Second try at OsRng documentation

src/os.rs

@@ -19,10 +19,17 @@ use rand_core::{RngCore, Error, impls};
 /// most applications. Commonly it is used to initialize a user-space RNG, which
 /// can then be used with much lower overhead.
 ///
-/// Sometimes there is the misconception that `OsRng` somehow provides 'better'
-/// random numbers than some good user-space cryptographic random number
-/// generator. This is not true, but `OsRng` has a role in initializing such an
-/// RNG. The initial seed of randomness has to come from somewhere.
+/// Generally you can say that the values generated by `OsRng` are not in some
+/// way 'better' than those generated by a good user-space cryptographic random
+/// number generator. It is not necessary to use `OsRng` if your only
+/// requirements are 'good' random numbers.
+///
+/// Where `OsRng` especially has a role is in seeding user-space RNGs. The
+/// initial seed of randomness has to come from somewhere. And it can also be a
+/// good choice if you need a small number of secure random values. We can
+/// expect the RNG of the operating system to have mitigations in place for
+/// vulnerabilities such as side-channel attacks. But in general it is best rely
+/// on security-oriented libraries in such cases, such as OpenSSL or Ring.
 ///
 /// You may prefer to use [`EntropyRng`] instead of `OsRng`. Is is unlikely, but
 /// not entirely theoretical, for `OsRng` to fail. In such cases `EntropyRng`