#jsmess 2018-03-30,Fri

↑back Search

Time Nickname Message
00:06 🔗 oscar Or really, better yet, a way to properly access the FileFlag class.
00:30 🔗 oscar Well, I patched it over by hardcoding the number like we always do, but that's obviously non-ideal.
00:37 🔗 SketchCow OK, to my best of ability, I've split out the CDs for Playstation and those are all being uploaded. (It's, like, 300 more CD-ROMs and we're down to the D's as it goes in.)
00:38 🔗 SketchCow But these are all the ones, ironically, so popular that they ended up getting multiple versions and revisions and so on.
00:56 🔗 SketchCow db48x: Are you about
01:07 🔗 bwn has quit IRC (Read error: Operation timed out)
01:07 🔗 SketchCow bai: ted Vito` Ops please
01:07 🔗 SketchCow db48x: When you wake up - the compilation job upgraded the mamepsx to 0.196, but that needs another patch from DFJustin that makes it work right.
01:08 🔗 SketchCow So as a result the PSX stuff stopped working (which is fine, it's not announced and still uploading anyway)
01:08 🔗 SketchCow But here's the relevant line. I don't want to just step in and mess things up.
01:08 🔗 SketchCow 11:29 < DFJustin> git cherry-pick 50c373be6f730eb3d1dc5efa20952fbc208a1cc2
01:08 🔗 SketchCow That fixes main.lua to load properly.
01:11 🔗 bai sets mode: +o SketchCow
01:11 🔗 bwn has joined #jsmess
01:17 🔗 DFJustin 0.196 should be a major improvement on the spectrum collection
01:20 🔗 SketchCow That's the hope
01:20 🔗 SketchCow Seems it
01:20 🔗 SketchCow But yeah, PSX is dead right now until we re-compile
01:20 🔗 SketchCow But I really don't want to mess up db48x life
02:21 🔗 SketchCow Uploader is down to the N's.
02:23 🔗 bwn has quit IRC (Read error: Operation timed out)
02:32 🔗 bwn has joined #jsmess
04:14 🔗 bwn has quit IRC (Read error: Operation timed out)
04:34 🔗 bwn has joined #jsmess
04:45 🔗 SketchCow It's at the Z's.
04:45 🔗 SketchCow And no db48x
04:45 🔗 SketchCow sets mode: +oooo arkiver azakai balrog db48x
04:45 🔗 SketchCow sets mode: +oooo DFJustin godane hook54321 jvilk
04:45 🔗 SketchCow sets mode: +oo Lord_Nigh oscar
04:53 🔗 Lord_Nigh has quit IRC (Read error: Operation timed out)
04:53 🔗 SketchCow I can't seem to build it to put in a replacement.
05:01 🔗 Lord_Nigh has joined #jsmess
06:00 🔗 bwn has quit IRC (Read error: Operation timed out)
06:18 🔗 bwn has joined #jsmess
06:18 🔗 SketchCow OK, I've got to hit bed. If you get a chance, db48x - throw in the updated PSX with the fix
06:18 🔗 SketchCow <3
06:18 🔗 SketchCow So many PSX games!!
08:10 🔗 bwn has quit IRC (Read error: Operation timed out)
08:12 🔗 db48x odd
08:12 🔗 db48x I kept the change to main.lua that you're talking about
08:13 🔗 bwn has joined #jsmess
08:49 🔗 wdna has joined #jsmess
11:09 🔗 SketchCow So, I guess we need to track that down.
11:11 🔗 SketchCow https://archive.org/details/psx_dinocrisd as usual is the easiest one to check (12mb, fastish load)
11:11 🔗 SketchCow Absolutely crashes out.
