsoliman> > the window size is the still the same old size - just handleResize() is called, which doesn't ultimately result in a SDL_SetVideoMode as far as I can debug
[05:39] <tsoliman> > my understanding is that handleResize is supposed to be called AFTER a resize request (and is not in itself a resize request) .. am I correct in this understanding?
[05:40] <tsoliman> ok thanks I'll hit you up tomorrow :)
[06:14] <-- Dark-Star left irc: Ping timeout: 255 seconds
[06:19] --> Dark-Star joined #scummvm.
[06:58] <-- _sev_ left irc: Quit: This computer has gone to sleep
[07:45] --> waltervn joined #scummvm.
[07:45] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[07:45] <waltervn> morning
[07:46] --> rootfather joined #scummvm.
[07:50] <-- rootfather left irc: Changing host
[07:50] --> rootfather joined #scummvm.
[07:50] #scummvm: mode change '+o rootfather' by ChanServ!ChanServ@services.
[07:52] <-- waltervn left irc: Ping timeout: 248 seconds
[07:56] <-- edheldil left irc: Remote host closed the connection
[07:57] --> edheldil joined #scummvm.
[08:01] --> _sev|work joined #scummvm.
[08:01] <-- _sev|work left irc: Changing host
[08:01] --> _sev|work joined #scummvm.
[08:01] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[08:06] <-- LittleToonCat left irc: Remote host closed the connection
[09:04] <-- kurtwr left irc:
[09:12] --> kurtwr joined #scummvm.
[09:24] <-- TMM left irc: Quit: Ex-Chat
[09:48] --> ccawley2011 joined #scummvm.
[10:01] <rootfather> hi folks
[10:01] <rootfather> just a quick question
[10:01] <rootfather> is it intentional that when compiling on Linux, scummvm doesn't use the modern theme as default until I set it manually in the preferences?
[10:02] <rootfather> when I run the compiled executable, it defaults to the "old" inbuilt theme without translations and such
[10:02] <rootfather> on Windows it detects the theme path properly
[10:07] <wjp> did you install the build properly, or running from the source tree?
[10:08] --> rootfather|linux joined #scummvm.
[10:13] --> jamm joined #scummvm.
[10:13] <-- jamm left irc: Changing host
[10:13] --> jamm joined #scummvm.
[10:13] <rootfather> running from the source tree
[10:13] <rootfather> but I do this all time on my mingw64 toolchain on windows without any problems
[10:13] <rootfather> therefore I expected it to work on Linux the same way
[10:16] <rootfather> it's obviously no big deal to set the path on time in the config file... I was just wondering *why* it behaves differently
[10:16] <wjp> on Windows I think we embed data files into the .exe
[10:16] <wjp> although I'm not 100% sure
[10:16] <rootfather> What I know is that at least the translations.dat is embedded into the .exe
[10:17] <rootfather> not sure about the theme data
[10:17] <wjp> scummmodern.zip FILE "gui/themes/scummmodern.zip"
[10:17] <wjp> so that seems to be added too
[10:17] <wjp> (scummvm.rc.in)
[10:18] <rootfather> could this be done for Linux too?
[10:18] <wjp> in theory
[10:25] <-- rootfather|linux left irc: Quit: Konversation terminated!
[11:13] --> criezy|Work joined #scummvm.
[11:13] #scummvm: mode change '+o criezy|Work' by ChanServ!ChanServ@services.
[11:13] <criezy|Work> Hi guys
[11:21] <criezy|Work> rootfather: what I am doing on macOS where we have the same issue is to set the path to a relative path to "gui/themes" in the config.ini so that it works whichever source tree I am running from, or when I am relocating them.
[11:44] <-- Storm-AFK left irc: Quit: KVIrc 4.9.2 Aria http://www.kvirc.net/
[11:51] --> GitHub143 joined #scummvm.
[11:51] <GitHub143> [scummvm] maikudou opened pull request #1044: IOS7: Updated app icons and splash screens to be more contemporary and clean (master...ios7-icons-splash) https://git.io/vd7SX
[11:51] GitHub143 (GitHub143@192.30.252.35) left #scummvm.
[12:04] --> TMM joined #scummvm.
[12:04] <-- TMM left irc: Changing host
[12:04] --> TMM joined #scummvm.
