#jsmess 2017-03-14,Tue

↑back Search

Time Nickname Message
02:33 🔗 phe has joined #jsmess
03:03 🔗 SketchCow So, what next goes into the machine
03:03 🔗 SketchCow bai: Did you finish compiling
03:04 🔗 bai no I ran into another problem, one which has cropped up in the past but I'm having a hard time recalling the solution
03:04 🔗 bai related to mame defaulting to -Werror and mame including sys/poll.h instead of poll.h
03:05 🔗 bai DFJustin do you remember the flag to fix that? -s NOWERROR=1 or something like that?
03:05 🔗 bai can't remember if it's a make option, a cflag, or an emscripten option
03:05 🔗 bai and googling isn't being very helpful
03:44 🔗 SketchCow By the way, it appears hitsthings "solved" the screenshot problem with phantom.
03:44 🔗 SketchCow The resulting output looks a little off, but I'm sure folks here can figure that out
03:44 🔗 SketchCow I have to reproduce it on my own system though
03:44 🔗 bai cool
03:46 🔗 SketchCow Also: Got Apple II working again, now ingesting Apple II floppies from a donated pile.
03:55 🔗 Vito` SketchCow: what was the solution?
03:58 🔗 SketchCow He deadfucked React
04:00 🔗 SketchCow ./emmake make -j4 SUBTARGET=$1 NOWERROR=1 REGENIE=1 SOURCES=src/mame/drivers/$1.cpp
04:01 🔗 bai cool thanks, that's what I needed
04:02 🔗 SketchCow I'm remaking the build script :)
04:02 🔗 SketchCow I deleted something from it by mistake once and was soooo sad
04:14 🔗 SketchCow I'm trying to remember how I built it.
04:14 🔗 SketchCow I basically need to have it find every emulator:* setting.
04:15 🔗 bai oh, like one of the mame -listsomething options?
04:37 🔗 SketchCow No, no, it's an IA thing
04:37 🔗 SketchCow Just have to get the drivers back
04:37 🔗 SketchCow So I can do the builds
05:00 🔗 SketchCow Yeah, my builds blow up I guess.
05:01 🔗 SketchCow Trying another one.
05:02 🔗 SketchCow I'd rather we spent time making the magic glowing machine work than my own craptastic builds
05:15 🔗 SketchCow The disk imaging is going along swimmingly
05:18 🔗 SketchCow I'm imaging while watching a Game Developers Conference 2017 talk on Oregon Trail
05:18 🔗 SketchCow http://gdcvault.com/play/1024251/Classic-Game-Postmortem-Oregon-Trail
05:23 🔗 SketchCow root@teamarchive0:/0/JSMESS/mame# du -sh mameapple2.js.gz
05:23 🔗 SketchCow 4.6M mameapple2.js.gz
05:23 🔗 SketchCow Does that seem large?
05:25 🔗 bai not sure what the revious builds were....but that sounds about normal
05:26 🔗 SketchCow mameapple2.js.gz 12-May-2016 16:32 3.7M
05:30 🔗 bai presumably building asmjs not wasm right?
05:30 🔗 SketchCow Right
05:30 🔗 SketchCow I mean, compile those wasms, sir, don't waste time on my little jaunt
05:31 🔗 bai hmm, someone in #emscripten mentioned making sure to compile with flags for release mode, but that may only apply to the compiler itself, not mame
05:31 🔗 DFJustin should be the default for compiling mame
05:32 🔗 DFJustin js compiles are bigger than they used to be too, I think it's mame bloat
05:35 🔗 SketchCow I definitely had arcade_mrdo fail
05:35 🔗 SketchCow I'm doing Apple2e right now
05:35 🔗 SketchCow Maybe my JSMESS environment on FOS is crusty
05:35 🔗 SketchCow I updated mame of course
05:37 🔗 SketchCow Oregon Trail postmortem does not disappoint
05:37 🔗 SketchCow And thankfully, these Apple floppies are really in good shape
05:38 🔗 SketchCow And this guy was a straight up pirate, so he has a lot of stuff
05:42 🔗 DFJustin in particular there's a thing where noto sans is compiled in as the default font now and it takes a meg or two, vas was going to make that optional and then forgot I think so I should probably poke him
05:54 🔗 DFJustin hmm no that only makes 30kb difference, I guess the worst of it was fixed
05:59 🔗 DFJustin but anyway stuff like this keeps being added https://github.com/mamedev/mame/commit/ff01b716711b97c2fcaa709ea97f7650f106aa10
06:02 🔗 SketchCow About to cold-swap AppleIIe driver, hold my veggie smoothie
06:04 🔗 DFJustin https://archive.org/details/EmulatorDetectionIn6502AssemblyKansasfest2014Presentation is a testcase for the updated code
06:04 🔗 SketchCow Build version: 0.183 (mame0183-356-g2110ca4)
06:05 🔗 SketchCow Boots.
