Theos-Talk Archives (June 1997 Message tl00383)

Remove all stock investments :)  see

> From: "ET" <et@nospam.com>
> Newsgroups: comp.software.year-2000
> Subject: Newsweek Article, for those who missed it
> Date: Mon, 16 Jun 1997 11:05:06 -0700
> Message-ID: <5o3vdg$ol2@dfw-ixnews5.ix.netcom.com>

Drink deep from your champagne glasses as the ball drops in times square
to usher in the year 2000. Whether you imbibe or not, the hangover may
begin
immediately. The power may go out. Or the credit card you pull out to pay
for
dinner may no longer be valid. If you try an ATM to get cash, that may not
work, either. Or the elevator that took you up to the party ballroom may
be
stuck on the ground floor. Or the parking garage you drove into earlier in
the evening may charge you more than your yearly salary. Or your car might
not start. Or the traffic lights might be on the blink. Or, when you get
home, the phones may not work. The mail may show up, but your magazine
subscriptions will have stopped, your government check may not arrive,
your
insurance policies may have expired.
   Or you may be out of a job. When you show up for work after the
holiday,
the factory or office building might be locked up, with a handwritten sign
taped to the wall: out of business due to computer error.
   Could it really happen? Could the most anticipated New Year's Eve party
in
our lifetimes really usher in a digital nightmare when our wired-up-the-wazo
o
civilization grinds to a halt? Incredibly, according to computer experts,
corporate information officers, congressional leaders and basically anyone
who's given the matter a fair hearing, the answer is yes, yes, 2,000 times
yes! Yes--unless we successfully complete the most ambitious and costly
technology project in history, one where the payoff comes not in amassing
riches or extending Web access, but securing raw survival.
   What's the problem? It's called, variously, the Year 2000 Problem, Y2K
or
the Millennium Bug. It represents the ultimate indignity: the world laid
low
by two lousy digits. The trouble is rooted in a seemingly trivial
space-saving programming trick--dropping the first two numbers of the
date,
abbreviating, say, the year 1951 to "51." This digital relic from the days
when every byte of computer storage was precious was supposed to have been
long gone by now, but the practice became standard. While any idiot
familiar
with the situation could figure out that the world's computers were on a
collision course with the millennium, no one wanted to be the one to bring
it
up to management. And, really, which executive would welcome a message
from
nerddom that a few million bucks would be required to fix some obscure
problem that wouldn't show up for several years?
   So only now, as the centurial countdown begins, are we learning that
the
digit-dropping trick has changed from clever to catastrophic. Because
virtually all the mainframe computers that keep the world humming are
riddled
with software that refuses to recognize that when 1999 runs out, the year
2000 follows. When that date arrives, the computers are going to get very
confused. (PCs aren't as affected; sidebar.) So that seemingly innocuous
trick now affects everything from ATMs to weapons systems. Virtually every
government, state and municipality, as well as every large, midsize and
small
business in the world, is going to have to deal with this--in fact, if
they
haven't started already it's just about too late. Fixing the problem
requires
painstaking work. The bill for all this? Gartner Group estimates it could
go
as high as $600 billion. That amount could easily fund a year's worth of
all
U.S. educational costs, preschool through grad school. It's Bill Gates
times
30!
   That tab doesn't include the litigation that will inevitably follow the
system failures. "You can make some very reasonable extrapolations about
litigation that take you over $1 trillion, and those are very conservative
estimates," says Dean Morehous, a San Francisco lawyer. (Conservative or
not,
this is more than three times the yearly cost of all civil litigation in
the
United States.)
   Come on, you say. Two measly digits? Can't we just unleash some sort of
roboprogram on all that computer code and clean it up? Well, no. Forget
about
a silver bullet. It seems that in most mainframe programs, the date
appears
more often than "M*A*S*H" reruns on television--about once every 50 lines
of
code. Typically, it's hard to find those particular lines, because the
original programs, often written in the ancient COBOL computer language,
are
quirky and undocumented. After all that analysis, you have to figure out
how
to rewrite the lines to correctly process the date. Only then comes the
most
time-consuming step: testing the rewritten program.