[12:04] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[12:04] <-- _sev|work left irc: Quit: This computer has gone to sleep
[12:06] <wjp> huh, changing logo/branding in a PR is a new one
[12:07] <-- user9 left irc: Ping timeout: 248 seconds
[12:11] --> _sev|work joined #scummvm.
[12:11] <-- _sev|work left irc: Changing host
[12:11] --> _sev|work joined #scummvm.
[12:11] #scummvm: mode change '+o _sev|work' by ChanServ!ChanServ@services.
[12:30] <criezy|Work> Urgh, a flat icon! I hate those. But that would indeed be more consistent with other apps in recent iOS versions.
[12:35] <wjp> I vote for skipping this whole flat design craze and waiting until some sanity returns to UI design
[12:39] <TMM> scummvm looks great as it is! :D
[12:39] <TMM> although I was never a fan of all the gradients on the site :P
[12:46] --> user9 joined #scummvm.
[13:25] --> Littleboy joined #scummvm.
[13:25] #scummvm: mode change '+o Littleboy' by ChanServ!ChanServ@services.
[13:26] <-- Cheeseness left irc: Quit: Leaving.
[13:51] --> GitHub56 joined #scummvm.
[13:51] <GitHub56> [scummvm] maikudou closed pull request #1044: IOS7: Updated app icons and splash screens to be more contemporary and clean (master...ios7-icons-splash) https://git.io/vd7SX
[13:51] GitHub56 (GitHub56@192.30.252.40) left #scummvm.
[14:04] --> TMM_ joined #scummvm.
[14:16] <-- TMM_ left irc: Read error: Connection reset by peer
[14:17] <-- user9 left irc: Quit: user9
[14:17] --> TMM_ joined #scummvm.
[14:17] <-- TMM left irc: Read error: Connection reset by peer
[14:18] <-- TMM_ left irc: Remote host closed the connection
[14:19] --> TMM joined #scummvm.
[14:19] <-- TMM left irc: Changing host
[14:19] --> TMM joined #scummvm.
[14:19] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[14:28] --> user9 joined #scummvm.
[14:37] <-- user9 left irc: Quit: user9
[14:45] <-- _sev|work left irc: Quit: This computer has gone to sleep
[14:50] --> user9 joined #scummvm.
[15:08] --> Henke37 joined #scummvm.
[15:14] <-- TMM left irc: Quit: Ex-Chat
[15:37] <-- jamm left irc: Ping timeout: 252 seconds
[15:56] --> waltervn joined #scummvm.
[15:56] #scummvm: mode change '+o waltervn' by ChanServ!ChanServ@services.
[15:56] --> ny00123 joined #scummvm.
[16:16] --> _sev_ joined #scummvm.
[16:16] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[16:23] <-- _sev_ left irc: Quit: This computer has gone to sleep
[16:31] --> _sev_ joined #scummvm.
[16:31] <-- _sev_ left irc: Changing host
[16:31] --> _sev_ joined #scummvm.
[16:31] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[16:31] --> TMM joined #scummvm.
[16:31] <-- TMM left irc: Changing host
[16:31] --> TMM joined #scummvm.
[16:31] #scummvm: mode change '+o TMM' by ChanServ!ChanServ@services.
[16:39] <-- _sev_ left irc: Quit: This computer has gone to sleep
[16:45] --> LittleToonCat joined #scummvm.
[16:46] --> _sev_ joined #scummvm.
[16:46] <-- _sev_ left irc: Changing host
[16:46] --> _sev_ joined #scummvm.
[16:46] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[16:48] --> ajax16384 joined #scummvm.
[16:48] #scummvm: mode change '+o ajax16384' by ChanServ!ChanServ@services.
[16:59] --> Guest75820 joined #scummvm.
[17:00] <-- Tomaz^ left irc: Ping timeout: 240 seconds
[17:00] --> Farmboy0 joined #scummvm.
[17:00] <-- Farmboy0 left irc: Changing host
[17:00] --> Farmboy0 joined #scummvm.
[17:14] <-- _sev_ left irc: Ping timeout: 252 seconds
[17:18] --> girafe joined #scummvm.