11:12 🔗 SketchCow exception thrown: 15989528 mamepsx_wasm.js.gz:17437:4
11:12 🔗 SketchCow failed to asynchronously prepare wasm: 15989528 mamepsx_wasm.js.gz:2117:5
11:12 🔗 SketchCow 15989528 loader.js:1153:42
11:12 🔗 SketchCow 15989528 mamepsx_wasm.js.gz:17536:3
11:12 🔗 SketchCow Unhandled promise rejection abort(15989528) at jsStackTrace@https://cors.archive.org/cors/emularity_engine_v1/mamepsx_wasm.js.gz:1734:12
11:12 🔗 SketchCow stackTrace@https://cors.archive.org/cors/emularity_engine_v1/mamepsx_wasm.js.gz:1748:11
11:12 🔗 SketchCow abort@https://cors.archive.org/cors/emularity_engine_v1/mamepsx_wasm.js.gz:17544:43
11:12 🔗 SketchCow instantiateArrayBuffer/<@https://cors.archive.org/cors/emularity_engine_v1/mamepsx_wasm.js.gz:2118:5
11:12 🔗 SketchCow u@https://archive.org/includes/polyfill.min.js?v=62577d7:2:26885
11:12 🔗 SketchCow I/<@https://archive.org/includes/polyfill.min.js?v=62577d7:2:27007
11:12 🔗 SketchCow a@https://archive.org/includes/polyfill.min.js?v=62577d7:1:18228
11:13 🔗 SketchCow DFJustin: If you get a chance, can you see if anything steps out
11:18 🔗 SketchCow db48x: So, the new versions of things are not showing up at the archive - I see the zxspectrum is, for example, 0.195
11:30 🔗 SketchCow Also, I see the PSX emulator is also 0.195
11:30 🔗 Lord_Nigh has quit IRC (Read error: Operation timed out)
11:30 🔗 SketchCow So I think something's weird with the update
11:32 🔗 Lord_Nigh has joined #jsmess
11:49 🔗 db48x yea, I reuploaded the previous version of psx
12:09 🔗 db48x ok, I rebuilt psx and now it works
12:13 🔗 db48x as for the others, the upload didn't work, and they weren't replaced
12:26 🔗 godane has quit IRC (Ping timeout: 260 seconds)
12:38 🔗 godane has joined #jsmess
12:46 🔗 oscar Hey db48x: when you get a chance, is there a reason flag_w in loader.js doesn't have the files open as truncating? I assume it was an oversight, but wanted to make sure before I made an obscure, possibly breaking change.
12:48 🔗 oscar That cost me a day or so until I actually looked at the contents of the mame.lua that was being loaded and found that it was a jumble of previous versions because it was opening in overwrite mode.
14:46 🔗 SketchCow thank you very much
14:46 🔗 SketchCow I have a psx in my phone
15:33 🔗 godane has quit IRC (Quit: Leaving.)
15:33 🔗 godane has joined #jsmess
15:52 🔗 db48x oscar: no reason that I can think of; probably just an oversight
15:53 🔗 oscar OK, cool.
15:53 🔗 db48x I did have a report of something similar happening, but hadn't been able to track it down
15:54 🔗 oscar It's truncate now. Pity there doesn't seem to be a good way to get at the proper classes with our current method... perhaps that's something that can be updated in BrowserFS to at least export that interface.
15:54 🔗 oscar Anyway, I have the socket interface mostly working between JS and Lua now, so depending on workload/family stuff today, I should have something to demo tonight or tomorrow.
15:55 🔗 oscar I'd like to make it something that would also work locally on Lua, but there doesn't seem to be a Lua-native way to open a socket non-blocking, and the Lua backend expects not to block.
16:02 🔗 db48x yea, jvilk was horrified when I wrote that
17:35 🔗 oscar Haha, well, perhaps the enums for file flags should have been exposed to the outside. ;-)
17:39 🔗 db48x indeed :)
17:39 🔗 db48x I'm also misusing the API in other, less obvious ways
18:30 🔗 SketchCow I've got someone worming on PSX covers
18:30 🔗 SketchCow I will work woth dopefish to make the other region psx games and crossref them in
19:09 🔗 DFJustin it's pretty simple, psu = usa, psj = japan, pse = europe, psa = asia-pacific
19:31 🔗 SketchCow Well, THAT's easy, but I mean the json files.
