Re: [IMP-dev] NBL cleanup and other small patches

[Daniel Russel <>]

> The only thing we can say for sure is that when NDEBUG is defined, 
> assert() is a noop. IMHO, IMP_assert() acts like assert(), so it 
> should do the same thing - that's how I put it in the patch I 
> committed. For the other runtime checks, it seems reasonable for it to 
> be like CHEAP or NONE, as you say.
I don't think we need to provide that guarantee (perhaps we should 
rename the macros to IMP_internal_correctness_check since they are 
checks which should not fail if there are no bugs in IMP).

Part of my goal was to get rid of having two different types of 
compilation which we need to pay attention to.