[17:38] <tsoliman> rootfather: this is what I do to run it in-tree on linux (and anything with "make" really) ... make a directory called build and cd to it .. then "../configure" .. then create symlinks to the themes
[17:39] <tsoliman> I also do this for the engine dat files ... this is because scummvm will look in the current working directory for all of this stuff (in addition to other paths)
[17:42] kusushi[m] <-- (kusushimat@gateway/shell/matrix.org/x-ktmkcyzltwqoostq) left irc: Ping timeout: 240 seconds
[17:42] trojan[m] <-- (trojanmatr@gateway/shell/matrix.org/x-jsbftrvedsnjfodb) left irc: Ping timeout: 240 seconds
[17:42] weezywinez[m] <-- (weezywinez@gateway/shell/matrix.org/x-ijjxxttpcxxxknnv) left irc: Ping timeout: 240 seconds
[17:42] Coldwine[m] <-- (coldwinema@gateway/shell/matrix.org/x-libsvmpcrnfdhdvi) left irc: Ping timeout: 240 seconds
[17:42] super_fox2[m] <-- (superfox2m@gateway/shell/matrix.org/x-ngzqrtapprpvkvxv) left irc: Ping timeout: 252 seconds
[17:43] tomazcuber[m] <-- (tomazcuber@gateway/shell/matrix.org/x-mqvwwovazgpjmdij) left irc: Ping timeout: 255 seconds
[18:02] <-- Guest75820 left irc: Ping timeout: 258 seconds
[18:22] --> SylvainTV joined #scummvm.
[18:22] #scummvm: mode change '+o SylvainTV' by ChanServ!ChanServ@services.
[18:25] <-- girafe left irc: Read error: Connection reset by peer
[19:25] <-- ajax16384 left irc: Read error: Connection reset by peer
[19:30] --> girafe joined #scummvm.
[19:37] --> CHCl3 joined #scummvm.
[19:39] <Henke37> https://www.youtube.com/watch?v=HqWrdIf69M8
[19:43] --> _sev_ joined #scummvm.
[19:43] <-- _sev_ left irc: Changing host
[19:43] --> _sev_ joined #scummvm.
[19:43] #scummvm: mode change '+o _sev_' by ChanServ!ChanServ@services.
[19:47] <-- _sev_ left irc: Client Quit
[20:12] <snover> would anyone care to assist in generating some docker images for buildbot?
[20:16] --> Tomaz^ joined #scummvm.
[20:41] --> Tomaz joined #scummvm.
[20:42] Nick change: Tomaz -> Guest97443
[20:43] <-- Tomaz^ left irc: Ping timeout: 246 seconds
[20:45] <criezy|Work> snover: what does that entail? I am completly new to this, so I don't know what is needed.
[20:45] <criezy|Work> (I am leaving now to go back home, so you have about 40 minutes to think of an answer)
[20:46] <-- criezy|Work left irc: Quit: Page closed
[20:51] --> Cheeseness joined #scummvm.
[20:58] <wjp> snover: what does your current done/todo list look like?
[20:59] <-- Henke37 left irc: Quit: ERR_SHUTDOWN
[20:59] <snover> wjp: everything is going to https://github.com/csnover/scummvm-docker as i am finishing. so far, it has not been blazing fast, since i am still learning how this works, and also trying to decide how much effort to spend on actually doing these builds properly from source so they are reproducible, versus just dumping the binaries from vm2 into containers and calling it a day
[21:00] <snover> since criezy asked, Im working on expanding on the instructions for adding new workers, so its clearer how to do this, for people not familiar with docker.
[21:00] <snover> right now i am working on the apple cross-compiler. macos is working, i am working on adding the compiler for ios.
[21:03] <snover> ill add some issues to the issue tracker for each of the outstanding things so its trackable.
[21:03] <snover> in the future it could be nice to require porters to give an image to the buildbot so the builds can always be made.
[21:05] <wjp> at some point TMM and I both had windows cross-toolchain setup scripts in various states of completeness
[21:05] <snover> windows cross-compilation using mingw-w64 from the debian repo is working now :)
[21:06] <wjp> great
[21:06] <wjp> ah yes, I see the debian Dockerfile installs g++-mingw-w64
[21:06] <wjp> cool
[21:06] <snover> ah, doing a bit of quick research here, it looks like all the (very few) x86 androids have an ARM compatibility later and google stopped providing updates for Play Services to MIPS around Android 4.0, so I guess just an ARM compiler will be fine
[21:07] <snover> There are 11571 ARMv7A-based, 359 x86-based, and 478 ARMv5-based device models. I've published an app to Google Play with MIPS as the only supported architecture, and that gave me 5 supported device models.
