The two Toms tell:

Tom E tale:
> As you allude to at the end, HPYYYYMMDD is a bad example -- it is really a
> function, not a variable, (and should have been implemented as such, but now
> that it exists, we're stuck with it).  As you note, the actual value of the
> variable will change over time, while a locally-created
> variable-with-the-same-name will not.

EXCELLENT!  Thanks for pointing out another key argument against
user-created HP... variables:  their semantics may not match those
of the real HP variables!  (In the case of HPYYYYMMDD, the semantics
are that the variable accurately reflects the current system date whenever
accessed, unlike a user created variable.)

Tom B's tale:
[re: problems that can occur by a user creating an HP... variable]
> This was a point I tried to make earlier.  Exactly this happened to me.  It
> took some time to diagnose what happened.

Thanks for the example!

Stan

Stan Sieler                                           [log in to unmask]
www.allegro.com/sieler/wanted/index.html          www.allegro.com/sieler