   It's a torturous process, but an absolutely necessary one. Because if
we
don't swat the millennium Bug, we'll have troubles everywhere.
   Electricity. When the Hawaiian Electric utility in Honolulu ran tests
on
its system to see if it would be affected by the Y2K Bug, "basically, it
just
stopped working," says systems analyst Wendell Ito. If the problem had
gone
unaddressed, not only would some customers have potentially lost power,
but
others could have got their juice at a higher frequency, in which case,
"the
clocks would go faster, and some things could blow up," explains Ito.
(Hawaiian Electric revamped the software and now claims to be ready for
the
year 2000.) Another concern is nuclear power; the Nuclear Regulatory
Commission says that the Bug might affect "security control, radiation
monitoring... and accumulated burn-up programs [which involve calculations
to
estimate the hazard posed by radioactive fuel]."
   Communications. "If no one dealt with the year 2000 Bug, the [phone]
network would not operate properly," says Eric Sumner Jr., a Lucent chief
technology officer. He's not talking about dial tones, but things like
billing (watch out for 100-year charges). Certain commercial operations
that
run phone systems by computer could also go silent if the software isn't
fixed.
   Medicine. Besides the expected mess in billing systems, insurance
claims
and patient records, hospitals and doctors have to worry about embedded
chips--microprocessors inside all sorts of devices that sometimes have
date-sensitive controls. The year 2000 won't make pacemakers stop dead,
but
it could affect the data readouts it reports to physicians.
   Weapons. NEWSWEEK has obtained an internal Pentagon study listing the
Y2K
impact on weapons and battlefield technologies. In their current state, "a
year 2000 problem exists" in several key military technologies and they
will
require upgrading or adjustments. One intelligence system reverts to the
year
1900, another reboots to 1969. The report confidently states that as far
as
nuclear devices like Trident missiles are concerned, "there are no major
obstacles which will prevent them from being totally Year 2000 compliant
by
Jan. 1999."
   Money. Banks and other financial institutions generally will go bonkers
if
they don't fix the year 2000 problem. The Senate Banking Committee is even
worried that vertiginous computers might automatically erase the last 99
years' worth of bank records. Some Y2K consultants are advising consumers
to
make sure they don't enter the 1999 holiday without obtaining hard-copy
evidence of their assets. According to Jack Webb of HONOR Technologies,
Inc.,
ATMs won't work without fixes.
   Food. In Britain computers at the Marks & Spencer company have already
mistakenly ordered the destruction of tons of corned beef, believing they
were more than 100 years old.
   Air-Traffic Control. "We're still in the assessment stage, determining
how
big the problem is," says Dennis DeGaetano of the Federal Aviation
Administration. One possible danger is computer lockup: while planes will
keep moving at 12:01 a.m. on Jan. 1, 2000, the screens monitoring them, if
not upgraded, might lock. Or the computers might know where the planes
were,
but mix them up with flights recorded at the same time on a previous day.
("You can bet we're going to fix it," says DeGaetano.)
   Factories. Ford Motor Co. reports that if the Bug isn't fixed, its
buildings could literally shut down--the factories have security systems
linked to the year. "Obviously, if you don't fix it, your business will
stop
in the year 2000," says Ford's David Principato. Even if a manufacturing
company aggressively solves its own problem, though, it might be flummoxed
by
a supplier who delivers widgets in the wrong century.
   Just About Everything Else. Larry Martin, CEO of Data Dimensions, warns
that if not adjusted, "on Jan. 1, 2000, a lot of elevators could be
dropping
to the bottom of buildings," heading to the basement for inspections they
believe are overdue. Similarly, automobiles have as many as 100 chips; if
they are calendar-challenged, experts say, forget about driving.