[21:11] <wjp> can I easily run these scripts to set up a buildbot on my local machine?
[21:11] <snover> yes
[21:11] <snover> get docker, clone the repo, run `build-images.sh all` and `docker-compose up` and youll be up and running
[21:12] <snover> er, and add a secrets.cfg.
[21:12] <wjp> how much disk space will it roughly take?
[21:12] <snover> lets see&
[21:12] <-- Guest97443 left irc:
[21:13] --> Tomaz^ joined #scummvm.
[21:13] <snover> the debian worker is 1.67GB, the buildmaster is 180MB, the nintendo worker is 868MB
[21:13] --> criezy joined #scummvm.
[21:13] #scummvm: mode change '+o criezy' by ChanServ!ChanServ@services.
[21:13] <TMM> wjp, the cross toolchain stuff I hda works
[21:14] <snover> since i havent published complete images to docker hub yet, the total size will be somewhat higher since you will be downloading the intermediate images
[21:14] <TMM> it's on github! :D
[21:15] <snover> there are also a couple of modifications to scummvms Makefile which I guess I will just commit to the master repository now, to get the list of dist-files and executables, since previously those were being hard-coded into the buildmaster config (hence why Starship Titanic needed some special instructions about downloading a titanic.dat)
[21:15] <TMM> snover, I can adapt my scripts to poop out a dockerfile?
[21:16] <snover> TMM: perhaps, though this is generating windows binaries already. what does your script do thats different?
[21:16] <TMM> snover, I'm also patching all the win32 dependencies and make sure we only have the stuff we need, particularly around the fluidsynth stuff that was a pain in the ass
[21:16] <TMM> snover, I don't know, depends on what you do. I didn't want to use the packaged dependencies due to issues with like 3 versions of libpng ending up in the .exe
[21:18] <TMM> snover, I'll have a look at your build environment maybe I can just merge my work on the dependencies in yours
[21:20] <criezy> snover: I assume that you only looked at intel (and no PPC) macOS cross-compiler?
[21:20] <wjp> hrm, my docker-build version seems to need the path (.) after the options (like -t, --build-arg)
[21:20] <snover> criezy: at the moment, yes. im not sure what would be needed to get a PPC one working.
[21:20] <TMM> just a gcc though? I can do that if that's a priority
[21:21] <TMM> building the dependencies is pretty easy
[21:21] <snover> criezy: it could be possibly as little as just running cctools with the correct processor type.
[21:21] <snover> (cctools: https://github.com/tpoechtrager/cctools-port)
[21:22] <snover> wjp: interesting. that is certainly an easy enough change to make.
[21:22] <criezy> I don't see PPC in the supported target though.
[21:23] <wjp> I'm updating docker now; probably have a version about a year old
[21:24] <snover> criezy: doh, i was looking at supported hosts. :) so not that simple&
[21:24] <TMM> snover, would you like me to have a look at this too? If you're just happy with what you have now I don't want to step on your toes. Also I have been rather absent lately so I can understand if you just want me to gtfo :P
[21:25] <snover> TMM: im open to whatever! im just trying to get this up and going so that its maintainable moving forward, and so i can get back to reverse-engineering stuff :)
[21:25] <wjp> snover: updated version is ok with current syntax
[21:25] <TMM> snover, fair :)
[21:25] <snover> criezy: the fact of the matter is that so long as you can get any OS environment that can do PPC builds, which is new enough that it can run a compatible version of buildbot-worker, it can be put into a container and it will work to do builds for us.
[21:26] <TMM> Since you guys are here, I don't know if I told you guys but I had a bit of a burnout for well, quite a while. I've started working on porting the mve stuff to the kingdom engine now though. I'm slowly coming back to doing 'things'. I was some days literally feeling to stressed to make tea, it was very strange.
[21:26] <TMM> I didn't really intend to disappear
[21:26] <snover> TMM: ugh, burnout is the worst. i am only now feeling mostly recovered from burning out 2.5 years ago.
[21:27] <snover> so i totally understand where you are coming from, no apologies needed.
[21:27] <wjp> sorry to hear that :-(
[21:27] <TMM> thanks, I was at one point feeling stressed about feeling too stressed to do things. And not doing things was as stressful as doing things. :-/
[21:27] <criezy> I have a OS X 10.6 environment setup for PPC builds. It probably doesn't qualify as "new enough" though.
