If I remember the circumstances correctly, a while ago I planned out a setup
for a client wherein LDEV numbers for printers were stored and then used in a
file equate executed within a program via calling the COMMAND intrinsic.  The
basic idea was to organize what output for which user would go where.

Initially I setup the date for say LDEV 123 as an eight byte sting like this:
"00000123".  I spent hours tracking down a problem with this.  I think that
the base problem was that the :FILE equate would "work" (no error) but then
the file still couldn't be accessed.  The problem that I found seemed to be
with the leading zeroes in my data were confusing some HP code.

I think that there is more to this latent bug than what ATOB is doing.  There
should be a consistant set of rules for what is a 'valid' device class name
and what will be accepted as a (potentially) valid LDEV number.  For example,
are these strings to be treated as valid LDEVs...?

123
0123
000000000000000000123
999999999999999999999

My thought was that MPE should perhaps allow any string of digits (max of 8
characters?) as a potentially valid LDEV #?

Wayne Boyer
Cal-Logic

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *