BKDLNY2K.RVW   990428

"Deadline Y2K", Mark Joseph, 1999, 0-312-20202-4, U$24.95/C$34.99
%A   Mark Joseph
%C   175 Fifth Ave., New York, NY   10010
%D   1999
%G   0-312-20202-4
%I   St. Martin's Press
%O   U$24.95/C$34.99 212-674-5151 fax 800-288-2131 www.stmartins.com
%P   294 p.
%T   "Deadline Y2K"

First of all, I have to say that after the halfway (or maybe two
thirds) point, I really got into this book.  The action is tense,
interesting, and well thought out.  The geeks save the day, even
outclevering the semi-evil capitalist.

That said, I'll go back to the opening I originally intended to use
for this review:

The Bad News is that Joseph figures we're all gonna die, or a
reasonable facsimile thereof.

The Good News is that Joseph gets so much else wrong that, based on
the fact that he predicts disaster, we'll probably have a fairly easy
time of it.  The author has a deep, profound, and abiding lack of
understanding of Y2K.

The one concept he does seem to understand is that the most serious
Y2K problems arise from the interconnected nature of modern systems,
and the fact that small problems, added, multiply themselves.  This is
good, and is used to good effect at times in the story.

The list of negative accomplishments is rather longer.  Lets start
with stereotypes.  We have magic bullet Y2K fixing software.  We have
the traditional, mythical salami scam.  We have the hygienically
challenged genius hacker.  (Who starts out as a completely
irresponsible flake, and suddenly transforms into a dedicated,
managerial, recruiting, social worker.)  We have the multi-millionaire
tycoon having his bagel every morning and shooting the breeze with the
boys from the old 'hood.  (This is the same guy who alienates his wife
and kid by never having time for them.)

Then there are specific technical errors.  The credit card companies
hit the Y2K wall in 1997 because of expiration dates, and therefore
have generally dealt with the issue.  There will not be a sudden
collapse of the credit card system in the hours just before midnight,
December 31, 1999.  (In fact, a large number of failures in the book
happen in the last few hours of 1999, rather than months or years
before, or slightly after.)  The failures of old GPS (Global
Positioning System) terminals that will happen in August of 1999 have
a very vague and conceptual relation to Y2K in that the problem
results from a "number of weeks" field that is limited to ten bits. 
It has nothing to do with a sudden change of format.

And I have no idea why you need to calculate dates in order to tell
the difference between O positive and O negative blood.

Then there are the more general mistakes.  One platform definitely
does not fit all: it is highly unlikely that a single IBM mainframe
could accommodate all the control programs from a power utility, civic
infrastructures, telecommunications companies, and transit
authorities.  (Nobody snuck a VAX or a UNIX box into the mix anywhere? 
Besides, I understand that the New York subway still has drivers.  But
you could check with Bombardier ...)  It is also unlikely that a
single mainframe could handle the processing load required to run all
of them concurrently.  (We will ignore, for the moment, the likelihood
of a small band of geeks being able to fulfill all the checking that
large corporations were unable to do.  *And* get it to run right the
first time they do a smoke test ...)

There is no reason that someone supplying utility software would have
any need to access full bank data records.  Yes, there is a T-4, and,
yes, it is bigger than a T-1 or T-3.  But almost nobody uses that
designation: it just isn't that useful in discussing the bandwidth
bundles that are provided.

And I don't care what von Neumann said, a data file, even a very large
data file, is *not* the same as code.  Yes, if someone was careless, a
very large data file could possibly overrun its bounds and crash a
system, but there would be no difference between a store completely
restocking and a store starting up for the first time, now would there
be?

Joseph's dialogue is readable, although not particularly great.  His
characterizations could use some work.  At one point (page 166)
someone goes from being "amused and thrilled" to being ready to "die
of anxiety" in the space of one sentence, with nothing happening in
the interim.

Entertaining, yes, but only after you get through the annoying bits at
the beginning.

copyright Robert M. Slade, 1999   BKDLNY2K.RVW   990428