19:31 🔗 SketchCow Like do we need special BIOS zips, etc.
19:36 🔗 DFJustin yes each one has a bios zip
20:02 🔗 Lord_Nigh has quit IRC (Ping timeout: 268 seconds)
20:06 🔗 Lord_Nigh has joined #jsmess
20:24 🔗 SketchCow OK, with luck I can do this tonight
20:34 🔗 SketchCow Duncan the Metadata Machine has added 1,505 mobygame links to the collection
20:34 🔗 SketchCow And has sent me 100 cover links, and I have a script that is doing all the right work
20:34 🔗 SketchCow So these covers are just BLASTING in
21:01 🔗 SketchCow db48x: While I'm here, there's no library/technology out there we can use for "download multiple chunks at once if available", is there
21:06 🔗 bai like multistream downloading of the same file?
21:21 🔗 SketchCow YEah
21:22 🔗 SketchCow I'm just trying to figure out if there's a way to speed up a download.
21:22 🔗 SketchCow I don't even know the name for the method we do a download with now, so I can't look up "multi-threaded version of _______"
21:25 🔗 bai well...if everything's operating properly, network-wise, it shouldn't be a problem to max out your connection on a single stream
21:25 🔗 bai I have noticed that streams coming out of IA tend to cap around 1 MB/s each, though
21:26 🔗 bai I know I can pull 6 MB/s on a single stream pretty easily on this connection, and I know IA has a nice fat connection, which is why I was wondering if there's any intentional QoS stuff going on on that end
21:28 🔗 bai the 1MB/s thing from IA has been consistent for me from a number of different networks, too
21:28 🔗 bai at least, over the past few days when I was really looking at it
21:33 🔗 DFJustin yeah my connection can do 20 MB/s and I'm not getting even close to that from IA
21:33 🔗 SketchCow It's likely a switch
21:33 🔗 SketchCow I have something from Jonah about it.
21:35 🔗 SketchCow After he does a switch replacement soon, we can find where it is.
21:38 🔗 i0npulse has quit IRC (Ping timeout: 252 seconds)
21:38 🔗 Lord_Nigh has quit IRC (Ping timeout: 252 seconds)
21:38 🔗 Rai-chan has quit IRC (Ping timeout: 252 seconds)
21:38 🔗 i0npulse has joined #jsmess
21:40 🔗 Rai-chan has joined #jsmess
21:42 🔗 Lord_Nigh has joined #jsmess
22:09 🔗 Vito` mm
22:09 🔗 Vito` can you web torrent from the IA web seeds?
22:12 🔗 bai ipfs :D
22:40 🔗 bwn has quit IRC (Ping timeout: 260 seconds)
22:45 🔗 bwn has joined #jsmess
23:00 🔗 Vito` I didn't think ipfs was supported in browser-only JS
23:00 🔗 bai there is js-ipfs but I haven't actually looked at it in a while
23:01 🔗 Vito` but if you make IA's emularity-related torrents discoverable over the webtorrent webrtc network or whatever, you should be able to download from their webseeds
23:01 🔗 Vito` it's a different tracker but I think normal webseeds work
23:01 🔗 bai yeah, maybe that would be a cool module for BrowserFS
23:01 🔗 Vito` or a different swarm? I don't know the terms any more
23:09 🔗 Vito` oh hm
23:10 🔗 Vito` it looks like webtorrent can just take a normal torrent file that specifies webseeds and figure it out
23:10 🔗 Vito` so maybe you wouldn't need to do anything at all
23:42 🔗 Lord_Nigh has quit IRC (Read error: Operation timed out)
23:42 🔗 Lord_Nigh has joined #jsmess
23:44 🔗 godane has quit IRC (Ping timeout: 633 seconds)
23:58 🔗 Vito` okay
23:59 🔗 Vito` the torrent file won't download in the browser without using the CORS proxy URL
23:59 🔗 Vito` but when it does, it seems to parse it and work
23:59 🔗 Vito` except then all the files it tries to piece together fail because of CORS

irclogger-viewer