#jsmess 2017-12-03,Sun

↑back Search

Time Nickname Message
00:40 🔗 azakai has quit IRC (Read error: Operation timed out)
01:11 🔗 A_Scanner I'm running ubuntu i386/x86 and trying to build. I keep getting some of the same errors. when I try and 'make TARGET=mess' I keep getting this error: "make: leaving directory ../mame/srs/devices/cpu/m6800' make:*** No rule to make target 'scripts/target/mess/mess.lua', needed by 'build/projects/sdl/mess/gmake-linux/Makefile'. Stop
01:13 🔗 A_Scanner I haven't been able to build MAME normally yet. I've been resolving different issues and right now when running a 'make' I get 'undefined reference to 'SEGAM1AUDIO' and 'undefined reference to 'ACCLAIM_RAX'
02:46 🔗 azakai has joined #jsmess
04:27 🔗 binji has joined #jsmess
06:39 🔗 SketchCow A_Scanner: DFJustin and db48x can help
06:42 🔗 bai oh A_Scanner - you don't really want to build "mess" as a target. mess isn't a system, it's just....well...it's hard to explain, but basically MAME tried to rename to MESS, it didn't really take, so they said, okay fine, we'll go back to just calling everything MAME
06:42 🔗 bai it's....well....a mess. but anyway, there is no target called "mess" - your target would be the individual system you want to build
06:43 🔗 bai or for the native built you can just run make without a target, and it'll build the big combined build file
06:43 🔗 bai build*
07:23 🔗 A_Scanner bai: Thanks! It looks like I'm not the only one with this issue on the full build. This person is facing about the same challenges that I am: https://github.com/mamedev/mame/issues/2874
07:31 🔗 A_Scanner Apparently there is an oddball change in mame that *might* break some builds. Sounds like compiling everything up until this commit works: https://github.com/mamedev/mame/commit/39176274946d70ff520f265dee8fbd16d5fe0000 I got latest in hopes it was fixed and no new errors and going for another build. Failing that I'll get up until that checkpoint and try building that.
07:35 🔗 DFJustin I know there have been issues with compiling on 32-bit because the project is so huge now
07:39 🔗 A_Scanner Thanks DFJustin, I'm starting an install of 64 bitness and giving that a try. But first. Sleep! that's where I'm a viking!
08:00 🔗 A_Scanner K.. got both VM's compiling mame right now. x86 vs amd64 fight!
08:27 🔗 DFJustin you'll need a goodly amount of ram too
10:42 🔗 binji has quit IRC (Read error: Operation timed out)
12:03 🔗 A_Scanner yay disk resize.. ran out of room on the 64 bit vm. Building again
15:49 🔗 SketchCow Great
15:50 🔗 A_Scanner I appear to be getting further on the 64 bit build of mame (at least different errors) so I think I'm on the right track
15:51 🔗 SketchCow Also great.
15:52 🔗 SketchCow Do you understand the basic problem we're trying to solve, A_Scanner?
15:53 🔗 A_Scanner There's a timing mismatch with mame/mess and browsers that could be streamlined.
15:55 🔗 A_Scanner Is my gist. There is a google doc started: https://docs.google.com/document/d/1FHmn1kO3PGYYCRJejZsZMFk_BxWer_pOaTUANRFroM8/edit#heading=h.pmazbh81wixz that could use some fleshing out
15:57 🔗 SketchCow I'm literally the one writing that.
15:57 🔗 SketchCow And yes. My hope is that a fresh mind with some time could nail out the problem pretty easily.
15:58 🔗 SketchCow For one thing, we're lucky - Javascript output is now a target built into MAME itself.
15:59 🔗 SketchCow That means it's a matter of making settings for when you're compiling with Emscripten
15:59 🔗 SketchCow I can't begin to say how much of a climbed mountain this is.
16:00 🔗 SketchCow We're just trying to put a little HVAC on the top of the mountain
16:00 🔗 A_Scanner That's pretty cool. Getting a new to you source/toolchain/build set up going is almost always a little fraught. Happy to help
16:00 🔗 A_Scanner Even if I can't fix it I'll document the shit out of what I find.
16:00 🔗 A_Scanner so the next person has good notes.
16:06 🔗 A_Scanner On a personal note, I'm curious if someone has figured a way out to access all the virtualized sound chips and get mame chiptune out of it.
16:07 🔗 SketchCow Well, MAME has lots of functionality for outputting what it generates.
17:02 🔗 SketchCow yay, bai is updating
17:08 🔗 bai :D
17:52 🔗 SketchCow I'll just keep piling people on this problem until it's solved
18:07 🔗 A_Scanner Sounds good.
18:38 🔗 A_Scanner Ok. So reverting to supposedly good check in. Rumor has it that this checkin: https://github.com/mamedev/mame/commit/6829ecb3b037d6bfbe0b84e818d17d712f71bce6 builds ok.
18:45 🔗 DFJustin there's a vgmplayer target that's a chiptune player
18:50 🔗 A_Scanner DFJustin: Thanks! I also saw a kurzweil k2000 emulation thing spin by. Not sure of it's current state but someone's trying to plumb in a k2000 board which I loved
19:18 🔗 binji has joined #jsmess
19:40 🔗 SketchCow bai: Let me know when you think it's "ready"
19:40 🔗 SketchCow And I will darken the skies
