Hacker News new | ask | show | jobs
by arinlen 1421 days ago
> Your point is a fair one, but defaults matter.

This sort of comment is far from fair and misses the whole point.

The thesis of this article is how programming languages compare with regards to safety.

It makes no sense at all to compare particular implementations and try to pass personal assertions on particular features of said implementations as broad assertions about the programming languages they support.

So a particular Rust implementation is bundled with a linter. That's pretty convenient.

It just so happens that the linter is provided by a compiler stack that is also one of the main reference implementations of C++, and said linter also supports C++.

Is it fair to thus claim Rust is somehow superior to C++ just because a particular implementation enables by default it's linter on Rust code but not C++?

> There's little reason for clang-tidy not to be part of the default clang invocation by now

It's perfectly fine that anyone forms their own personal opinion on what defaults a tool should ship with.

That says anything nothing about the programming languages and their safety though.

1 comments

As is clear from my other comments I also find this stuff unfair.

I try to keep in mind that on balance it’s a good thing that performant programs have become dramatically more accessible recently, and that most of the C/C++/Fortran/Haskell antagonism is a result of enthusiasm around that. For me it was BASIC -> C, but I imagine JS -> Rust is every bit as exhilarating.

But I do hope at least a few people read your comment and are inspired to learn a little background. Rust is cool because it remixed the broadly-accessible FP algebra, a kickass C++ toolchain decades in the making, and a big bet on linear typing.

I didn’t set out to be the jaded, un-hip old guy but here we are :) It’s a nifty new LLVM front end with type classes and the Either monad. Neat. Get off my lawn ;)