Encode Time Problem

[egdorf@zaphod.lanl.gov (Skip Egdorf)]

I have just run into a problem with universal times.
The system is a large simulation system that runs on
Suns (Allegro, Lucid, and CMU Lisps)
PCs (Allegro CL/PC)
and
Macs (MCL 2.0.1, built by patching 2.0, not from CDROM)

The system was just moved to a Quadra 800 with System 7.1

After running for some time (and doing a large number of
encode/decode universal times in the process) the system dies with
the error

> Error: Universal time is not defined for year: -27989
> While executing: ENCODE-UNIVERSAL-TIME
...

The numbers for this call come from an earlier decode-universal-time.
The input scenario for this run runs correctly with the same source
code on the other systems listed above.

It seems that the times are getting screwed up. I seem to recall
some discussion about this in the last year on this mailing list,
but do not recall if the issue was resolved.

Is this a bug? Is there a patch? Is there a workaround?

					Skip Egdorf
					hwe@lanl.gov