by Brian Grainger: brian@grainger1.freeserve.co.uk

We are into the new millennium and somehow the world has not come to an end. The electricity is still working, no nuclear bombs went off and most important, the cash machine says I still have money in the bank. Was it all a joke? Did we spend too much money on solving the bug? Since the New Year I have been reading bits and pieces from various sources, making an attempt to try and determine what really happened. This article tries to summarise where we are with the Year 2000 (Y2K) problem at this time.

It is a sad reflection of the times that although there is an explosion of information available it is more difficult to determine the truth. In times past the English were basically an honest lot and the BBC was regarded as the sole custodian of the truth. Now our present government is acknowledged as the master of being able to spin any information into what it wants it to mean. Then one only has to consider how much air time is devoted to discussing the merits of the pound against that for the merits of the euro to show how reporting from the BBC has become increasingly biased. Add to this the fact that everybody thinks they understand IT but in reality most don't and is it any wonder the Y2K situation is confuse?

Last year the media was keen to predict the disasters that were going to occur because of the Y2K bug. They were also keen to blame programmers for the problem. The media always likes bad news over good and if they can castigate somebody all the better. Because they do not understand IT they did not realise that programs were written to meet user requirements. In the early days users demanded 2 digit representation of the year to save storage space, which was expensive. How many people actually wrote the year as 19xx when required. I started to make a conscious effort to do so only last year to get used to the idea of 4 digit years for this new millennium. However I have started to regress already - the folder where these files are stored on the server is called ej100 not ej12000! In a year at most we will all be writing 2 digit years again and programs will reflect what people want. The Y2K problem was the fault of everybody, not just the programmers. Of course, the predictions of doom were a useful tool to get something done about the problem. Once it is realised there is a problem then it has to be solved in some way. If it was said at the outset that it would only cause minor disruption then, in these cost conscious times, there would be less incentive to do anything about it. Doom predictions helped to get effort expended. It also helped the government, because if they predicted doom and then nothing happened they could say what a wonderful job they have done - despite the fact that is the IT industry that has done the work and the government have done nothing at all. It is quite interesting to note that in Italy, where I doubt that even a nuclear attack would get them to work much quicker, they appear to have expended much less money than the UK on the bug, although not quite as low as has been reported!

OK then. Was there a problem? Well, the answer is obviously yes. Even your man in the street realises that 00 could be interpreted as 1900 rather than 2000 and therefore there might be an impact.

Was the problem likely to lead to catastrophe? At the start we had predictions of even washing machines being affected? Anything that might have a chip in it was a potential disaster area. No consideration by the media, or general public, was given to whether (a) the chip involved the storage of the date and more importantly (b) was the date used to control the function? If it didn't store the date then no problem. If it did store a date but didn't actually use this information to control operation then no problem either. Finally, even if the date was stored and acted upon would it make any difference if the date was thought incorrectly to be 1900. I personally could not see how a lift was going to break down even if it did think we had gone back to 1st Jan 1900. I could see there might be a problem with time locks that, say, did not open when it thought it was a weekend so maybe some safes, or even buildings, maybe at risk. Financial institutions, who use dates all the time in their calculations of interest payments, eligibility for pensions etc. were clearly at risk. To be honest, I do know how fly by wire avionics works so I was not sure whether planes would fall out of the sky. At the end of the day, when I was asked by my sister whether I thought anything would happen, I said that yes, things will happen, but I could not foresee disaster. However, I was going to make sure I was not in a plane on January 1st 2000. I think I was right.

In order to confirm that problems have occurred, but they have not been catastrophic, it is necessary to view what has been reported. I list here those things I have read about in various media sources which I believe can be put down to the Y2K problem. They are in no particular order.

