On Monday, 28 July 2014 at 19:16:24 UTC, Ola Fosheim Grøstad
wrote:
Why should the situation be different if I use the builtin
`assert` instead?
No difference. Except with a builtin you can get additional
debugger support, so that you don't have to recompile when
following a complicated trace that trigger asserts.
Well, actually, that is not correct. There are other advantages
too.
If the compiler guarantees that asserts don't have side effects
(on optimization) and aren't affected by the optimizer then you
are more likely to see the same behaviour in debug and release
builds even with heavy optimization turned on (such as various
math approximations). E.g. you can verify with exact math in the
assert() that the fast-math approximations the optimizer makes on
running code are within the acceptable tolerance level.
