academics are fundamentally extremely conservative, and any work that doesn't fit the current party line faces a major uphill battle to get accepted
-
-
Replying to @johnregehr @pcwalton and
But the borrow checker, if nothing else, is a huge advance? Surely that should break through? He said, naively
1 reply 0 retweets 2 likes -
Replying to @fugueish @johnregehr and
It was seen as too similar to existing work
2 replies 0 retweets 5 likes -
I think the Rust pitch is made more complicated by Rust's lack of ambition relative to prior work. Rust basically says "don't try to make it possible to write a GC in fully safe code, instead make a usable language" but that's harder to describe.
2 replies 0 retweets 7 likes -
The interesting stuff in Rust is all of the stuff necessary to make substructural type systems practical…
1 reply 0 retweets 2 likes -
e.g. Rust tries to solve the dichotomy in Grossman’s “Existential Types for Imperative Languages” by tracking ownership and enforcing uniqueness of mutable borrows per use site, which is the only way to make a usable language IMO.
1 reply 0 retweets 9 likes -
In that case, the novelty is Rust has its cake and eats it too. Is that enough for novelty? I think so, because the complications that this creates are interesting. But reasonable people (and reviewers) can disagree.
1 reply 0 retweets 1 like -
I think that's exactly it, and it is of course enough for novelty, but it's a trickier pitch than "we strictly improve on prior work" even though it's more important.
1 reply 0 retweets 3 likes -
I should clarify: It’s the only way to make a usable language *that doesn’t require GC*. Cyclone was usable because it fell back on GC for a ton of stuff.
1 reply 0 retweets 6 likes -
Also true of Rust for the first few years
1 reply 0 retweets 0 likes
Nobody was trying to publish anything on pre-borrow-checker Rust.
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.