Conversation

Patrick Walton
@pcwalton
Maybe the biggest unanswered question about Rust is how much the borrow checking rules can help performance by improving compiler alias information. (There is some use of noalias but it's not very aggressive yet.) If they do by a lot, that changes the calculus for using Rust...
21
284
Patrick Walton
@pcwalton
Like, if the borrow check helps runtime performance, then why wouldn't you use Rust over alternatives? It'd be safer *and* faster.
5
98
CheeseLip
@lip_cheese
Replying to
@pcwalton
Have a newbie write a linked list in rust to see the problem. I'm a fan, but you really have to understand things before you jump in. Obviously we all WANT devs to know their tools before writing a line of code, but other languages can be a bit more forgiving while folks learn.
2
1
CheeseLip
@lip_cheese
Replying to
@lodurel
and
@pcwalton
It's hot garbage - I'm not defending C. That rust is a complex and unforgiving language is both a strength (it's super powerful) and a weakness (harder to ramp up). Haskell is in the same boat.
2
Daniel Micay
@DanielMicay
Replying to
@lip_cheese
@lodurel
and
@pcwalton
Rust is not complex and unforgiving, especially compared to C. It's dramatically easier for someone with no experience with either to learn Rust. Reading comments on Hacker News threads about Rust or Haskell for that matter doesn't mean you have an informed opinion about them.
1
CheeseLip
@lip_cheese
Replying to
@DanielMicay
@lodurel
and
@pcwalton
So would years of experience writing programs in these languages allow me to have an informed opinion? If you insist that rust is simple and forgiving, you are doing a disservice to the community. It'd be like me insisting that it's easy to write safe and secure programs in C.
1