[21:27] <criezy> And there might be some licensing issue with using macOS in a container I suppose.
[21:27] <TMM> Anyway, I'll be coming back, slowly. :)
[21:29] <criezy> TMM: welcome back. I heard a bit about it from Strangerke as we were both in SF last weekend.
[21:29] <TMM> criezy, thanks :)
[21:30] <TMM> I did some super low-intensity stuff for godot, but that was it. Played a lot of games :)
[21:30] <-- Tomaz^ left irc: Read error: Network is unreachable
[21:31] <wjp> snover: looks like this needs an ever more recent Docker since you use an ARG in the FROM
[21:32] --> Tomaz^ joined #scummvm.
[21:32] <snover> wjp: i didnt intended to be quite so bleeding-edge! :) these things can be backed off if needed so that a wider range of development environments can be supported.
[21:32] Action: wjp updates even further
[21:34] <wjp> now this one just needs some extra kernel options...
[21:35] <wjp> but let me just ignore those for now
[21:35] <snover> see? easy! &
[21:35] <TMM> lol
[21:36] <TMM> docker is such a trainwreck :P
[21:37] --> Stormkeeper joined #scummvm.
[21:42] <wjp> snover: - for dockerfile in $(find workers -name Dockerfile); do
[21:42] <wjp> + for dockerfile in $(find workers -name _template -prune -o -name Dockerfile -print); do
[21:43] <snover> wjp: 👍
[21:43] <-- CHCl3 left irc: Ping timeout: 252 seconds
[21:48] <wjp> ok, workers built
[21:51] <snover> ok. now you just need to create a secrets.cfg (you can just set all the keys to None) next to the master.cfg, per the README.md, and then docker-compose up should start the whole thing and you can access it at http://localhost:28453
[21:54] Action: wjp installs docker-compose
[21:56] --> CHCl3 joined #scummvm.
[21:57] <wjp> hurray
[22:00] super_fox2[m] --> (superfox2m@gateway/shell/matrix.org/x-binfrlxcsdqjposv) joined #scummvm.
[22:09] <-- Farmboy0 left irc: Remote host closed the connection
[22:10] <Lightkey> criezy: so, what do you think about an additional platforms entry in the infobox to indicate which of the original releases are supported?
[22:13] <criezy> For the game pages on the wiki?
[22:14] <criezy> Maybe this should be specified in the support field (especially as supported platforms may depend on the ScummVM version)?
[22:14] --> rsn8887 joined #scummvm.
[22:28] tomazcuber[m] --> (tomazcuber@gateway/shell/matrix.org/x-rgzajpxhdltoopbh) joined #scummvm.
[22:28] trojan[m] --> (trojanmatr@gateway/shell/matrix.org/x-agjiuomkggthzkut) joined #scummvm.
[22:28] weezywinez[m] --> (weezywinez@gateway/shell/matrix.org/x-atoegeswccwsplsy) joined #scummvm.
[22:28] kusushi[m] --> (kusushimat@gateway/shell/matrix.org/x-uaxrintlzyxndxks) joined #scummvm.
[22:28] Coldwine[m] --> (coldwinema@gateway/shell/matrix.org/x-muaexiozdlfkocyv) joined #scummvm.
[22:39] <snover> wjp: which version of docker were you using that did not work?
[22:42] <snover> also, criezy, wjp, TMM, ive updated the repository with a better default docker-compose.yml for development, added a quick-start section in the readme, and added some tips at the end of the readme for common things youll probably want to do with docker that might not be immediately apparent
[22:44] <wjp> 17.03 was too old for the ARG/FROM thing
[22:47] <wjp> do I need to manually clone the scummvm repo for the first build?
[22:47] <wjp> looking at docker history, 17.06 might be new enough for that. I'm on 17.09 myself now
[22:47] <snover> you should be able to just run a manual build from inside the Buildbot UI once youve logged in
[22:48] <snover> its unclear to me if there is a page to run a manual build against all builders at once or not. so far it seems like you have to do it against one builder at a time, which seems unhelpful, so theres probably a way to do it differently.
[22:48] <wjp> I've seen that, but it seems to be lacking 'changes' to build?
