That's the policy, but we've been a bit more conservative in practice. `main` currently targets 1.88, but that's only because a security issue in the time crate has forced our hand (one reason I don't like the time crate all that much). Before that, it was 1.83 (from November 2024). Our last release targets 1.71 (from July 2023).
I think the way Rust checks borrows also makes it a lot more feasible to avoid allocations/copies; not because it is impossible to do in C, but because doing it in C requires writing very careful documentation and the caller to actually read that documentation. In (safe) Rust this is all checked by the compiler such that libraries can leverage it without blowing their complexity budget.
There’s also a compounding effect: I’ve heard from a hardware vendor that they spend a lot of time optimizing OpenSSL to get the most out of their silicon, so for their customers they suggest using OpenSSL to get the most out of the hardware.
To quote one of the lead GTK/GNOME developers, "what makes you think font sharpness is a metric (of font rendering quality)"? It's absolute madness: https://gitlab.gnome.org/GNOME/gtk/-/issues/3787.
It's always interesting how these large organizations can bring in tens of millions of dollars in excess of expenses, yet still manage to "have no money"
reply