The Y2K Computer Bug -- Update September 1999

by Ben Best

 

  • "Panic now and avoid the holiday rush."

    This essay updates information in my original Y2K essay written in May 1999 dealing with the y2k computer bug. That essay provides essential background for this update and should be read first.

    Since I have now delved more deeply into the subject, I can give better historical perspective, and then proceed to updates and current thoughts.

     

  • HISTORICAL PERSPECTIVES ON THE Y2K PROBLEM

    In the 1960s there was dispute among computer authorities concerning whether date formatting should be standardized for two digits or four digits. Computer memory was expensive at that time, and a four-digit standard could have resulted in huge expense for organizations that store large amounts of date information. Nonetheless, the International Standards Organization (ISO) decided upon a four-digit date standard. But in the United States the Defense Department imposed on the National Bureau of Standards to adopt a two-digit date. Since the US government was the largest consumer of hardware & software at the time, suppliers felt it was more prudent to ignore the ISO standard.

    In 1993 the South African-born Canadian Peter de Jager wrote an article entitled "Doomsday 2000" for the September issue of COMPUTERWORLD magazine, which argued that disaster is inevitable in January, 2000 and that it was too late to prevent it. That same year the Defense Department began to become concerned when NORAD (NORth American aerospace Defense command) set its network computers to 31-December-1999. At the simulated stroke of midnight all the computer screens locked-up.

    The first IBM PCs had no hardware clock, and MS-DOS required the user to manually set the date and time whenever the system booted-up. IBM's PC-AT introduced a Motorola Real Time Clock (RTC) chip which included CMOS RAM with 50 bytes of registers which could be used to store configuration information in a power-off condition. When the PC-AT was turned-on, the BIOS (Basic Input/Output System) chip code would examine the two-digit year information received from the RTC, and apply century digits of 19 or 20 depending on whether the RTC information was greater than 80 or not.

    Unfortunately, the subsequent manufacturers of BIOS chips (AMI, Award Software and Phoenix Technologies) did not begin to achieve Y2K compliance until the mid-1990s. Only BIOS systems that date from 1998 or later can be relied upon as being unlikely to have a Y2K problem, because AMI was so late in implementing compliance (Award and Phoenix were compliant by 1996).

    Computers using 80286 and 80386 processors are very likely to have a Y2K hardware problem. An 80486 is still likely, although not as likely, to have a problem. And although hardware problems are even less likely with an early Pentium or Pentium Pro, Y2K problems can still occur. BIOS is no longer in dedicated chips, but has been replaced by "flash" BIOS which can be software updated.

    How BIOS or the Operating System handle the information received from the RTC is highly dependent upon the manufacturer. Although it would seem that a date field of 1900 or 1980 or 00 could do no harm at Y2K rollover, if a chip is used in an embedded system that calculates time intervals for control functions, there could be serious consequences. Coding for chips used in embedded systems has often been customized for the machinery in which it was used. A BIOS or embedded-system code that increments the RTC digits could cause complete malfunction/nonfunction if incrementing from 99 to 100 at Y2K rollover (although the problem may not occur at rollover, but on the first re-start after rollover -- which could be days, weeks, months or even years later).

    Aside from hardware, many old software packages are not Y2K compliant. ACCESS 2 is not compliant and MicroSoft has announced that it offers no fixes -- users are expected to upgrade to the newest version of ACCESS. (Actually, I understand that compliance can be forced on ACCESS 2.) MicroSoft WORD version 5.0 is not Y2K compliant and problems have even been found with Windows 98. Users can upgrade their Windows 95 or Windows 98 systems with a patch from the MicroSoft website ( http://www.microsoft.com). Netscape Navigator 2.02 and greater are Y2K compliant, but earlier versions are not.

    The oldest Macintosh operating systems are accurate to the year 2040 and current models will be good until 29,940 A.D. (which may be important for life-extensionists who are a little slow to upgrade their computer systems). Leap days were added to the calendar by Julius Caesar in 49 BC. In the 16th century pope Gregory made the adjustment that years divisible by 100 are not leap years unless they are also divisible by 400. True date compliance requires that years evenly divisible by 400 be a leap-year only if the century is divisible by 9, with a remainder of 2 or 4. Thus, 2000 and 4000 are leap years (20/9=>2 remainder & 40/9=>4 remainder), but 2400, 2800, 3200 and 3600 are not leap years.

     

  • THE STRUGGLE FOR Y2K COMPLIANCE ON MY PC

    A March 1999 Gallop phone survey of Americans found 28% were planning to prepare for Y2K by stockpiling supplies, but only 4% had plans to update their home computers. Even I have not attempted to check my PC until this summer, partially because of the warning that setting the computer forward to January, 2000 might prevent being able to set it back ( http://www.usatoday.com/life/cyber/tech/ctf493.htm). But wandering-around Toronto's COMDEX computer-fair in July I found 4 or 5 booths (out of hundreds) that were offering "Y2K solutions".

    A couple of these booths were offering free diskettes which could be used to test your PC for compliance without having to manually change the date. Another was selling a CD ROM for $20 which would check all PC software for Y2K compliance and update it via the Internet.

    The first diskette involved a power-on test that automatically set the PC date forward, checking BIOS and the RTC for Y2K and leap-year compliance. The software informed me that my PC had passed all the tests except the RTC century rollover test. The second diskette did much more thorough testing -- both power-on and power-off -- for a wider selection of dates, including leap year for 2000, 2001 and 2004. My PC passed all the tests, except gave me "Anomaly(4)" for the "CMOS only 01/01/2000" test -- which was their way of saying another RTC failure.

    The vendor behind the first diskette was Micro2000 ( http://www.micro2000.com) who were selling Dallas chip-sets with a 4-digit RTC (http://www.dalsemi.com) to replace the Intel 2-digit model. Several times I asked the salesman at the Micro2000 booth why I should be concerned that my 2-digit Pentium chip only supplies the year if BIOS is correctly adjusting the century. There may have been some sense to his explanation, but I couldn't understand it.

    The vendor of the second diskette was Fernlink2000 (http://www.fernlink.com). This vendor's salesman was much more sophisticated. He explained to me that some software will bypass BIOS, getting the year from the RTC and then applying the wrong century digits. Even a 4-digit RTC is not a complete solution, because the century information for a 2-digit RTC is stored in memory location "32H", whereas for a 4-digit RTC it is stored at "48H" (??? -- this is not hexadecimal). Some software could conceivably check 32H and get the wrong information. (See http://www.fernlink.com/download/IB04.txt for a detailed explanation.)

    I mentioned to the Fernlink salesman an article I had read by Peter de Jager in which de Jager said that the only software in widespread use that bypasses BIOS and goes directly to the RTC is the NT operating system and and old version of Adobe Acrobat (the last may be incorrect -- de Jager's article was removed from the web for copywrite reasons). The vendor told me that de Jager is a "politician" who has an "ax to grind". He acknowledged that only a tiny percent of software would probably be affected, but he said that a tiny percent could still have a devastating effect. He said that being a programmer, I should understand how easily programmers can choose to bypass BIOS and go directly to the RTC.

    The package that was marketed as a "silver bullet" for all my application software was called "Year2000 Now" (http://www.imsisoft.com). I ended up being more paranoid about their product than of Y2K software on my PC. I had to go through the usual installation rigmarole, which included putting their icon on my desktop and requiring me to enter my product serial number, a valid e-mail address, a username, password, a "secret word" (in case I forgot my password). When I connected to the Internet, the software then automatically downloaded a plug-in. I probably now have a "Trojan Horse" in my PC, which may be another reason to get a new PC next year. The software gave me a Y2K "thumbs-up" for my copies of WORD, EXCEL and CorelDraw, but failed to find my ACCESS 97 and gave me "undetermined status" for a lot of other software.

