[00:44] good idea with the showcase [00:49] added support for the caps-api. now it does want capsimg.dll :) [04:27] n8 [04:27] *** naTmeg has quit IRC (Leaving) [08:31] Hey hey hey [08:32] So, going to wander around London, then take a 5 hour plane ride, then home, then I'm more accessible. [08:44] db48x: So, we have a "thing" happening. [08:45] If you look at screenshots, a pile of "failed to load game data" errors. [08:46] I can't tell if one or more screenshotters are doing this now, so I'm going to let them keep screenshotting, and then I can do an algorithmic removal of the fails (they all have orange in the same place), crush and rebuild the boxes, and start again. [08:47] But something goes south in firefox and then the things stop working. That's obvious. [08:47] It happened to my usual screenshotter, it's happening here. [08:54] We'll figure it out, I'm sure. [09:13] did a check, they seemed fine... weird [11:33] free wifi at heathrow, might be able to write cleaner now [11:33] so much beauty [13:44] *** naTmeg has joined #jsmess [15:25] OK, running the script [15:25] I wrote it! [15:25] It's cleaning [15:30] (That'll fix the problem.) [15:32] Well, you know, stop the thing looking like an orange nightmae. [16:14] *** Coderjoe has quit IRC (Read error: Connection reset by peer) [16:16] I found the clunker. [16:16] (The one making orange delights) [16:18] hrm [16:19] what's the difference between the broken one and the ones that work? [16:20] *** Coderjoe has joined #jsmess [16:26] I terminated it. [16:26] (SOrry) [16:26] ah [16:26] I suspect that if I had to guess, it's going to be either the cache or the cookies [16:26] Since these were all exactly the same when started. [16:27] Firefox is 100% killed between screengrabs [16:27] Cache and cookies are all that survive. [16:28] If it happens again, I will save off all the crap. [16:30] Good news is I can make another screenshotter in about 3 minutes. [16:31] apt-get update [16:31] for each in firefox imagemagick Xvfb python3-pip xdotool fdupes unzip; do apt-get install $each;done [16:31] pip3 install internetarchive [16:31] scp jscott@fos.textfiles.com:screenshotter-aws.zip . [16:31] mkdir /SCREENSHOT [16:31] ia configure [16:31] cd /SCREENSHOT [16:31] unzip ~/screenshotter-aws.zip [16:31] Xvfb :1 -screen 0 1280x1280x24 & [16:31] Boom baby [16:31] :) [16:33] aaaand it's up [16:34] aaand it's screenshotting [16:34] Hopefully that won't happen again, but maybe [16:35] Also, the metadata split worked great! Except where it didn't. [16:35] But I can easily find those. [16:39] The Orange-B-Gone(tm) Deleter is down to the Crs. [17:20] The thing is now fixing the metadata that got weird (because of the wrong parenthesis) [18:00] Hackernews... you never disappoint [18:21] :) [18:26] Metafilter also hopped in I see [18:26] Definitely something with firefoxx, another person reporting it [18:28] Also the internet says it's not perfect [18:28] * SketchCow starts burning [18:28] heh [18:31] Oh no db48x you're responding to them [18:31] They're going to get stuff on you [18:32] well [18:33] I'm not going to add more stuff quite yet, but I probably have a few thousand more to add. [18:35] Mostly because we will empty my bank account on my screenshotting [18:35] Best reason [18:35] :) [18:36] * SketchCow holds up sign "SCREENSHOTTED TOO MUCH AMIGA - ANYTHING WILL HELP" [18:36] hmm [18:36] naTmeg: Not surprisingly the masses have... ideas. [18:36] SketchCow: will IA refund you? [18:37] maybe [18:37] If I ask [18:40] I thought IA was spinning up its own EC2-equivalent [18:41] oh maybe archive labs was what I was thinking of and that just looks like traditional hosted projects [18:43] Yeah [18:43] I also don't deal with archive labs [18:43] This is pure amazon [19:28] SketchCow: for the mute button to start working, more than just the loader must change [19:29] we'll have to impose on Tracey or someone to update the theater code so that it actually calls the loader's mute/unmute functions [19:30] in the mean time, could you update the emulator from http://erebor.db48x.net:8000/emulators/sae/scriptedamigaemulator.js? [19:32] back in a bit [19:48] Updated [19:53] I must say, people are pretty hilarious. [19:53] Just by the