Computerized
sprinkler systems could initiate icy midwinter drenchings.
   Like leaves rustling before a tornado, there have already been
harbingers
of a bureaucratic meltdown. At a state prison, a computer glitch misread
the
release date of prisoners and freed them prematurely. In Kansas, a
104-year-old woman was given a notice to enter kindergarten. Visa has had
to
recall some credit cards with expiration dates three years hence--the
machines reading them thought they had expired in the McKinley
administration.
   The $600 billion question is whether we'll fix the Bug in time. The
good
news is that the computer industry is finally responding to the challenge.
For months now, squadrons of digital Jeremiahs have been addressing tech
conferences with tales of impending apocalypse. The most sought-after is
Peter de Jager, a bearded Canadian who scares the pants off audiences on a
near-daily basis. "If we shout from the rooftops, they accuse us of hype,"
he
complains. "But if we whisper in an alley, no one will listen." Last week
in
Boston de Jager demonstrated the rooftop approach: "If you're not changing
code by November of this year," he warned, "you will not get this thing
done
on time--it's that simple. We still don't get it."
   But we're starting to. Most major corporations now have year 2000 task
forces, with full-time workers funded by multimillion-dollar budgets, to
fix
a problem that their bosses finally understand. They're aided by an army
of
consultants and specialized companies. Some, like Data Dimensions, offer
full
Y2K service, providing tools, programmers and guidance. Others, like
Peritus,
sell special software to help find offending code and, sometimes, even
convert it. (The final, most arduous stage, testing, still defies
automation.) These firms are the new darlings of Wall Street. But buyer
beware--consultants are coming out of the woodwork to exploit the
desperation
of late-coming companies. Someone might promise a phalanx of brilliant
programmers to fix the Bug, but "for all you know, it could be 10 people in
a
garage doing it by hand," says Ted Swoyer, a Peritus exec. Still, the
creation of a Y2K-fixing infrastructure is encouraging.
   It's not uncommon to find gung-ho efforts like the one at Merrill
Lynch:
an 80-person Y2K division working in shifts, 24 hours a day, seven days a
week. It'll cost the company $200 million, a sum that could hire Michael
Eisner and fire Mike Ovitz. "Our return on investment is zero," says
senior
VP Howard Sorgen. "This will just enable us to stay in business."
   So maybe we're not in for a full-scale disaster. Let us assume--oh God
let
it be true--that those in charge of life-sustaining applications and
services
will keep their promises to fix what needs fixing. The costs and
liabilities
of not doing so are too huge not to. (On the other hand, when did you last
see a huge software project that met its deadline and worked perfectly?
Just
asking.) Still, there will almost certainly be severe dislocations because
of
the mind-boggling enormity of the problem.
   Even the most diligent companies don't have total confidence they can
fix
everything. Consider BankBoston, the 15th largest commercial bank in the
United States. Early in 1995, the company realized that "it was a problem
that could bring an institution to its knees," says David Iacino, who
heads
the bank's Team 2000. To stop a meltdown, BankBoston has to probe 60
million
lines of code. The harder BankBoston works at solving the problem--it now
has
40 people working full time on it--the more complicated it seems. "Every
day,
when we see something new we haven't thought about, we get additional
angst,"
says Iacino.
   Of the 200 BankBoston applications that need revamping, only a handful
have been completed so far. BankBoston is now separating the essential
work
from the noncritical, and if the Bug causes less dire problems, like the
heavy vault doors swinging open on New Year's Eve, it'll just cope:
"Vaults
are physical things," says Iacino. "If push comes to shove, we can put a
guard in front."
   Now, if BankBoston, which started early and has been driving hard, is
already thinking triage, what is going to happen to institutions that are
still negotiating in the face of a nonnegotiable deadline? The Gartner
Group
is estimating that half of all businesses are going to fall short.
"There's
still a large number of folks out there who haven't started," says Matt
Hotle, Gartner's research director.
