 |
Date with Destiny: The Year 2000 Computer Bug
Matthew L. James
Technology Adviser, Science, Technology, Environment and Resources Group
23 March 1998
Millennium Scaremongering
There are predictions of gloom and doom for the arrival of the centesimal
year in 2000 relating to modern computer systems. These may not act as
usual when the first day of Year 2000 arrives. Prudence may be wise for
certain crucial sectors such as airlines. Qantas and Ansett may spend
$150 million working on fixes. Aircraft might not fly on the night in
question. Pacemaker wearers may be wise to stay at a hospital overnight.
Stockmarkets may best remain closed for the duration.
Computer systems of various forms exist within cars, hospitals, aircraft,
offices, communications networks, financial institutions, missiles, shops
and in almost any other sphere of modern business and technology. They
need the correct date to work properly or they may cause system failures.
It is the embedded code within the microchips and computer circuits where
the devil lies with the Year 2000 (Y2K) computer bug. In the early days
of computers, it was necessary to limit the amount of input data as much
as possible, in order to maximise processing storage. As a consequence,
when inputting years, the prefix '19' was often left out. The Y2K problem
comes with the Year 2000 when computers may instead think it is 1900.
Banks might then calculate account interest over 100 years!
Countdown to Armageddon?
The doomsayers predicting mass computer failures in 2000 have served
the purpose of alerting the world to possible effects that require remedial
action now. By the year 2000, the world might have spent around $3-600
billion on fixes for the Y2K bug. Organisations have to allow funds to
address and fix the problem.
Fortunately, most systems built since 1996 are compliant but the remedies
for others are time consuming and labour intensive. The solutions are
proving to be relatively straightforward. An easy way to determine any
adverse outcomes is to run operations with test data to see the possible
effects. Anyone can do that on a personal computer by changing the date.
More subtle changes may be required, e.g. in lifts, or in large systems
that may require extensive debugging. A date appears on average once in
every 50 lines of software code.
Concerned agencies are providing computer consultants, lawyers and insurers
with good incomes. Compliance conferences also abound. Sydney company
MFX Research claims to have the only product that automatically opens
files, examines them, diagnoses and corrects for the Y2K bug. Legal firms
advise on matters of contract, negligence, insurance and liability for
Y2K.
Australia Prepares for Doomsday or Salvation
As each sector of the economy views the future, their role in the debugging
process comes under scrutiny. Australian businesses may spend up to $3-6
billion to fix the problem, but perhaps half have yet to do anything at
all. Senior staff should be aware of the problem and should devise a well-resourced
risk management plan. For those that haven't acted, time is fast running
out. The Y2K National Strategy Committee guides Australia's progress and
there is also a national standard. The Australian Taxation Office has
not yet finalised its position on the application of tax law to the Y2K
issue but will do so by July 1998. So, RU Y2K OK?
Australian governments may have to spend up to $300 million to minimise
any effect on their systems. The Commonwealth's Year 2000 compliance project
is a resource for overall monitoring at http://www.y2k.gov.au/html/. The
Federal Office of Government Information Technology (OGIT) has overall
responsibility for Y2K. It suggests that it may be easy to fix large computers
but not the embedded codes in chip circuits. In December 1997, the Australian
National Audit Office reported on weak departmental progress in meeting
compliance. It found that Commonwealth agencies risk revenue and security
breaches if they don't act fast.
Millennia: Come Again?
We now live in an age where failure of complex systems may have widespread
effects. On 22/8/99, receivers for the Global Positioning System of navigation
satellites may fail due to a program date error. This may disrupt transport
and information systems globally. The date 9/9/99 may cause problems too
as it has been used as a product expiry date. This might be a useful prelude
to the ultimate test, but experts suggest a more useful test action deadline
of 31/12/98.
If all of this is not confusing enough, imagine the possible effects
stemming from the Y2K computer bug. It just might be enough for us all
to want to stay home and not join in the global celebrations for what
is a purely arbitrary date(1). The first day of 2000 is a Saturday, which
is probably just as well for all of us! A Monday was the first day of
1900. The year 2000 is a leap year, having a February 29, unlike the years
1700, 1800 or 1900. Note that the new millennium actually starts on Monday
1/1/01, one full year later, not 2000.
Many nations have jumped on to the centesimal bandwagon with plans for
grand events. London intends celebrations at the centre for world timekeeping
at Greenwich in a huge and costly dome. Fiji wants to usher in the new
century with a light show, claiming to be the first country in the world
to see each new day. New Zealand's Chatham Islands also have a claim,
as do Tonga and Kiribati. Sydney has joined in to say that it is the first
major city in the world to greet a new day.
Daylight Saving Chaos or Twin-Zone Sanity?
All this depends on the setting of the International Dateline and time
zones in 1884. The history of time zones and daylight saving in Australia
is fascinating, reflecting parochial interests.(2) Much the same applies
to the setting of international time zones(3). However, there are arguments
for daylight saving in terms of road accident reductions, energy savings,
crime decreases and leisure increase, but perhaps at the expense of farming
practices(4). But it has also cost business in States that stay on standard
time.
It is conceivable that on the night of Friday, 31/12/99, some computer
systems will suddenly fail at midnight in New South Wales, the Australian
Capital Territory, Victoria and in Tasmania. Half an hour later, the same
thing might occur in South Australia, while half an hour later still,
some computers may die in Queensland. The final death throes would occur
two hours still later in Western Australia. Such is the legacy of our
time zones. Queensland, Western Australia, the Northern Territory, Norfolk
Island, Christmas Island and the Cocos (Keeling) Islands do not observe
daylight saving time. Lord Howe Island clocks move ahead 30 minutes to
match NSW.
Such daylight saving and Y2K chaos could be overcome with a little lateral
thinking. In the north, there is less difference between summer and winter
day lengths and the sun rises later. The Commonwealth Government has constitutional
powers over weights and measures. It could create a daylight saving zone
in the south-west and south-east only leaving the rest of rural north
Australia on standard time. The zone would not match state boundaries
but may well better match demographic time desires. All major cities would
be an hour ahead of standard time with the main rural areas not changing.
Times could change at equinoxes.
- Gould, S. J., Questioning the Millennium: A Rationalist's Guide
to a Precisely Arbitrary Countdown, Jonathan Cape, 1997.
- Baker, G. 'Daylight Saving Time', Research Note no.17, Parliamentary
Research Service, Department of the Parliamentary Library, 16 October
1995.
- Edwards, P. 'A Brief History of Time Zones', Sky and Space,
Sydney, December 1996.
- Hughes, P. 'British Time (Extra Daylight) Bill', Research Paper
96/4, House of Commons Library, London, January.
|
 |
