On Thu, 15 Apr 2004 14:21:12 -0500, Jeremy Siek <jsiek@osl.iu.edu> wrote:
Hi Gennaro,
On Apr 14, 2004, at 4:06 AM, Gennaro Prota wrote:
I was thinking, however: what about returning the index of the first bit in common (or npos), instead of true/false?
That functionality sounds useful,
Besides usefulness, the idea was to not "throw away" information that the function already has (well, almost: it knows the two blocks which have a non-null bitwise and; then it's just a matter of calling my lowest_bit function). Of course one could want a mechanism to get, then, the second bit in common, then the third, etc... If anyone has ideas... :)
but I think the current intersects() function should not be removed since it provides for a common use case. Adding another function with the semantics you describe would be good.
I see. But my worry is that we'll have a much dirtier interface if we add utility functions (shortcuts to other functions) light-heartedly. Since dynamic_bitset has already a lot of functions I think we should have a fairly minimalist policy. Genny.