RolandRB wrote: > On Oct 3, 3:05 pm, "Lou" <lpog...@hotmail.com> wrote: >> "RolandRB" <rolandbe...@hotmail.com> wrote in message >> >> news:7367742a-7f44-45dc-a96e-a401108d5a5e@j68g2000hsf.googlegroups.com... >> >>> I'll have a special macro containing the above macros plus a >>> declaration of the global macro parameters so that it gets called at >>> job start and compiles the macros. These macros are for macro testing >>> purposes only and will never be used for anything else. >> Does anyone else see an infinite regression gathering steam here? If there >> are macros to validate macros, how do we know that the validation macros are >> valid? Somewhere this process has to bottom out. > > The validation macros themselves are aimed to be very simple and not > used for anything else. In fact I desingned them to be useless outside > their own construct. For example, nobody in their right mind would use > a sas method to compare two flat files.

Huh? Earlier this week, you asked "Has anyone got a file compare utility written in sas? ... This is for "files" and not datasets."

What did you mean by "files" if not flat files?