Morten Welinder wrote: >>>This is a joke, you are kidding, right ? >> >>No, we're not kidding. RTFM: Section, 5.12 Arrays of Length Zero > > > He is kind of right, though. Outside struct (or perhaps union), > zero-sized arrays > make little sense and could be rejected. Or else I am missing something too.
If you've got an array that is a cache of items and you have a constant defined that specifies how many cache entries to use, then you can tune the cache size to any number except zero. That doesn't make any sense, but being able to define a zero length array *does* make sense. The alternative is to hide the array in an extra layer of templates, and to specialise it for zero to not use a real array in its implementation, and that's so nasty. IMHO, this is a good extension, at least until boost::array gets standardised. -- Tristan Wibberley
