Hacker News new | ask | show | jobs
by matis140 2773 days ago
On the upside I was pretty sure this was going to be a Rust to the rescue opinion piece before I clicked on it. On the downside this is a Rust to the rescue opinion piece though it does also mention swift for memory safety so that is nice. Nothing against the language, but it has been around a few years and I would like to see something more concrete than opinions. Give me a write up about a vulnerability in the wild fixed with the new language.

Something new at any rate I see plenty of opinion piece espousing this or that new language PR reading points, but most fall into fluff.

1 comments

But it's a bit difficult to prove, isn't it? I mean, you can just say "We didn't have a buffer overflow error in our software written in Rust", but that's basically it. You can't fix this kind of a bug in a project without rewriting it, and as the author says, it's hard for several reasons. People simply don't have enough incentives to do it.
Not entirely and I know Mozilla is in the process of doing this now. They could take some small portion of the original bit of the browser they are rewriting and put something technically interesting together instead of more opinions like this. One of the things they have to overcome has been stated above and that is that c and c++ have a lot of inertia. They are also still cool languages. This will take a lot more than opinion to overcome.

I have written a few small things in rust and explored crates.io and like the language and ecosystem. I am much less certain about the community surrounding the language, but it might just be that they are mostly younger people looking for safe places so they often feel unauthtically welcoming. It is a bit off putting.

Anyway I would just like to see something like here is a unit of the browser we replaces with rust servo. It took us 2 week to do this unit. before the rewrite this unit would crash or leak at some rate after the rewrite these have been reduced x amount where these would be numbers from the wild based off their metrics. They could Also throw in a couple bench marks over those code units for good measure, but i am certain Mozilla captures enough to probably put together something with interesting real world numbers.

AFAIK, they are not rewriting the browser in Rust. They are instead writing new components in it, so it wouldn't be an apples-to-apples comparison.