3 ms·
But: they are holding it wrong and it got fixed regardless because if you can fix such stuff you should. For Rust such a CVE is nearly free given the installed
by jacquesm 3y ago
But: they are holding it wrong and it got fixed regardless because if you can fix such stuff you should. For Rust such a CVE is nearly free given the installed base of Rust and the fact that most of that code is fresh and in maintenance. But that definitely isn't true for all of the instances of glibc that are currently deployed.
- megous 3y agoWell, a fix in glibc fixed it across the whole system and for all already compiled programs. Can you say the same for Rust?
- nequo 3y agoDynamic linking is something that I would personally like to see in Rustland too. But parent's point is that some issues that count as a CVE for Rust do not count as such for C/C++ because the boundaries of undefined behavior are drawn elsewhere. This means that there are many fixes that are simply never made for C/C++, even though dynamic linking would make such a counterfactual fix easier to deploy.
- megous 3y agoParent point was that such CVE could be fixed in Rust super cheaply (free) because code is maintained. If I have to reinstall the whole system to fix CVE in Rust stdlib in a brave new world where everything is written in Rust, then it's certainly not free fix.