Whereas David <[log in to unmask]> wrote:

If there are only resources to add one new system defined variable for
the next century, make sure it's for the full century and year.  I
suggest:
     HPCCYY
which I think reads slightly better than HPYYYY.  Having the century
in a separate variable HPCENTURY variable would be nice, but not
necessary if we had HPCCYY.

If there is to be a fundamental shift from elementary variables to more
complex variables, David's basic suggestion gets to the heart of it.
Implement variables that are obviously useful by themselves: four digit
year, hh:mm:ss.mmm as a number for both calculation and comparison, and
so forth. Then, if one needs fully qualified time (CCYYMMDDHHMMSSMMM),
concatenate it; conversely, if one needs only the hour or only the
minutes, trim that out. But just as important as utility is consistency.

>----------
>From:  Wirt Atmar[SMTP:[log in to unmask]]
>Sent:  Friday, November 22, 1996 10:53 AM
>To:    [log in to unmask]
>Subject:       Re: [HP3000-L] Proposal for new HPCENTURY CI variable
>
>Jeff Kell writes:
>
>>But on the "micro-management" issue... I'd much prefer ONE variable >with
>>all clock information you could substring from as opposed to the >growing
>>number of possible individual date/time/etc variables.
>
>That would be my strong vote, too. The string would contain
>CCYYMMDDHHMMSSMMM (where the last three MMM's are milliseconds). However
>I would be prone to name it something other than HPYYMMDDHHMMSSMMM.
>Perhaps just HPTIME.
>
>Wirt Atmar
>