Hacker News Reader

I think just like other languages, writing Rust leads to a familiarity with what the compiler wants. Then you spend more time thinking about design and the problem at hand, and less fighting the compiler. I think Rust will ultimately be easier than C++ to write because while there are tricky concepts to master, there should be much fewer of them than C++ has.

View on HN

I disagree. While you get familiar with what the compiler wants (certainly I did, after using Rust for more than 2 years now), it's just that Rust compiler wants more from you than C++ compiler does.

The number of tricky concepts largely do not affect writability, because you don't have to use them. The number of mandatory compile time checks do, because you have to pass them.

One way to improve writability is a mode to turn off some of checks while you are exploring. I already gave a good example from Haskell. When you are done exploring, you can turn on all the checks.

Hmm, so like a Rust flag like --allow-leaks or --corrupted-ownership?

It seems like editor support would totally dominate here. That is, realtime checking of lifetimes and autocomplete/inline errors should be a much bigger improvement in write times than some way to emit invalid code.

Reply on news.ycombinator.com