Discussion > Y2K thingummy
Also bear in mind that memory and storage were expensive, so data parsimony was a virtue in IT back in the day.
woodentop
Those who enjoyed the Y2K event can look forward to UNIX time wrapping in 2038
BitBucket
Please, not the Y2K thing again, I was in my apartment in Wapping watching the revellers while Mrs. geronimo and her bosom buddy quaffed Dom Perignon and chatted idly with the gangs of merrymakers who'd decided to sit on the wall of our balcony to watch the fireworks. I, with my special pass that would allow me to get anywhere in central london, watched as New Year crossed the globe without a single sign of the famous bug. God knows what I was expected to do if anything went wrong, take the blame I suppose, I couldn't do anything else. I don't know what would have happened if we'd have done nothing, but I am reminded of a childhood joke, where a man comes upon another man tearing up little pieces of paper and scattering them on the ground. "What are you doing?" asks the first man. "Scattering pieces of paper on the ground to keep the elephants away," says the second man. "But there are no elephants round here!" says the first man. "I know," says the second man, "it must be working."
geronimo
geronomo: the Y2K problem had little, if anything, to do with what might have happened "as New Year crossed the globe". See my posts on Unthreaded (link) on Wednesday - 4:42 PM (note my examples of non-trivial problems) and 7:18 PM.
James Evans: I think I answered your "whose fault was it?" question in my reply to Jiminy Cricket on Unthreaded yesterday (link) - 8:27 AM (to my shame after the Bish's decree).
On a far lighter note, these incidents are amusing:
1. In 1992, Mary Bander of Winona, USA, was invited to join an infant class as she was born in "88" and was thought to be 4 . However, Mary was born in 1888 and was 104. (link - page 2). A Y1.9K problem?
2. In 2012, Anna Eriksson of Tierp, Sweden, was invited to join an infant class as she was born in "07" and was thought to be 5. You guessed it: Anna was born in 1907 and was 105. (link)
Robin Guenier
Am I right in thinking we all share common ground in noting that Y2K catastrophism was unfounded?
Robin argues that it was a serious problem and that unsung heroes of IT laboured into the night to remove or reduce associated risks. Others, like me, are not convinced that it deserved so much special attention. Here is some support for that view, from New Zealand:
So the scene was set here in New Zealand for midnight on Dec. 31, 1999. We are just west of the dateline, and thus would be the first to experience not only popping Champagne corks and fireworks, but the Y2K catastrophe, if any. As clocks hit midnight, Champagne and skyrockets were the only explosions of interest, since telephones, ATMs, cars, computers and airplanes worked just fine. The head of the government’s Y2K Readiness Commission declared victory: “New Zealand’s investment in planning and preparation has paid off.”Confident that our millions were well spent, we waited for news of the calamities sure to hit countries that had ignored Y2K. Asia, a Deutsche Bank official had predicted, was going to be “burnt toast” on New Year’s Day — not just the lesser-developed areas of Vietnam and China, but South Korea, which by 1999 was a highly computer-dependent society. South Korea, one computer expert told me, had a national telephone system similar to British Telecom’s. But where the British had wisely sunk millions of pounds into Y2K remediation, South Korea had done next to nothing.
However, exactly 10 years ago today, as the date change moved on through the Far East, India, Russia, the Middle East and Europe, it became apparent that it made little difference whether you lived in Britain, which at great expense had revamped many of its computer systems, or the lackadaisical Ukraine, which had ignored the issue.
With minor glitches that would have gone unnoticed any other day of the week, the world kept ticking on. It must have been galling for computer-conscientious Germans to observe how life continued its pleasurable path for feckless Italians, who had generally paid no attention to Y2K. There were problems, to be sure: in Australia, a bus-ticket machine stamped the wrong date, while in Britain a tide gauge in Portsmouth Harbor failed. Still, the South Korean phone system came through unscathed.
Professor of Philosophy, Denis Dutton [http://www.eschatology.org/index.php?option=com_content&view=article&id=733:guest-article-it-is-always-the-end-of-the-world-as-we-know-it&Itemid=92 ]
Where we may also have common ground, is a shared hope that there are lessons to be learned from the Y2K episode that might help us understand the astonishing spread of alarm over carbon dioxide and how best to separate substance, speculation, and various degrees of unprincipled opportunism. Is that right?
