3 ms·

Assuming that tests exist, and have good coverage, wont the checked iterator implementations mentioned above catch most issues with invalid iterators?

• dikaiosune 10y ago
  A big part of what makes Rust great is that it catches those issues at compile-time, no test coverage required, no extra run-time checks required (not that it moots tests, of course). In my experience so far this is a very powerful ally to have when writing software, as it's very useful to surface those issues as early as possible. EDIT: Also: > Assuming that tests exist, and have good coverage That's a very big assumption. Granted my current job is often to clean up codebases that are the opposite of that, so maybe I've got a bias here.
• pcwalton 10y ago
  I don't know of any large, widely used, security critical, heavily scrutinized C++ program that uses those flags in production. I also don't know of any such C++ program that hasn't had game-over use after free bugs. So the answer to your question is somewhere ranging from "we don't know, but unlikely" to "no".
• Someone 10y ago
  For some definition of ‘good coverage’ and ‘most’: yes. I do not think 100% coverage on the code you write alone is enough to get to ‘all', though. For example, if your test declares a vector and calls a function that does v.erase( i, v.end()); with i == v.end(), you get 100% coverage, but in production, that part of the code might be called with i == v.begin(), and that invalidates all iterators on the vector.