The Year 2000 bug and its social effects
by Paul H Ray <PaulHRay@aol.com>
15 Apr 1998
I'm writing because I've just become aware of the extent and seriousness
of
the year 2000 bug problem. You know, when mainframe computers with old
programs in them go from 1999 to 2000 they only look at two digits: 99 goes
to 00 and the whole program goes to hell--about every 10 lines of code.
The
result is either nonsensical output or the program crashes. It's mostly
true
of "legacy" programs written decades ago that giant corps and
govts tend to
run on, ones that have never been replaced, for which there's no longer
documentation, in languages their current programmers don't know, and the
original programmers are long gone. Most people with techie background like
me have heard that much. That's what I thought I knew, that a few big corps
were going to take it in the ear to replace these.
This past week I've been reading all the tech-talk on the internet and
browsing the tech books at the bookstores to see how real it is, and how
big a
problem. IT IS TOTALLY REAL, and IT'S A DISASTER.
The first piece of bad news is that 2/3 of corps and govts are still in
denial
and not fixing it--even though it's been known well since 1991-93--and there's
so much computer code that at this date its not even fixable. We're
talking
hundreds of millions of lines of code, and too few programmers, and direct
hits to all the profits they've got, and a date certain when it goes down,
in
a technology where only 14% of major systems projects are done on time.
World
estimate of cost of fixing, just to stay in business, $600 billion. So
they're postponing action in the hopes of a silver bullet--and the best
systems people are saying that there isn't any. It's just tons of old, stupid
code, and it's too late for anything but triage at this stage.
The second piece of bad news is: As soon as the stock market realizes that
these corps will pay billions just to stay in business, they'll try to yank
their money out, probably in the second or third quarter of '99. Crash of
'29
all over again, call it the crash of '99. Big bank runs too. Lawyers are
all
licking their chops at lawsuits flying in all directions, and can paralyze
all
action today, and stop the law cold later. From what I've seen so far, most
executives and systems people have been commanded by their lawyers to say
nothing at all to outsiders for fear of lawsuits and stock prices going
down--so the word isn't getting out.
The third piece of bad news is that the U.S. is less bad than the rest of
the
world in facing this. Japan is doing nothing because of its recession.
Europe is totally preoccupied with reprogramming for monetary integration.
It's an unknown problem elsewhere. Third world countries are going to suffer,
unless their economies are so primitive that mainframe computers are nowhere.
The fourth piece of bad news is year 2000 bugs imbedded as date functions
in
chips. About 50 billion chips that aren't in computers, and 3 to 5% are
affected. Say 1.5 to 2.5 billion chips--hard to identify--imbedded in
electrical power grids, power stations, phone lines, nuke plants, oil
pipelines, train tracking systems, subways and subway cars, new models of
cars, chemical plants, nuclear tipped missiles, elevators, medical equipment,
etc. The infrastructure that is the lifeblood of all advanced and some
developing countries. This has the potential to make most cities look like
Beirut.
The fifth and worst piece of bad news is the "network" effects
of this and
they're awful. My first reaction was "oh, shit!" Like everyone
else, I
hadn't been paying close enough attention.
Network effects are what happens even if you have year 2000 compliant
programs/pcs, because you're part of a network, where if some parts fail,
there's no way out, and no way to buffer or isolate yourself. The good parts
crash along with the buggy ones. Like a car is 90% finished, but key
components don't show up, so it's unusable. Try this: phone networks go
down,
electrical networks and power plants go down, your bank goes down, along
with
all its ATMs, and all international finance, trade, payments, phone traffic,
inventory shipping all go down, air traffic control systems go down, stock
markets stop dead. We're in a time of "just in time" deliveries
not only to
factories but to grocery stores (which run out of food every 27 hours).
People starve, freeze, can't go anywhere, or run out of gas if they do.
Riots
in the streets. It has more potential to make most cities look like Beirut.
Don't take my word for it. Read the dozens (hundreds if you've got the
stomach) of articles at the following sites:
http://tmn.com/~doug (Doug Carmichael's
personal website, click at the "y2k"
link to get a slug of articles: read the one by Frautschi on Embedded
Systems, and the Doug's scenario "Year
2000: who will do what and when will
they do it? Towards actions" For all the current
articles read http://www.year2000.com/articles
and also
http://www.garynorth.com. ((These
guys sound extreme only because they've
been at it for years and have watched a once-fixable problem pass beyond
the
date of fixability.)) There's a helpful not-technical book by Yourden called
*Time Bomb 2000* that spells it all out. He is a top DEC systems guy. The
computer section of your bookstore will have a half dozen 'what to do' books
that will convince you of its reality. I used to have a software company
1970-72 so I can verify the above are true.
You and I may have wanted for the big corporations and governments to be
cut
down to size, or have less power. We're going to get our wish. I for one
am
not so happy about it. I've been in a kind of grief reaction all week. I
have nowhere to run and hide. My business, and I suspect yours too,
disappears whenever there's a depression this deep. We may ask whether we
ourselves can survive in cities (or anywhere offshore).
I saw the Titanic movie recently. That's a prescient movie symbolic of our
time! The iceberg's straight ahead in our path, the ship can't turn in time
(going too fast and rudder too small), it's not really unsinkable, most
of the
crew and passengers are in denial, and everything the captain "knows"
from 25
years at sea doesn't apply to this problem. And there's lifeboats for at
most
half of us (if we're lucky). And it is entirely possible that when our own
ship goes down, those in the lifeboats will not lift a finger for those
who
aren't.
The automatic reaction of "how do I protect me and mine?" is going
to come up
with everyone. The second reaction ought to be "how do we create a
pro-social
response instead of the Mafia-ization that will come with selfish and
individualistic responses? How do we avoid catastrophe?"
That's what I'm struggling with. Maybe the Cultural
Creatives offer some hope
as the kind of people who can be mobilized on the basis of values. Are you
interested in working on some on-line conversations about what to do? If
so,
Who else should we involve? I'm talking about serious, "what to do
and where
do we go live if we can't make a pro-social response" stuff, not just
on-line
gab and breast-beating.
-----
PS
Last night I read something personally useful, which is to frame it in a
risk
analysis format, mostly at a personal-response level. For each aspect of
the
problem, you say: "What if that aspect leads to systems going down
for a
couple of days, a week, a month, a year, ten years--what does that do to
me"?
And they have good discussions of each level for about 20 different parts
of
the system. It's the book by Yourdon, Time Bomb 2000. Very
balanced
treatment--you can see lawyer's warnings to the authors coming out around
the
edges of every other chapter. Be sure to read appendix B.
Summarizing their assessment: Probably a deep recession, with local effects
at
about the level of a month long winter blizzard, but a real possibility
of
some serious breakdowns with food riots by the poor in some cities (Detroit
and LA riots as archetypal examples, but with the added filip of no fire
and
police services). Bare minimum: get your money out of the stock market by
early '99, have several months of cash on hand, have a month's worth of
food
stocked, maybe get out of town if you're too close to riot/fire/no-food
prone
areas, and review your dependencies on big systems that are likely to fail.
There's a huge range of uncertainties because you don't know what goes down
and what the ripple effects are. The ripple effects can be worse than some
keystone system going down.
That's part of the problem, but fails to deal with the socially responsible
issues. So can we address that? Ya wanna play?