by Chizen Chou

Chizen Chou, a computer engineer specializing in integrated
circuit design methodology and automation for 16 years in the
West Coast, wrote this article to help the public to scrutinize
the myth of "personal electronic library of this nation's nuclear
weapons secrets" portrayed by FBI Director Louis Freeh.

The Congressional testimony of Mr. Freeh's showed how information
technology illiteracy can seriously hinder one's ability to
interpret evidences correctly and produce ill-advised judgments
in the new information age.

He stated "Dr. Lee used an elaborate scheme to move the
equivalent of 400,000 pages of extremely sensitive nuclear
weapons files".  His odd way of describing the 806 Megabyte (MB)
of computer files appears to be intentionally misleading
non-technical persons to believe it is a very huge amount of data
and thus the back-up operations very abnormal.  If you know
scientists commonly deal with data in many thousands of MB and
mere 806 MB of computer files can virtually fit into one CD,
would you still think that is a huge amount of data?

What's the real meaning of "extremely sensitive" nuclear weapons
files?

They weren't even classified secret when Dr. Lee downloaded.
They were labeled as "Protect As Restricted Data," or PARD,
calling for lower levels of security.  The wording of "extremely
sensitive" is apparently extremely exaggerating.

Was the downloaded data really sensitive at all if others cannot
understand and use them?  Let's analyze further.

It's been alleged that the data included computer codes
simulating a thermonuclear explosion; files on the physical
properties of materials in U.S. warheads; and "input decks" or
mathematical descriptions of the geometry of nuclear devices.
However, are these files comprehensive enough to give other
country any lead in designing, testing, or improving nuclear
weapons?  Or they are just a narrow scope of certain explosion
testing data under contain conditions?

Furthermore, could these programs and data be possibly decoded
and understood without US weapon experts and the original
developers' collective assistance and supervising?  Could those
computer codes easily ported to other machines for possible
reverse engineering attempt?  Complex computer codes are
impossible to be run on other machines without very significant
effort from well-informed experts to properly set up all the
software and hardware foundations.

Even if any team of foreign experts would possibly crack on the
codes and data will they eventually found out most of the
algorithms and ideas are already available in open scientific
literature?  John Richter testified at the court that 99 percent
of the information was available in open scientific literature.
We probably should welcome any rogue country to waste their
resource to crack on these publicly available secrets.

Dr. Lee's download process consumed nearly 40 hours over 70
different days.  Mr. Freeh made it appeared to be a very
deliberate effort to conduct such a long series of downloads.
But, can you imagine a thief would go to the same bank in 70
different days to steal money 70 times?

More logically, it might imply that the downloads are very common
in the Labs.  The downloads were sparse thus might be related to
other ordinary purpose, such as backup certain research works.
To download 806 MB into one single CD will be much faster and
easier to be hidden or transported.

The wording of "secret, portable, personal electronic library of
this nation's nuclear weapons secrets" and "the complete design
capability of Los Alamos at that time -- approximately 50 years
of nuclear weapons development, at the expense of hundreds of
billions of dollars" are apparently very extreme over-statements
since it indicated the downloaded simulation programs and data
together is a complete weapon design library. 806 MB probably
only represents a very small fraction of all the required
simulation programs and data to fully describe an extremely
complex nuclear explosion.

Take another example, can you imagine the information of a
complete simulation of a Mercedes-Benz crash test is equivalent
to a library of its complete design or the "electronic library"
of Mercedes-Benz engineering secret?

Freeh emphasized Lee's repeated attempts at 3:30 a.m. on
Christmas Eve, after his access had been revoked, to delete
files.

First I wonder how one could re-gain access to secure computers
after the revocation.  Either this is another "honest mistake" or
our national Labs' security is a joke.

Freeh said Lee's conduct was in an attempt to cover his tracks
before he was caught.  I have some doubt that would be Lee's
motive since we know all critical computer files have backup
schedule (normally once a day).  Unless Freeh also described that
Dr. Lee somehow created the files after the previous backup and
then deleted before the next backup and the Labs system backup
was normally scheduled after 3:30 a.m.

Logically, I would wonder what did Lee do beside deleting files.
Did Mr. Freeh withhold other information which could help better
determining Lee's real motive?

Working on Christmas Eve may not sound abnormal as many may
think.  Too many programmers work on odd hours to maximize
productivity, not to mention crazy scientists who even work in
their dreams.  If Freeh is kindly enough to disclose the national
Labs employees' working patterns on Christmas Eves, holiday
nights it will definitely prevent us from making ill-informed
judgments.

Chizen Chou
P.O.  Box 2066, Cupertino, CA