John Shade
Some overhyping by the CIA:
https://www.cia.gov/news-information/speeches-testimony/1999/gershwin_testimony_102199.html
Some collations of what was then 'projected':
http://www.countdown.org/y2k/specific.htm
http://www.countdown.org/y2k/countries.htm
For me, the bottom line is that it was overhyped.
It was something that IT departments dealt with, much the same as they deal with other things that come up - introduction of the euro, changes in regulations, migration to new hardware or operating systems. Much of it had to be - and was - dealt with in the previous years (eg handling multi-year contracts, guarantees and so on).
Martin A
John Shade: yes, predictions of catastrophe were unfounded. But warnings of potential catastrophe if nothing was done (such as the Bank of England Governor's I quoted on Unthreaded) were not.
Your New Zealand example is based the common misunderstanding (see my comment to geronimo above) that Y2K was all about what might have happened at midnight on Dec. 31 1999. It wasn’t. Remember the two examples I gave you on Wednesday (4:42 PM)? One came to light in 2002 and the other in 2010. And the South Korean “computer expert” didn’t know what he was talking about. See this report from early 1999 (link):
South Korea Telecom Operators Almost Y2K ReadySouth Korea's telecommunications operators are on track to be fully Y2K compliant by the end of August, the Minister of Information and Communications Nam Kyong-sok said Tuesday. The government is monitoring progress at the country's major telecommunications operators to ensure there are no outages as the new year rolls in.
At Korea Telecom, Dacom, SK Telecom and Shinsegi Telecom, judged the four most important operators in the country, Y2K work is expected to be finished before late August. Three operators of PCS mobile telephone networks, KT Freetel, LG Telecom and Hansol PCS, are already Y2K ready.
The telecom operators performed a series of tests in May that tested their own networks and the interconnections with other carriers to ensure wireline to wireline, wireline to wireless and wireless to wireless calls still flowed.
However, taking no chances, the Ministry said it plans to establish a Y2K division in December to handle any problems that might arise.
Er ... not exactly "next to nothing".
Ukraine could afford to be “lackadaisical” because, like most countries in Soviet-controlled Europe, it had had introduced digital computing on a substantial scale relatively late; therefore, unlike developed Western economies, they did not rely on systems incorporating software originally developed in the 1960s and 1970s. And the claim that the “feckless Italians” had “generally paid no attention to Y2K” is simply untrue.
Convinced now, John?
Robin Guenier
Convinced of what?
That there was a software problem due to 2-digit dates, and quite possibly still is in many applications. Yes, I am convinced of that.
That there was a plausible threat of catastrophe around the world overnight on 31st Dec 1999? No, I am not convinced of that, and neither, it would seem, are you.
That complex software is often poorly understood and virtually impossible to construct without 'bugs'. Yes, I am convinced of that.
That computer consultants and software companies had a vested interest in hyping up Y2K threats? Yes, I am convinced of that.
That computer programmers often have a reach that is well short of their grasp? Yes, I am convinced of that?
What nuance am I missing here that is exercising you so?
John Shade
Well, John, at least there’s a measure of agreement between us.
I think we’re agreed that focussing on what might have happened at midnight, 31 December 1999 misses what Y2K was really about. However, I have no doubt that, in 1996 and 1997, “there was a plausible threat of catastrophe around the world” arising from potential Y2K failure when 20thC dates came to interact with 21stC dates. Fortunately, because of action taken, that threat was largely dissipated by 1999.
Whereas it’s probably true that “consultants and software companies had a vested interest in hyping up Y2K threats”, few in fact did. Most of the hyping was done by the media – as I’ve explained elsewhere.
OK, I’m a Y2K bore. That’s why I haven’t provided the following link before – but have a look at this:
http://fm2x.com/The_Century_Date_Change_Problem.pdf
If nothing else, read the Executive Summary.
Robin Guenier
JS:
"What nuance am I missing here that is exercising you so?"
It's not a nuance. There was a huge potential problem. Fact. :)
In the UK, we tended to address that possible problem by checking the code. In my own small sphere of experience most of the checking found that the code was OK. But I think it was sensible to have a look.
Was the problem over-hyped? I don't remember that, but maybe I wasn't paying much attention. Too busy working, probably. I remember a few jokes on The Simpsons.
James Evans
I worked for a company that did find a serious Y2K probem. One that would have resulted in a serious river pollution incident which would have supposedly meant a prison sentence, since that was what the government threatened us with.
To prove the possibility was real, a few months after Y2K came and went the valve, which would have stayed open had the bug not been fixed, physically stuck open and did pollute the river. Beyond weird.
