[00:59] <Voldenet> I don't mean that bugs should be replicated, I mean that optimizer could be generating very inlineable code which happens to trip the spesh

[01:00] <Voldenet> and with the same optimization, spesh could act the same

[01:08] <Voldenet> the bug seems to be that some inlined attribute access isn't guarded, becasue self.^name is properly set to dynvar if you use it

[01:46] <Voldenet> (i didn't actually read spesh.log, it's just how it behaves)

[02:03] *** japhb left
[02:20] *** japhb joined
[04:07] *** vrurg_ left