24 May 96 04:18, Frank Adam wrote to Peter Collis:

 FA>> Talking about strcmp() i have noticed something that nobody mentioned,
 FA>> it actually returns the ascii code of the character it fails on in
 FA>> str1, and negates it if *str2 > *str1.

 PC>> I've never noticed this, what compiler are you using.

 > BC4.02, and TurboC 2, TurboC does it, but i've been told i'm just lucky.

Of course, if you find other RTLs don't support such behaviour (which
wouldn't be suprising) you could always write your own version... 

--- Msged/2 3.36ac alpha 1