RE: Subroutines vs Subprocedures was RE: Indicators

On Wed, 2004-08-25 at 13:21, CWilt@xxxxxxxxxxxx wrote:
> Joel,
> 
> Actually this works ok even at v5r2 (v5r1?)  as long as the local variable
> has a (globally) unique name.

Thanks Charles, I tested this at V5R1 and it certainly does work.  So
once again I learn something before 10am... my work here is done :-)

Wouldn't you want to avoid this though?  You would have to ensure that
the local variable name is unique, which kind of defeats the purpose of
making it local... although, back to my other statements about
guaranteed initialization and similar items, I guess I can see a benefit
to this.  In fact, this could come in very handy when using columnar
functions and such for calculations... hmmm... now you've got me
thinking...

As an aside, I wonder how I got hung up on the idea that SQL couldn't
use ANY local variables.  I obviously believed it enough to assert it,
both here and in my articales, so I have to wonder what made me so
confident about that.  Does that ever happen to anyone else?  You think
you understand something completely and then find out you've been
mistaken all along?

My apologies to all for spreading an inaccuracy.

Joel
http://www.rpgnext.com