06:06 🔗 SketchCow Still can tell it's MESS
06:06 🔗 SketchCow So, it gives some interesting console errors at the top:
06:07 🔗 SketchCow Driver is listed as compatible with nonexistent driver apple2
06:07 🔗 SketchCow Well, just that error.
06:08 🔗 DFJustin expected
06:08 🔗 DFJustin apple2e and apple2 are in separate driver files now
06:08 🔗 SketchCow I like it now just mentions emscripten in the boot
06:09 🔗 DFJustin oh you know what it is, it can tell it's MESS because we're using the old -sl6 diskii still
06:09 🔗 SketchCow Yep
06:09 🔗 SketchCow Is that still super slow
06:09 🔗 DFJustin probably
06:10 🔗 SketchCow I'll bring up other boot failures soon
06:10 🔗 SketchCow For now, doing a800 just to jack the booting up to current
06:10 🔗 DFJustin I'm heading to bed
06:11 🔗 DFJustin also, a thing
06:12 🔗 DFJustin mame has been moving to a more release-oriented development cycle http://mamedev.org/?p=437
06:13 🔗 SketchCow Compressing src/mame/layout/comp4.lay...
06:13 🔗 SketchCow make: *** build/projects/sdl/mamea800/gmake-asmjs: No such file or directory. Stop.
06:14 🔗 DFJustin hmm annoyingly that post doesn't tell you how to actually pull the tagged release version, maybe someone here can pipe in
06:16 🔗 bai SketchCow: remind me which args you're using to build specific systems like a800
06:17 🔗 bai I'm doing a full SUBTARGET=mess build right now, we'll see how big that ends up being
06:18 🔗 bai but once it's done I can crank out all the other systems, I just can never for the life of me remember which options I use to build one specific system, is it just the SOURCES= argument?
06:29 🔗 SketchCow #!/bin/sh
06:29 🔗 SketchCow # Builds the machine in the argument, having cleaned out the previous build.
06:29 🔗 SketchCow echo "----------------- CLEAN BUILD"
06:29 🔗 SketchCow ./emmake make -j4 SUBTARGET=$1 SOURCES=src/mame/drivers/$1.cpp clean
06:29 🔗 SketchCow echo "----------------- NEW BUILD"
06:29 🔗 SketchCow ./emmake make -j4 SUBTARGET=$1 NOWERROR=1 REGENIE=1 SOURCES=src/mame/drivers/$1.cpp
08:53 🔗 SketchCow http://teamarchive0.fnf.archive.org/PHANTOM/frames/screeny.gif
08:53 🔗 SketchCow That happened
10:44 🔗 yipdw has quit IRC (Read error: Operation timed out)
10:48 🔗 me has joined #jsmess
11:05 🔗 me has quit IRC (Read error: Operation timed out)
11:12 🔗 me has joined #jsmess
11:26 🔗 bai oh SketchCow, http://assets.metacade.com/emulators/
11:26 🔗 bai all updated
11:27 🔗 bai well all wasm updated, didn't do any asmjs or wasm-imprecise
11:53 🔗 phe has quit IRC (Quit: Leaving)
12:44 🔗 godane has quit IRC (Ping timeout: 633 seconds)
12:58 🔗 SketchCow That's fine, asmjs is dead to me
12:58 🔗 * SketchCow looks to moon
13:12 🔗 SketchCow Thanks, this helps a lot.
13:12 🔗 SketchCow What else needs to be contributed towards the build machine stuff
13:30 🔗 SketchCow Also, Paperboy blows up
13:36 🔗 phe has joined #jsmess
13:39 🔗 godane has joined #jsmess
17:08 🔗 db48x morning
17:09 🔗 db48x so, next for the build machine is to try it out
17:09 🔗 db48x to see if what I've written so far is working
17:12 🔗 db48x also, I feel like it needs a real name
18:55 🔗 bai SketchCow: yeah, the paperboy one is interesting, it might be a regression or I might have just compiled with the wrong systems
18:55 🔗 bai I think it's the right system because otherwise it would be a different error
19:02 🔗 bai LinkError: Import #3 module="env" function="_EEPROM_PARALLEL_2804" error: global import must be a number
19:02 🔗 bai I should try a native build of this version of mame with just the paperboy system, I wonder if it's just missing a driver
22:31 🔗 SketchCow db48x: I can make a name or someone else can have the honor
22:31 🔗 SketchCow Or we can be straightforward
22:37 🔗 SketchCow https://twitter.com/codinghorror/status/841775015917887488/photo/1
22:37 🔗 SketchCow So my guy who was helping with Phantom got it working, sort of
22:37 🔗 SketchCow And then said "You know, you REALLY REALLY REALLY should be doing chrome headless"
22:38 🔗 SketchCow So he's kind of doing that now
23:00 🔗 db48x` has joined #jsmess
23:00 🔗 db48x has quit IRC (Read error: Operation timed out)
23:59 🔗 phe has quit IRC (Quit: Leaving)

irclogger-viewer