[l-albertovich]

I felt the same way when I read the bold part that says "But that C codebase is an issue" so I quickly checked out the public databases and couldn't find a single serious vulnerability in the past 7 years.

Admittedly I stopped after going through a bunch of useless stuff related to CVE-2017-8823 (which was initially reported as remotely exploitable with no proof at all).

I went through the tor repository (not vidalia though) and read a bunch of conversations about some of the memory related bugs but none of those were exploitable either (exploitable as in remote execution, not a DoS) and most of the (not so many) bugs were actually logical bugs.

I really don't care what they decide to do with their project and honestly anything that can potentially improve the security of such a system is fine by me but I really think they're doing themselves and the language a disservice by communicating the way they do.

Also, as a side note, even with a C codebase there is SO MUCH you could (and should) do to minimize the impact of a vulnerability that the fact that some choose to present just rewriting code in a different language is not even funny.

[uecker]

And of course, "impossible to refactor" just is very deep in the bullshit territory. "more fun to write new code" would probably be more honest, and the Rust proponents created a marketing narrative that allows them to do this while pretending (and probably also believing themself) to do a good thing.