TinyCO2
Robin, my views don't come into it, I personally didn't think anything catastrophic would happen, with, or without, action, but my company did, and if it had happened on the change over from 1900s to 2000s then it inistend senior, and by definition, useless, staff were on hand to take control. I don't believe that the stories of the two elderly ladies would have frightened the bejasus out of anybody, what had happened is that the large corporations had been told that catastrophic disasters could result in the change of century. I have no issue with the hard work put in by many of my staff, I believed then, and I believe now, there time was wasted. Finding problems in software written 10 or 20 years ago, with the loss of documentation, the lack of the original programmers etc. isn't easy, and I'm certain that we missed shedloads of them, but still no disaster. Add to that the sight of Mrs. geronimo winking at me as she downed her Dom Perignon to let in the New Year and you will appreciate my irritation.
geronimo
geronimo: poor you – having focussed on the marginally important matter of whether or not there would be any problems at the transition from one century to another, I understand your irritation when there were none. But good for Mrs geronimo.
And, of course, you’re entitled to your belief that taking action was wasted effort – although I'm glad you weren’t giving advice at the time. But here’s a suggestion: read page 6 of my paper. Then answer this: do you really think it would have been best if the National Health Service had not instituted its compliance programme?
Robin Guenier
Feb 15, 2013 at 5:04 PM | John Shade
Well put John. I was doing IT for a bank at Y2K time and the whole thing was overblown to a ridiculous degree. We had contractors in on ridiculous rates for months to 'certify' software that had nothing to do with dates at all. As you and others put these kind of issues//fixes come along all the time in IT and need urgent solutions at various times. Also dates issues in most cases would have become apparent in the preceding years up to the impending year zero anyway and been fixed in due course.
I'm sure there were serious problems due to the issue that needed fixing but as I said these kinds of think occur routinely in IT anyway.
Rob Burton
Rob Burton: “… dates issues in most cases would have become apparent in the preceding years up to the impending year zero anyway and been fixed in due course.”.
So was the Bank for International Settlements at Basel (the epitome of a sober and conservative organisation) mistaken in 1997 in issuing this detailed instruction to the world's banking institutions?
Here's the opening paragraph:
The Year 2000 poses a significant challenge for financial institutions because many automated applications will cease to function normally as a result of the way date fields have been handled historically. Failure to address this issue in a timely manner would cause banking institutions to experience operational problems or even bankruptcy and could cause the disruption of financial markets. As a result, banking institutions must take the necessary steps to ensure that problems and disruptions are minimised.
Most banks were compliant by early 1999 (probably including yours) – subject in most cases to testing and contingency planning. But not all: for example one small German bank had adopted a temporary “windowing” fix for credit cards. It failed in 2010. Correcting it is reported to have cost £270m.
Robin Guenier
The finance industry had this covered many years in advance.
Trade Finance (Letters of Credit and such like), mortgages, commitments, forward/swap deals, commercial loans. All had dates going past 2000 many, many years in advance of the millennium.
International payments? I was in charge of the team that had the contract to roll out the new SWIFT system in Central Europe - that was in 1996-1997, part of the reasons being Y2K issues.
Y2K compliance was overblown and great opportunity to make money. The international banking system I was responsible for had few Y2K issues and its core was written in early 1970's, and that was based on systems from the 1960s. It was never certified compliant. It would have failed compliance, but it had ZERO issues with the millennium.
My friends in the UK were laughing their heads off at the craziness of those times. The rates and the jobs and overblown nature of the whole escapade.
Y2K was a typical Anglo-Saxon/old Europe event. It is typical example of why empires fall. They eat themselves from within. Climate change is another example.
For all the talk of Government intervention, Y2K compliance/auditing, the ex-Soviet block, the third world basically just got on with risk. And with systems that had more older and more diverse foundations.
Jiminy Cricket
Well, Jiminy, I agree with you about climate change – “a typical Anglo-Saxon/old Europe event” puts it well.
But not Y2K: there were enough failures to demonstrate that. Take banking: one small bank got one relatively minor aspect of its systems wrong and it cost them £270m (see reference above). BTW I agree many financial systems had dates going beyond 2000 that were, where necessary, fixed on a case-by-case basis long before the late 1990s. And, as for the ex-Soviet block and the third world, they either used old analogue systems or had introduced digital computing relatively late; therefore, unlike developed Western economies, they did not rely on systems incorporating software originally developed in the 1960s and 1970s.