1.	U.K.	This one happened BEFORE Y2K, which made me disbelieve when I first heard the report, but the explanation given below convinced me. In the post Christmas sales credit card swipe machines, manufactured by Racal and distributed by HSBC, refused to process the date 1/1/2000. The machines contain a reminder program so that 4 working days ahead it informs the retailer if the transaction has not been processed. Obviously calculating 4 working days ahead into Y2K caused the glitch.
2.	South Korea	Heat supply cut off to families in an apartment block in Pyongchon.
3.	Japan	Radiation monitors and temperature gauges were affected in nuclear power plants although no serious problems resulted.
4.	Malaysia	Some medical equipment failed to function.
5.	Australia	Bus ticketing systems reported to have failed
6.	Sweden	Heart monitoring equipment failed. The problem was resolved without any casualties.
7.	Denmark	Birth registration software took 2000 to mean 1900 and registered the age as 100.
8.	Germany	It is said an online bank customer had an account dated 30/12/1899 and had DM 12,999,997 deposited in it. Apparently the bank took the money back! I am a bit sceptical of this one. Why did it not affect all customers?
9.	USA	Contact lost with a military satellite.
10.	USA	Several computer problems in uranium storage sites and nuclear power stations.
11.	USA	Wells Fargo Bank and a water utility issued documents dated 1st January 1900.
12.	UK	The Portman Building Society sent out statements displaying the date 1990.
13.	UK	A kidney dialysis machine malfunctioned in Glasgow.
14.	UK	A PC in the Aberdeen weather centre failed (no details)
15.	UK	A tide gauge failed in Portsmouth harbour (no details)
16.	Turkey	Six companies reported Y2K problems to their stock exchange.
17.	Italy	Prison records gave incorrect birth, trial and release dates. Again I am sceptical of this. Surely the (release) dates would have been wrong before Y2K. Perhaps the implementation of a bug fix caused 19xx dates to fail??
18.	USA	A man was asked for $91,000 for returning a video 100 years late!
19.	USA	The Y2K Information Centre web site had pages dated 3 Jan 1900
20.	USA	The Philadelphia stock exchange web site was dated Jan 3 100 on the return to work
21.	USA	Al Gore's web site carried the date Jan 3, 19100
22.	USA	Some VISA and MasterCard users were charged twice, because retailers had failed to update their systems for Y2K.
23.	Mali	Freight train management system had problems. Apparently they went back to manual methods.
24.	France	The web weather service displayed the date as 19100
25.	Spain	Nuclear power plants suffered minor computer problems
26.	Denmark	Unidanmark bank payment and information system affected. (No details)
27.	UK	Users of Pegasus Senior software who had not taken annual licence fees would not have a bug fix that was sent out to fix 'age debt' calculations overlooking debts more than a year old. The error would start to occur from Y2K.
28.	UK/US	Lotus Development revealed that a document sent in Notes/Domino messages that contained a year date beyond 2000 (or prior to 1950?) can lead to software crashes. A patch exists on the web site.
29	UK	Ikea accounting systems would not accept its own store cards with expiry dates in the year 2000 so they were all coded as expiring in December 1999. Unfortunately the accounts software had not been updated by January 2000 so manual authorisation was required.
30.	UK	Birth Certificates issued by the Registry Office showed dates as 1900. Hand-written documents will be issued until the bug is removed.
31.	UK	An Office for National Statistics survey showed, in the first week of January, that 74 out of 1114 businesses had suffered date related problems. Well 74 admitted to it!
32.	Windows 3.x and Windows 95 Users	File Manager displays file dates such as 03/01/:0. (I guess this is related to the fact that ':' is next after 9 in the ASCII table!)
33.	WS_FTP Users	Displays file dates such as 1000103

 

The last two items mentioned are problems I have had with my own computers. Not life threatening, but problems nevertheless. As we can see, problems have occurred from around the world. It would seem that a number of problems relate to the date being displayed incorrectly. While serious for a financial institution it is no problem for a web site page, except to lose credibility. However these glitches ultimately have to be solved.

Did the UK spend too much money solving the Y2K problem? It is very difficult to know precisely how much we spent purely on the bug. It is also very difficult to know how much other countries really spent. Do we believe what they tell us? Do we believe the media reports? Apparently the figures quoted for the Italian spend only covered the setting up of their governments equivalent to Action 2000. It did not cover any costs incurred in solving the problem itself. Since the UK is more forthcoming with information I think it is fairly clear a lot of money was spent. However, how much of it was purely to solve Y2K? One way of solving the problem was to update software, which generally required updated PCs. This was an ideal opportunity to buy new equipment on the back of a perceived threat, as it is incredibly difficult to get company accountants to spend money on new hardware at other times! All these minor glitches mentioned above, and those not reported, are going to have to be solved and paid for eventually. I would argue that the UK spent more of its money prior to Y2K whereas the rest of the world will be spending it after Y2K, when it is no longer newsworthy. No, the UK did not spend too much money. No, the IT community is not to blame for wasting millions of pounds. (The media obviously had to find someone to blame again.)

Is it all over? Sadly, no. It is tempting to get complacent now we have had no serious problems. However, those minor glitches have to be solved. Manual solutions and work-arounds are OK but if they are left the efficiency of the business will be impaired. Errors upon errors will accumulate into something serious. Finally, we still don't know whether February 29th 2000 is going to cause any problems. I think all the other dates bandied around, such as the end of the financial year, are not really a problem. However, problems that are already there may not be noticed until the year end computer runs are done at this time! The February 29th problem is very unlikely to be a serious problem. Indeed, problems may go unnoticed. Would you know if your bank did not consider February 29th when paying the interest on your savings account? I wonder how much extra profit they could make if they conveniently forgot to solve that one! I'll leave you with that sobering thought.