    On the weekend of September 24-26 I attended the Toronto Computer Fest (http://www.compfest.com/). One theatre presentation dealt with PC software, including Y2K issues. When I asked the presenter about Y2K, he only knew about operating system issues. His solution to pre-pentium chips was to upgrade hardware. He referred me to Miles White, a local Toronto PC guru, to answer more detailed questions.

    Miles was standing nearby listening, so I went over to him. He told me that embedded systems in nuclear power plants have all been replaced -- as if that were the end of the embedded chip problem. He said that the chip makers could not provide information on BIOS Y2K compliance because the software of the computer manufacturer would determine how BIOS was used. He said that the same argument applies to the manufacturers of embedded systems.

     

  • THE FEDERAL Y2K ACT

    In July, 1999 President Clinton signed the Y2K Act into law. The Act had been pushed by a Republican congress, with Clinton resisting until a compromise was reached. Clinton is well positioned politically on this issue (as he so often is) since this will give Republicans a greater liability for being blamed for many of the Y2K problems that may arise.

    Essentially, the Y2K Act provides for a 90-day "cooling-off" period before lawsuits can be initiated in connection with problems related to Y2K computer failures. Punitive damages are capped at $250,000 for companies that employ 50 or fewer workers. Municipalities are protected from punitive damages arising from failure to provide water, sanitation, fire-fighting, etc. due to Y2K problems. Many people may be breathing easier by being freed from responsibility, but this could easily translate into a more complacent approach toward finding possible Y2K bugs. This may be another reason to think seriously about having reserve water supplies.

    Detailed information on the Y2K Act can be found at:

  •    http://www.zdnet.com/pcweek/y2k/y2kliability.html

     

  • LANDMARKS

    On Sunday, August 22nd the Global Positioning System (GPS) rollover occurred. Only GPS systems manufactured before 1994 were noncompliant, and most of these were in Japan. Although many Japanese car & truck GPS screens went blank at 9am Sunday, there were no reported accidents. Car navigation by GPS is particularly useful in Tokyo because most of the streets are unnamed. The main reason there were no problems in North America is because General Motors and Ford did not start installing GPS navigation until 1996. Comparison of the GPS rollover with the Y2K problem seems strained.

    I have often seen mention of the "September 9th, 1999 problem", and I repeated the idea that this could be serious in my previous essay. I regret doing this because I have little belief in it. In my years of programming, I have never seen 9/9/99 signifying a null date or end-of-file. I have seen negative 9999999 to signify null, however. But in the case of a date, I would sooner expect 99/99/99, a date which could never be reached. September 9th could be written 09/09/99, but this would be a very unlikely choice for a null date. I have not worked with COBOL very much, and this problem evidently is mainly associated with COBOL programs. The only reported 9/9/99 failure I heard-of was for a pharmaceutical company in China ( http://dailynews.yahoo.com/h/nm/19990921/tc/yk_china_9.html). I find it difficult to believe that there were no other problems, and suspect that problems in North America were covered-up by companies unwilling to admit shortcomings in Y2K compliance. Suspiciously, I returned a rented car to Hertz that day and their computer systems were down. However, if the problem was easy to cover-up, it couldn't have been devestatingly serious.

    There are also alarmists mentioning the Leonid meteor storm expected in November, which could knock-out some satellites. But there is not good reason to believe that the 1999 Leonid storm will be worse than the 1998 storm. And even if a few satellites were knocked-out, it is doubtful that there would be any cataclysmic consequences. Those interested in Leonid meteor showers can visit: http://www-space.arc.nasa.gov/~leonid/history.html

     

  • THE STOCK MARKET

    I am expecting a stock market crash in October. October is often a stressful time for the stock market. Last year the Dow Jones Industrial Average index fell from a July high of over 9000 to an October low of under 8000. In some years, October has been uneventful, but I don't think 1999 is going to be such a year. My theory is that retailers, manufacturers and distributors experience a capital crunch in October in preparing for Christmas. Capital that could be in the stock market is invested in stock on shelves.