[22:49] <snover> even if there are no changes in the system, it should still work, by pulling the branch you tell it in the manual build form. that is how i tested things before i set up the change hooks.
[22:50] <wjp> or maybe the red crosses before the workers are the problem
[22:50] <snover> ah, yeah, i think that means the worker isnt connected
[22:51] <wjp> oh here we go
[22:55] <tsoliman> snover: can you give me a test case that you were concerned about yesterday with this change? https://github.com/tsoliman/scummvm/commit/2fea095fc7202a6707c6f5499565f125927ae544
[22:55] --> dreammaster joined #scummvm.
[22:55] #scummvm: mode change '+o dreammaster' by ChanServ!ChanServ@services.
[22:55] <wjp> I'm not convinced this new interface is an improvement over the version we're currently running on vm2 :-)
[22:56] <wjp> everything's jumping about
[22:56] <snover> yeah, i am not crazy about this UI, it is pretty terrible. at least it is ajaxified so you can watch live build output.
[22:57] <snover> the waterfall view is somewhat better.
[22:57] <snover> the rest of it is& poor.
[22:57] <wjp> oh, that's where I am
[22:57] <wjp> that makes me afraid to look at the rest :-)
[22:58] <snover> for whatever reason, they decided that when you reconnect to the websocket after losing a connection it will reload the entire page
[22:59] <wjp> anyway, I have a first successful build. Hurray
[22:59] <wjp> Very nice work setting this up
[22:59] <snover> other than that, the rest of the UI issues should be fixable by beating their CSS into submission. though frankly, im more concerned about the functional aspects, like running failed builds first, and actually being able to maintain the thing, so i am willing to forgive the awful parts.
[23:00] <snover> tsoliman: games like DreamWeb CD which changes virtual screen resolutions during the intro
[23:00] <tsoliman> I'll give it a spin
[23:01] <snover> there are fixes in the SDL2 code to stop it from also resizing the window, which i didnt bother to get working with SDL1 because i consider SDL1 to be a legacy backend so didnt have a concern about ensuring total feature parity
[23:02] <wjp> hm, 21GB docker directory :-)
[23:02] Action: wjp will have to look into moving that somewhere with more space
[23:03] <wjp> oh, debian-i686-windows build failed
[23:03] <snover> once the prebuilt images are stable and get uploaded to Hub, things should end up being less crazy in that regard.
[23:04] <wjp> with an X11.h error of all things
[23:04] <wjp> uh, X11/Xlib.h
[23:05] <wjp> I'd look closer, but I really should be going
[23:06] <snover> i guess SDL2 is a little confused about which platform its really building on
[23:07] <-- CHCl3 left irc: Ping timeout: 246 seconds
[23:10] <tsoliman> it's crashing on a mixer assertion - my VM doesn't have a soundcard :(
[23:10] <snover> #undef SDL_VIDEO_DRIVER_X11 is in the future of this backends code i think
[23:13] <snover> and/or, the backend code needs to #define SDL_config_h_ and then provide the win32 version explicitly?
[23:14] <snover> or myself or TMM or someone needs to do a bit more work to compile the third-party libraries, which in my crazy night-time stupor i apparently failed to do
[23:17] <TMM> snover, I don't mind looking into that, all the domain knowledge for that is already in my scripts. I can just translate it to your docker thing.
[23:18] <snover> sounds good to me!
[23:19] <snover> TMM: i need to run for a bit, but i added some stuff to the README which hopefully makes the Docker thing approachable
[23:23] --> CHCl3 joined #scummvm.
[23:23] <snover> if not i will be around for questions. i only learned this last week so its still a bit of an exploration for me.
[23:23] jpcf[m] --> (jpcfmatrix@gateway/shell/matrix.org/x-gfdkyleqtgonwtqh) joined #scummvm.
[23:24] <TMM> we're doing docker stuff at work
[23:24] <TMM> maybe I can do some work on there in general, I haven't really looked at your setup yet
[23:24] jpcf[m] (jpcfmatrix@gateway/shell/matrix.org/x-gfdkyleqtgonwtqh) left #scummvm.
[23:33] <-- ny00123 left irc: Quit: Leaving
[23:33] <-- CHCl3 left irc: Quit: Saliendo
[23:34] --> Zaibitsu joined #scummvm.
[23:43] <-- criezy left irc: Quit: criezy
[00:00] --- Fri Oct 20 2017