Hi Roger! :-)

 RS>  It's a very broad statement.  I should probably change it to "don't 
 RS> make assumptions about anything that is not defined in the language 
 RS> standard".    Since this is a C conference, I think we need to be 
 RS> careful to qualify statements that take in to account implementation 
 RS> dependent features. 
Agreed, also taking into account that different versions of the C standard
exist, and not all compilers may be up to full C99 yet.

Ciao
Pascal

--- Msged/LNX 6.1.1