    I believe that the normal stress on capital that occurs in October combined with the fact that people with Y2K fears will be withdrawing their capital is going to have a snowball effect. I have spoken to several people who have little knowledge of markets or of computers who are planning to move their money out of equities and into secure bonds -- or even just in the bank.

    Since July of this year the Dow has been up-and-down around 11000. Has the American economy really grown 22% since July, 1998? According to a recent issue of THE ECONOMIST [24-July-1999], the Information Technology sector accounted for 35% of America's economic growth in the 1995-1998 period, despite accounting for only 8% of the GDP. Stripped of computers, growth in manufacturing of durables has actually slowed down in the 1995-1998 period.

    Many companies are putting a freeze on computer development this coming fall because they have already tested their systems for Y2K compliance and they don't want to risk introducing new, untested, systems. Moreover, many Internet stocks dominating the market have astronomical price-to-earnings ratios because their price is driven by anticipated rather than actual earnings. Stories of people who became millionaires with a modest investment in MicroSoft stock nearly 20 years ago has inspired heavy buying.

    I have become so convinced of a fall crash that I have invested a significant portion of my personal wealth in short-selling futures in market indices. I have also bought a few options (LEAPS), but this seems more risky to me. I don't recommend this to others, but I would be happier to see cryonics organizations with their patient care funds in T-Bills for the next 8 months.

    Being as heavily invested as I now am in a bear market, I can no longer be viewed as an objective source of information. The market weighs heavily on my mind. I feel sad & foolish when the market goes up and joyful & emboldened when the market goes down. This is not a very healthy mental state, especially for a technophilic & economic optimist like me. I do believe that computer technology may be on the verge of accelerating economic growth rate to unprecedented heights. But not until we trip over the threshold of the 21st century.

    I'm not foolish enough to think I am influential enough to think I am going to profit by creating panic. I have nothing to gain by duping myself. I short-sold index futures because of a deep conviction, which has not changed enough for me to disinvest. If I really believed in Y2K doom, I wouldn't be trusting that any profits I make will not be lost in the investment company's records. I worry more about loss of value of money.

     

  • WILL THERE BE A Y2K POWER FAILURE?

    Electrical power is the lynchpin of technological civilization. Without it, computer bugs cannot be fixed, water cannot be pumped, gas & gasoline cannot be pumped, telecommunications is nonfunctional, etc. The idea of a power-grid makes excellent sense for the normal condition in which failures of individual units occur randomly while other units are able to take-up the slack. But in a Y2K situation of simultaneous failure of many units, power grids may be unable to stand the load.

    That Y2K will occur in winter is fortunate, insofar as electricity usage is no more than 75% of summer highs, even in extremely cold weather. This is because people use electricity for air-conditioning, but use fossil fuel for heating. As of July, however, only 68 of the 103 commercial nuclear power plants in the US were claiming to be "Y2K ready". Many of the plants will have to wait until the Fall to be taken off-line for repairs. If the Nuclear Regulatory Commission is not certain by the end of September that a plant can be made Y2K ready, the plant may be shut-down. It would not surprise me if technophobes living within a 50-mile radius of a nuclear power plant did not begin to demand that plant be shut-down as a precaution. If many plants are taken off-line, there will be an increased dependence on the electrical utilities that use coal to supply 60% of US electricity.

    Banks and investment companies have done the most toward Y2K compliance of their systems. Utilities have not done nearly so well -- they are less well financed and are heavily dependent on embedded systems. Railroads may even be worse than utilities, and railroads supply most of the coal required by electrical utilities.

     

  • WILL THERE BE NUCLEAR ACCIDENTS?

    Russia has the equivalent of about 60 Chernobyls -- which could potentially result in a nuclear accident. And Russia isn't the only country with nuclear reactors and not enough time or money to ensure Y2K preparedness.

