[01:36] *** azakai has joined #jsmess [04:47] *** azakai has quit IRC (Ping timeout: 255 seconds) [05:47] *** taisel has joined #jsmess [05:59] *** taisel has quit IRC (Quit: ok going for the next half year. \o/) [08:10] *** godane has quit IRC (Read error: Operation timed out) [08:16] *** godane has joined #jsmess [15:30] *** azakai has joined #jsmess [18:20] *** taisel has joined #jsmess [18:23] so apparently my theory on MS Edge having a threading/buffering bug with scriptprocessornode seems to be on to something. [18:23] cpu load on document thread at barely 2/3rds load causes audio stutter [18:23] offload entire codebase to webworker = no stutter [18:24] should only occur if cpu load is at 100% or there's thread blockage longer than buffer duration [18:24] webkit used to have this bug, until they rewrote their threading in webkit, as per some webkit blog article [18:25] firefox has its own bug, but it's entirely different, it just seems to *not* skip javascriptprocessor events when it should [18:26] implementation details are tiresome [18:28] javascriptprocessornode is officially a dumpster fire of implementation differences and bugs. [18:28] *** azakai has quit IRC (Quit: Ex-Chat) [18:35] isn't that the one that's technically deprecated but no one's implemented the new node that replaces it? [18:35] yeah [18:35] lovely :D [18:36] because its replacement is YOU scheduling buffers on your own, still afflicted by implementation details on how to handle buffers that start and end next to each other [18:36] rather than something in its nature supposed to be gapless [18:36] it turns into an implementation detail [18:37] welcome to hell [18:52] *** taisel has quit IRC (Quit: ChatZilla 0.9.93 [Firefox 55.0.3/20170824053622]) [19:42] *** azakai has joined #jsmess [23:03] https://team-europe.blogspot.com/2017/09/100-years-dumped-and-preserved.html [23:18] *** godane has quit IRC (Quit: Leaving.) [23:34] *** godane has joined #jsmess