Why the fuck does everybody claim that rust was initiated by mozilla?
It was a hobby project of graydon hoare. Yeah he worked at mozilla at one point, but this wasn’t a mozilla project…
Of course Graydon should be credited, but 2010 was also an important moment for the language, because that’s when it graduated from being a hobby to being the Servo Project. The Rust language has evolved a lot since then, and the borrow-checking GC-free Rust that we know today has been developed at Mozilla.
I’d be very surprised if Tor (as in the project) plans to maintain small-t tor once arti is feature complete. Using C for security relevant network applications in 2022 … is questionable.
Zig is a major improvement in language quality and ergonomics, but in terms of safety I don’t think Zig is above safety of modern C++ with sanitizers and a hardened runtime. If modern C++ doesn’t get a pass, Zig shouldn’t either.
Reading through the announcement and repo it sounds more like they intend to completely replace their current implementation. If not for “RIIR”, then simply because their old implementation got really hard to maintain and is the opposite of modular or embeddable. Might as well call it a complete rewrite which also switches the language.
Why the fuck does everybody claim that rust was initiated by mozilla? It was a hobby project of graydon hoare. Yeah he worked at mozilla at one point, but this wasn’t a mozilla project…
Of course Graydon should be credited, but 2010 was also an important moment for the language, because that’s when it graduated from being a hobby to being the Servo Project. The Rust language has evolved a lot since then, and the borrow-checking GC-free Rust that we know today has been developed at Mozilla.
Not the best phrasing, but “at” and “by” are different words, and it did receive a lot of support from Mozilla down the line.
While I dislike Rust, it’s always great to have a diversity of implementations for a given protocol/format.
I’d be very surprised if Tor (as in the project) plans to maintain small-t tor once arti is feature complete. Using C for security relevant network applications in 2022 … is questionable.
Indeed, this is the plan; we’re looking to slowly move away from doing work on C-tor, preferring instead to divert resources to arti going forward.
(source: I’m on the arti team)
Using C/C++/Zig/etc on any project for anything but historical reasons is questionable
I’m curious, why do you include Zig?
Zig is a major improvement in language quality and ergonomics, but in terms of safety I don’t think Zig is above safety of modern C++ with sanitizers and a hardened runtime. If modern C++ doesn’t get a pass, Zig shouldn’t either.
Reading through the announcement and repo it sounds more like they intend to completely replace their current implementation. If not for “RIIR”, then simply because their old implementation got really hard to maintain and is the opposite of modular or embeddable. Might as well call it a complete rewrite which also switches the language.