    Russia is special, however, in that it has 6,000 nuclear weapons pointed at the United States. Prior to the war in Yugoslavia a joint early warning system was under development between Russia and the US. But co-operation ended when NATO began bombing, and did not resume until mid-September ( http://www.wbn.com/y2ktimebomb/International/rvann9937.htm). Although failures are most likely to render systems inoperable, there is enough uncertainty to keep many people worried. Thailand is shutting down its missile computers out of fear of accidental firing (http://www.timesofindia.com/today/16busi19.htm).

    Pentagon officials have said that Y2K power failures in Russia could result in a loss of early-warning systems that would make the Russian military feel dangerously vulnerable to sneak attack. Russia and China are both listed as having high risk of failure.

     

  • WORLDWIDE IMPACTS OF Y2K PROBLEMS

    North Americans suffered little during the recent economic turmoils in Asia, and this has no doubt contributed to the feeling that Y2K readiness elsewhere in the world is not so important for the domestic economy. But Y2K problems may be very different, in part because exports account for 20% of the American economy and a much higher percentage in Canada (although in large part exports are to the US).

    Although the White House is exuding confidence of American Y2K readiness, the State Department released a survey in July stating that half of 161 nations have a medium-to-high risk of breakdowns in energy, transportation and telecommunications. In Latin America, the Gartner Group has ranked Costa Rica, Ecuador, El Salvador and Uruguay as being least prepared. An Intel spokesman has identified Malaysia as being most at-risk in Asia, stating that disruption of supplies from that country could seriously impact Silicon Valley. Others worry not only about the suffering that Y2K could cause abroad, but of the political and military upheavals that could be triggered.

    The OPEC-induced oil crisis in the 1970s resulted in 4% shrinkage of the American economy. The US still depends on imported oil for most of its needs -- 15% comes from Venezuela, 14% from Saudi Arabia and 7% from Nigeria. Canada supplies 16% and Mexico supplies 13%, but these countries are considered low-risk. For the other countries, The Gartner Group has estimated a Y2K failure rate of 50% or more for organizations in those countries.

    A Y2K Global Readiness Status can be found at http://www.iy2kcc.org/CountryWeb.htm

     

  • SOME SURVIVAL NOTES

    My major concern is water. I have bought 3 large plastic garbage cans which I plan to fill in December. Because of my concern that these could be knocked-over, I also bought two rugged 25-litre gasoline containers from Canadian Tire. An article in the "Y2K Survival Handbook" recommends storing 4 litres of water per person per day -- 2 for drinking and 2 for food preparation & sanitation. All stored water should be treated with 5.25% sodium hypochlorite bleach, about 2 drops per litre or a teaspoonful per 10 gallons. Waterbeds can hold up to 400 gallons, but they can contain toxic chemicals unless thoroughly cleaned.

    I have already begun stocking-up on cans of tuna fish, roasted soybeans, dried tomatoes and dried milk. As it gets closer to December I will buy oatmeal and other foods that might not last so long. I am stocking-up on plastic bags for solid excrement disposal (kitty litter is recommended). I rarely eat cooked foods anymore, so I am not worried much about sources of heat. I have bought a very warm sleeping bag.

    For some people, Y2K preparations appear to have the quality of a game -- like building a treehouse. But for me it is an inconvenience, driven by my fears. I wouldn't do this for fun. I've got much better things to do. And, ironically, I do not expect a power-outage to last over two weeks at the very most -- probably much less.

    In December, 1999, 7-Eleven stores will begin promoting "Y2K Ready -- 4U" displays of bottled water, canned tuna, candles, flashlights and batteries. James Keyes, the chief operating officer of 7-Eleven Inc has said, "...this may be the single biggest opportunity we've ever had"

     

  • CELEBRATIONS

    The Year 2000 has been called a momentous milestone devoid of meaning, comparable to an automobile odometer hitting 100,000. Nonetheless, the more people choose to impute meaning to it, the more meaningful it becomes. "What will you be doing for the Millennium?" is a question some people take as a challenge to do something extraordinary.