   As businesses finally come to terms with the inevitable, it's going to
be
panic time. In about a year, expect most of the commercial world to be
totally obsessed with the Bug. "Pretty soon we have to just flat stop
doing
other work," says Leo Verheul of California's Department of Motor
Vehicles.
   But no amount of money or resources will postpone the year 2000. It
will
arrive on time, even if all too many computers fail to recognize its
presence.
   "It's staggering to start doing mind games on what percentage of
companies
will go out of business," says Gartner's Hotle. "What is the impact to the
economy of 1 percent going out of business?" Or maybe more: Y2K expert
Capers
Jones predicts that more than 5 percent of all businesses will go bust.
This
would throw hundreds of thousands of people into the unemployment
lines--applying for checks that may or may not come, depending on whether
the
government has successfully solved its Y2K problem.
   What is the U.S. government doing? Not enough."It's ironic that this
administration that prides itself on being so high tech is not really
facing
up to the potential disaster that is down the road a little bit," says
Sen.
Fred Thompson. If Y2K indeed becomes a calamity, it may well be the vice
president who suffers--imagine Al Gore's spending the entire election
campaign explaining why he didn't foresee the crisis. (Gore declined to
speak
to NEWSWEEK on the Y2K problem.)
   Here's the recipe for a federal breakdown: not enough time and not
enough
money. While the Office of Management and Budget claims the problem can be
fixed for $2.3 billion, most experts think it will take $30 billion. Rep.
Stephen Horn held hearings last year to see if the federal agencies were
taking steps "to prevent a possible computer disaster," and was
flabbergasted
at the lack of preparedness. His committee assigned each department a
letter
grade. A few, notably Social Security, were given A's. (The SSA has been
working on the problem for eight years and now has it 65 percent licked;
at
that rate it will almost make the deadline.) Those with no plan in
place--NASA, the Veterans Administration--got D's. Special dishonor was
given
to places where inaction could be critical, yet complacency still ruled,
like
the departments of Labor, Energy and Transportation.
   State governments are also up against the 2000 wall. California, for
instance, finished its inventory last December and found that more than
half
of its 2,600 computer systems required fixes. Of those, 450 systems are
considered "mission critical," says the state's chief information officer
John Thomas Flynn. These include computers that control toll bridges,
traffic
lights, lottery payments, prisoner releases, welfare checks, tax
collection
and the handling of toxic chemicals.
   As bad as it seems in the United States, the rest of the world is
lagging
far behind in fixing the problem. Britain has recently awakened to the
crisis--a survey late last year showed that 90 percent of board directors
knew of it--but the head of Britain's Taskforce 2000, Robin Guenier,
worries
that only a fraction really understand what's required. "I'm not saying
we're
doomed, but if we are not doing better in six months, I really will be
worried," he says. He expects the cost to top $50 billion. On the
Continent,
things are much worse; most of the information-processing energy is
devoted
to the Euro-currency, and observers fear that when countries like Germany
and
France finally tackle 2000, it might be too late.
   Russia seems complacent. Recently Mikhail Gorbachev met with
Representative Horn in Washington, expressing concern about how far behind
Russia is in dealing with the Bug; Gorbachev raised its possible impact on
the country's nuclear safeguards.
   The list can go on, and on and on. "It's like an iceberg," says Leon
Kappelman, an academic and Y2K consultant. "I would certainly be
uncomfortable if Wall Street were to close for a few days, but I can live
with that. But what if the water system starts sending water out before
it's
safe? Or a chemical plant goes nuts? Anybody who tells you 'Oh, it's OK'
without knowing that it's been tested is in denial."
   It's tough out there on the front lines of Y2K. And in less than a
thousand days, it might be tough everywhere. "There are two kinds of
people,"
says Nigel Martin-Jones of Data Dimensions. "Those who aren't working on
it
and aren't worried, and those who are working on it and are terrified."
   Tick, tick, tick, tick, tick.