03:33 | <ljharb> | psyched for a nonfictional plenary |
16:55 | <bakkot> | could we add a blocking version of await, only available off the main thread |
16:55 | <bakkot> | we already have atomics.wait |
16:55 | <bakkot> | is this a terrible idea |
16:57 | <Ashley Claymore> | the win being not worrying about red/blue functions? Or could there also be performance wins to blocking over using await mechanics ? |
17:00 | <shu> | bakkot: probably a terrible idea? |
17:01 | <shu> | what are the proposed semantics? remember Atomics.wait only gets woken up by another thread or timeout |
17:01 | <shu> | i don't think await s are like that |
17:05 | <legendecas> | IIUC we can not resolve a promise off the thread, so blocking version of await sounds like a deadlock. |
17:06 | <shu> | it doesn't have to be, but you'd have to invent a radically different set of event loop semantics |
17:06 | <shu> | which makes me think it is probably an infeasible idea |
22:58 | <Richard Gibson> | GetSuperBase makes a supposedly fallible call MakeMethod($f, object) sets $f.[[HomeObject]] to object and is called from only
MakeMethod($f, homeObject) sets $f.[[HomeObject]] to object and is called from only
DefineMethod is called with argument object from only
DefineMethod is called with argument proto from only
MethodDefinitionEvaluation is called with argument object from only
ClassFieldDefinitionEvaluation is called with argument object from only
ClassStaticBlockDefinitionEvaluation is called with argument object from only
ClassElementEvaluation is called with argument proto from only
ClassElementEvaluation is called with argument F from only
PropertyDefinitionEvaluation is called with argument obj from only
|
23:04 | <bakkot> | Richard Gibson: your analysis sounds right to me |
23:05 | <bakkot> | I strongly suspect there's a number of remaining cases where there's a ! which should be a ? |
23:05 | <bakkot> | or, I should say the non-throwy-ness sounds right; I haven't looked for how it might be undefined |