    A $2 billion Millennium Dome will be unveiled at Greenwich, England at the stroke of midnight, 2000. Millions of balloons will be released in Europe, a Millennium Wall is being build in Fiji along the dateline, a million people are expected in Jerusalem and Ottawa is planning the "largest millennial rock festival in Canada" at the Corel Centre.

    Sydney, Australia is claiming to be the first major city in the world to celebrate the new Millennium (an hour before Tokyo, but two hours after Wellington, New Zealand -- which is also claiming to be the first "major city" to celebrate the Millennium). Sydney is spending $5 million for the "biggest fireworks show in the world". The theme is to highlight the Sydney Harbour in a television broadcast expected to reach a billion people.

    In addition to the fireworks, Sydney is sponsoring a "City of Light" project in which the owners of up to 150 buildings in the business district are being encouraged to build light displays costing from $35,000 to $100,000 each. Ironically, while celebrants are making plans for the "world's biggest lightshow", EnergyAustralia has issued warnings that if all Sydney residents test their electrical equipment at midnight, it could bring down the power grid. If the world's biggest light show turns into the world's first Y2K power failure, it will give North Americans a jump-start on the morning of New Year's Eve.

     

  • HYSTERICS, SKEPTICS AND SPIN-DOCTORS

    Prior to the last few years, "Millennialism" has been associated with the prophecy in the Book of Revelations of a thousand years of peace & harmony following the second coming of Christ. Serious students of the BIBLE don't confuse this reference to a duration with a reference to the timing of the second coming. But associative Apocalyptic thinking is easily jump-started by circumstance.

    As the days progress toward January and media attention increases, the average citizen is likely to feel caught between lunatic ravings and spin-doctoring. The credibility gap will become a broad breach.

    Gary North (http://www.garynorth.com) has become the Web's most notorious doomsayer. Not only has he claimed that "We are headed for a disaster greater than anything the world has experienced since the bubonic plague of the mid-14th century", but he has said "I want to see Y2K bring down the system, all over the world. I have hoped for this all of my adult life". His ravings have inspired "The Gary North is a Big Fat Idiot Page" (http://206.28.81.29/) and http://www.garysouth.com. Other skeptics of the "Y2K hype" can be found at http://www.equipped.com/y2ktoc.htm

    The press is full of "We're Y2K ready" stories. But these are mainly just unsubstantiated claims, not proven facts. The chairman of the Florida government Y2K task force has suggested that the "Y2K lying rate" may be as high as 50% ( http://www.orlandosentinel.com/news/y2k/s070199_y2k01.htm). And even where honest claims are made of Y2K compliance or Y2K support for products have been made, these claims have later been withdrawn by vendors upon discovery of problems in a high percentage of cases ( http://www.computerworld.com/home/news.nsf/all/9908032trackr/).

    The director of the White House Y2K council said last December that "one of our priorities for the coming year is the prevention of public overreaction to the Y2K problem". But public officials lose their credibility when their statements are predictably designed to create a certain effect rather than reflect the truth.

    On the other hand, there is refraining from media activities which might inspire panic. Warner Brothers, evidently afraid of being held responsible for a modern "War of the Worlds" type hysteria, cancelled plans for its movie "Y2K", which was to depict a Y2K bug that made New York City vulnerable to terrorist attack ( http://www.wbn.com/y2ktimebomb/Media/jy9930.htm). No such fear affected musician Loudon Wainwright III, who released a song entitled "Y2K" which warns of currency collapse, phone failure, etc. ( http://www.rosebudus.com/wainwright/y2k.html).

    For an update on information in this article go to The Y2K Computer Bug -- Update December 1999


    [GO TO BEN BEST'S HOME PAGE] HOME PAGE