19:41 🔗 bai SketchCow: well, I'm trying to retrace my steps on what I did before. having a hard time picking up the trail again. I can't find my original experimental changes, I know the change had to do with how the sleep functions were handled, but now I'm having a hard time finding the art of either emscripten's, mame's, or stdlib's code I was looking at that had be going "a hah!"
19:42 🔗 bai part*
19:44 🔗 A_Scanner I'm still compiling mame on x86 and amd64 of that checkin I linked to. I'll keep you guys posted if that compiles on ubuntu for me.
19:44 🔗 bai I wouldn't recommend doing both at once, heh. just do the 64-bit build
19:45 🔗 bai and use as much parallelism as you can throw at it, -j4 or -j8 depending how many cores you have
19:46 🔗 A_Scanner I'll keep that in mind for future builds.
19:47 🔗 SketchCow In other news, I'm thinking of paying a pixel artist to make icons along a style for the computerlab.archive.org site.
19:47 🔗 SketchCow And I'm wondering if I can find a web designer who wants to help
19:47 🔗 bai cool
19:51 🔗 chartreus has joined #jsmess
20:00 🔗 SketchCow I do fairly believe that once this is on a table in a cogent explanation, people will just fix it
20:05 🔗 bai yeah, it's an interesting enough problem that if people get the explanation, it gets the wheels turning, but there's so much ramp-up on getting a working dev environment
20:06 🔗 bai having the explanation lets people think about it while things are compiling
20:12 🔗 bai oh wow, I have an audience, haha
20:21 🔗 SketchCow Better type faster!!
20:21 🔗 bai I'll throw in some fancy pictures, that'll keep 'em busy!
20:23 🔗 Lord_Nigh has quit IRC (Read error: Operation timed out)
20:23 🔗 Lord_Nigh has joined #jsmess
21:31 🔗 SketchCow Looking good
21:33 🔗 bai I think I covered all the details
21:34 🔗 bai I'd ask if it all makes sense or if I missed something, but, that's the whole problem, it doesn't entirely make sense and we definitely missed something :D
21:39 🔗 SketchCow Looks like the next thing is to come up with the requirements for compiling the environment. Maybe DFJustin or others can help
21:39 🔗 SketchCow I think that's what slows down people who want to help
21:40 🔗 bai a lot of that is covered in mame's docs now, but yeah we could do with more detailed instructions
21:40 🔗 bai especially about hooking up what we've built to emularity so people can test
21:41 🔗 SketchCow Even some links to how to build from the docs is helpful
21:41 🔗 SketchCow I notice that's what's slowed down A_Scanner
21:42 🔗 bai yeah I put that up at the top, under "Compiling prerequisites"
21:42 🔗 bai although there's a bad page break so one of them kind of gets lost
21:43 🔗 bai and I kind of trail off there too, like you said, needs a full section later on
21:43 🔗 SketchCow Found it, added a page break
21:43 🔗 bai cool, better
21:45 🔗 SketchCow Still people looking at it, so that's something
22:53 🔗 SketchCow A_Scanner: Need help with anything?
22:53 🔗 SketchCow Does the document help now?
22:53 🔗 SketchCow anyone else in the channel help?
22:59 🔗 i0npulse has quit IRC (Ping timeout: 248 seconds)
22:59 🔗 Rai-chan has quit IRC (Ping timeout: 248 seconds)
23:00 🔗 i0npulse has joined #jsmess
23:02 🔗 Rai-chan has joined #jsmess
23:08 🔗 A_Scanner SketchCow: I just got it all to compile on ubuntu! x86 and amd64.
23:15 🔗 A_Scanner From a new ubuntu VM I had to install the following to get mame to build. git make gcc mesa-common-dev qt5-default qttools5-dev-tools libsdl2-dev libsdl2-ttf-dev emscripten libfontconfig1-dev
23:16 🔗 A_Scanner I tried to build from the latest multiple times over the weekend and in ubuntu that didn't work for me. I found out that this version builds ok: https://github.com/mamedev/mame/commit/6829ecb3b037d6bfbe0b84e818d17d712f71bce6
23:17 🔗 bai good to know, yeah there's always the possibility of either mame, emscripten, or the browsers themselves changing things which break the whole system
23:17 🔗 bai part of the fun I guess :P
23:22 🔗 A_Scanner One dev thinks its just an 'environment' issue but an awful lot of people run ubuntu.. I don't have the cycles to devote to figuring out why the latest mame doesn't compile right on ubuntu.
23:24 🔗 A_Scanner Speaking of cycles I've got day job work I need to get done this weekend still. The google doc does a much better job of describing the problem and I think it's an interesting problem so it should attract developer interest.
23:29 🔗 SketchCow Great
23:31 🔗 bai yeah, and it helps me to put it all in to words, the act of writing it down made me go back and remember most of what I'd forgotten
23:33 🔗 A_Scanner has quit IRC (Leaving)
23:45 🔗 SketchCow ....and then fix it!
23:45 🔗 * SketchCow waits hopefully
23:45 🔗 SketchCow I'm sure it's 3 lines
23:53 🔗 sebbert has joined #jsmess
23:54 🔗 SketchCow https://skeletonplanet.com/test/timeout_tester.html
23:54 🔗 SketchCow https://twitter.com/cairnsh/status/937468413697495040

irclogger-viewer