3 ms·
Partly true, but this doesn't apply to many rewritten tools, like cat or ls where memory safety bugs are tolerable
by kangalioo 5y ago
Partly true, but this doesn't apply to many rewritten tools, like cat or ls where memory safety bugs are tolerable
- ansible 5y agoWell, I don't know about the cat command (though that _is_ appropriate for something like the Redox OS, which aspires to be a pure-Rust desktop operating system). But there is a broad-based desire to rewrite libraries in Rust for a couple (IMHO valid) reasons. One is the API. While it is possible to have a nice and idiomatic Rust API to a C library, it will often be nicer to use when it is designed from the ground-up in Rust. Traits, interfaces, etc. It is also often desirable to avoid C when cross-compiling (such as for embedded systems), as that is generally a hassle to set up and get working seamlessly.
- detaro 5y agoDoesn't sound like things where you'd want bugs either. But also the rewrites for this kind of thing are usually not that "serious". A few people do it because they can or because they think it's interesting, but it's not like there are large pushes to replace such tools (E.g. there's been "lets rewrite basic tools in Go" projects too, and nobody is surprised that they aren't in widespread use"). The winners usually are not in that category, but tools that stand alone.
- kaba0 5y agoI see no evil here. For many it is probably just a hobby project they can practice their rust skills on. Also, it’s not like rewriting such trivial tools as cat in rust would cause any conceivable amount of useless churn. They are not rewriting complete ecosystems.