interesting! Any up to date notes?
-
-
Replying to @Gankra_
so that nonthrowing functions can be physical subtypes of throwing ones. Caller that expects an error zeros before call
3 replies 0 retweets 0 likes -
Replying to @jckarter
As you're no doubt aware from Rust's struggles, implicit unwinding also has massive systemic cost on language semantics
1 reply 0 retweets 0 likes -
Replying to @jckarter
the API and ABI concerns seem orthogonal in this regard -- all 4 combos of implicit|explicit X unwind|return work.
3 replies 0 retweets 0 likes -
Rust's struggles aren't unwinding per se, it's the fact that unwinding is an implicit thing everyone can do.
1 reply 0 retweets 0 likes -
I won't be surprised to find nounwind annotations finding their way into Rust to compensate.
1 reply 0 retweets 0 likes -
sadly panic=abort compilation strategy changing what code is *allowed* to compile is probably not gonna happen...
1 reply 0 retweets 0 likes -
why not? specialisation/#[cfg] for panic=... has been discussed, so that unwinding is actually zero-cost
1 reply 0 retweets 0 likes -
painful to write things 2x but eased by eg fn read_or_swap(&mut T, T) -> T that only read with abort
1 reply 0 retweets 0 likes -
mostly politics -- fans of unwinding ("embedders" like Servo and Ruby gems) want unwinding to be first class.
2 replies 0 retweets 0 likes
It’s not “politics” and we’re not “fans”.
-
-
didn't intend it in a negative way. You have legit usecase and don't want the community to fragment.
0 replies 0 retweets 1 likeThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.