On Wed, Dec 05, 2001 at 06:34:00PM +0000, John Levon wrote: > > especially if there are more "cases" or (a) and (d) are "enforcing" > > prerequisites for "//stuff"... > > but this loses the semantic information that a and d are mutually exclusive. > That's important.
The semantics are the same (if not, it was a typo). It just makes reading the code "more linear". If you reach a certain point in the body, you can be sure that all "entry tests" succeeded. I find this less obvious in (possibly nested) 'else' branches since I always suspect some hidden "fall through cases". This always makes me thinking about "obvious" code and forgetting the thing I was going to do... Andre' -- André Pönitz .............................................. [EMAIL PROTECTED]
