3 ms·
I tend to agree with his point that you may perhaps not solve design and driver problems just changing language, but can anyone dispute that a memory-safe langu
by Santosh83 9y ago
I tend to agree with his point that you may perhaps not solve design and driver problems just changing language, but can anyone dispute that a memory-safe language can at least help eliminate routine programming pitfalls like off-by-one, null deferences, double-free and so on? And these are the types of innocuous and inadvertent errors that are exploited by black-hats, sometimes with costly consequences.
I don't see a strong argument against a memory-safe language, although granted it won't solve architectural and hardware issues. Perhaps the effort to support, say Rust, in the kernel is just not worth the trade-off at this point? Perhaps no one has taken on the mantle for doing so? But I can't see how one can categorically argue against it...