On 12/07/2016 12:17 PM, Martin Sebor wrote:
OK. So is the hangup really a problem in how the return type is
documented? I parsed the comment as essentially saying we return true
if the range gets adjusted in any way -- thus a sign change in the first
block would qualify, but we returned false which seemed inconsistent.
Looking at it again, what I think it's saying is we're returning true
only for a subset of adjustments to the range, ie, those cases where the
postcondition does not hold. Correct?
Correct. Would changing the description of the return value to
this make it clearer?
Return true when the range has been adjusted to the full unsigned
range of DIRTYPE, [0, DIRTYPE_MAX], false otherwise.
Yea, that does help. I see you've got an updated version posted. Let
me take a final (?) looksie now that I have a better understanding of
the return value. I think that was the biggest stumbling block on my side.
Jeff