screenshots, I can see thousands are working. [20:06] OK, date and creator metadata is transferred over. [20:06] Now fixing the titles to rip out the elsewhere metadata. [20:06] So it'll look less computery [20:55] ok, we can mute SAE correctly now [20:56] by typing AJS.emulator.mute() or AJS.emulator.unmute() in the console [20:57] *** n00b125 has joined #jsmess [21:01] SketchCow: oops, you updated scriptedamigaemulator.js, but not the source map (scriptedamigaemulator.js.map), and perhaps not the source (the sae directory) either [21:08] I need a routine to do this. [21:08] And super bad timing... I need to land [21:08] But give me a routine and I'll happily do it [21:08] I can also give you access to the thing as well. [21:09] that would work for me [21:11] is it db48x and net [21:11] FOr the account [21:12] db48x@db48x.net, yes [21:12] db48x@db48x.net now has the following privs: [21:12] emulatorlab [21:12] emularity_engine_v1 [21:13] Log out and log back in and you own it [21:15] hmm, still doesn't let me edit [21:16] Be sure to log out completely. [21:16] I'll work with you if it doesn't work, tonight [21:20] ok, I can edit the emulatorlab collection, but not the emularity_engine_v1 item [21:56] naTmeg: ping? [21:56] *** n00b125 has quit IRC (Ping timeout: 268 seconds) [22:02] * naTmeg is busy working on the caps-support :p [22:02] :) [22:11] if you have a minute, I've tried calling sae.start() followed immediately by calling sae.mute(true), but it doesn't actually mute [22:17] lol @ https://archive.org/details/100_Printer_Drivers_19xx_Anglia_PD_PD [22:20] SketchCow: some suggestions for the showcase: https://archive.org/details/Liberator_1992-03_Scorpius, https://archive.org/details/Titan_1989_Titus, https://archive.org/details/Minos_1989_Turtle_Byte_en-de, https://archive.org/details/Nicky_Boom_1992_Microids [22:32] db48x: how do you know it does not mute? was there audio-output? [22:44] yea, I was testing with https://archive.org/details/Killing_Game_Show_The_demo-playable_1990_Psygnosis_cr_Conqueror_Zike [22:44] go to https://archive.org/details/Killing_Game_Show_The_demo-playable_1990_Psygnosis_cr_Conqueror_Zike&external_js=1, then before you click Play, type "AJS.emulator.mute()" in the js console, then click Play [22:46] (AJS.emulator.mute() is what the mute button would do if it weren't somewhat broken at the moment) [22:47] w8 [22:48] SketchCow: https://archive.org/details/Amazing_Intros_19xx_17-Bit_Software is interesting [22:52] i think because if sae is not running, an error is retured. see amiga.js:590 [22:55] potentially, but I did just call start() [22:56] oh ofcos.. [23:03] loader.js:668 "var runner;" should be "var runner = true;"? cos i got the SDL-part called here.. [23:04] yea, that's a side effect [23:04] runner is still null, so it falls back to the Emscripten way of doing things [23:04] but it sets muted = true [23:05] then on line 950 it calls runner.mute() if muted is true [23:06] i get: Unable to change audio state: TypeError: Cannot read property 'cwrap' of null(…) for line 685 [23:07] yea [23:07] it's completely expected for that branch to fail in this case [23:07] it gets caught, we log it, and things continue [23:10] oh ofcos. maybe should not give suggestions while just overfly code [23:19] line 948: you call runner.start(); and immediately after runner.mute(); but start() does take some time and it is not running until cfg.hooks.running() get called. so the call to mute will fail [23:21] splash.finished_loading = true; should be set in the hook [23:22] runner.mute() and splash.hide(); too [23:23] oops, i meant cfg.hooks.started() ofcos [23:40] I don't suppose you have a list of the hooks? [23:41] just see example.js:275 [23:41] in the root of sae.net [23:42] that has cfg.hook.event.started [23:46] the hook was never called... [23:47] yes, you need to "this._cfg.hooks.started = function() { }" [23:48] i think at least :) [23:49] not this._cfg.hook.event.started = func? [23:50] typeof this._cfg.hooks.started must be function [23:50] yes, func is a function [23:50] then it should be ok [23:51] does not get called? [23:51] nope [23:51] reload the page to get the updated script [23:56] *** n00b125 has joined #jsmess [23:57] mom