And the Y2K exercise had other benefits. For example, here’s a comment from an IMF report on the post 9/11 resilience of the US dollar Payment System:
Contingency planning advanced substantially in the run-up to the century date change (Y2K). Y2K planning incorporated the involvement of senior management and boards of directors and asked business managers to consider how they would operate their businesses (and not just their back offices) in the event of a Y2K-related disruption. In a letter to financial institutions in March 2000, federal bank regulators noted that detailed contingency plans developed for Y2K specified the minimum level of output and services necessary for each major business process. They also noted that simulated operational failures and scenario building helped reduce the time needed to respond to operational problems and improved decision making and internal communication.
But I suggest we drop this. Maintaining pressure on the CAGW madness is a better use of our energies.
Robin Guenier
The biggest obstacle was that, in most organisations, there was no one responsible 'at board level' for the information systems on which the organisation depended. The exceptions were usually those that specialised in software development or were in a closely related field. Where I worked, the Y2K work was quite a small project that was completed TWO years before the big day and the changes were included in the regular update schedule. This wasn't as early as you might think as the software then had to be installed on client sites and they had to test it to their satisfaction!
If vast sums were paid for Y2K specialists, it only highlighted the organisation's poor management. Not a problem for the IT professionals, just some good pay days!
The government realised that until people at the top of the organisation knew how important it was, the risks would not be addressed. That is why there was such a big effort to inform the public. There were no lists of IT directors. In many cases there wasn't even an IT director!
The main problem was that when the sewage pumps failed and the emergency number was dialed, the phones wouldn't work and the maintenance workers' cars wouldn't take them to the scene of the emergency. If the cars needed petrol, tough! If someone became ill, the same problems would be evident in trying to get an ambulance or trying to fix the phones. I know that one phone company used old fashioned radios to ensure they could communicate if needed. Doing nothing would have been as bad as a pilot not bothering with a pre-flight check. After all, so few problems are found, why waste that time when everyone could get to their destination sooner!
I have heard of several stories where BIG BIG Y2K problems were found. There were also some BIG BIG problems found that were not Y2K related. For some, it was the first time that the software was audited. (See Robin Guenier's last two posts!)
There was also the problem that testing was sometime difficult. Some monthly billing systems did not not allow the date to change so, to bill on the 11th of the month, you just had to wait until the next 11th of the month. If it was a Sunday, there was a choice to be made that involved delicate consideration.
In the last few years there have been several IT problems at British banks and they may have been Y2K related, but were too embarrassing to be made public. Remember that a system for driving licences could be struck down 17 years after 01/01/2000, when people born on that date become eligible to drive a car, or after 18, or 21, or 24 when they can drive a bus. So the problem has not yet been fully left behind! And then there are the IT systems for pensions to consider!
I was told sewage pumps and lifts needed special attention as many flagged when servicing was required. Ironic that these H&S requirements caused even greater risks than those they were trying prevent.
There was also the problem of finding people within the company who knew the technology of the old equipment.
Robert Christopher
It wasn't a made up problem. It was talked about from the mid 80s in IT circles.
IT departments started checking their code at least a couple of years before. There was something of a last minute dash, where COBOL programmers were hauled out of retirement and were making silly rates contracting.
I think part of it was that there was uncertainty as to whether systems would be in trouble when the last two digits of the year couldn't be lazily assumed to be OK, so it was mainly testing and applying stock solutions where necessary.
The media got hold of it and started to blow it up into a big deal. There was an old bloke with a young wife on one news feature. He'd been in IT for 30 years or something and they'd moved to a smallholding in the middle of Scotland on the basis of his judgement that there was going to be a meltdown.
There were silly stories about how hospital equipment, central heating controllers and anything with a processor iwas going to stop working. Err, these things may keep track of time, but have no record of the date.
cosmic
As the good Bish has decreed that IT talk is distracting us from our normal devotions, I will waffle here for a second.
JS writes:
"Part of the Y2K problem was the basic carelessness of programmers using 2-digit representations for calendar years. So that is embarassing for the software profession, a profession in which the foolishness of that apparently lay out of sight to many. Another embarassment is poorly documented and hard-to-test software. A third embarassment is the writing of code without routine checks to prevent, or failing that to detect, 'impossible' values such as negative lifetimes or other durations. So, I can quite understand computing folks being a bit touchy here."
I get touchy about this subject purely because the Y2K thingy was a huge issue. And I get irritated by people saying that it wasn't.
I was only in the IT industry for about 8 years, and didn't write any of the suspect code. So I have no sensitive spot there.
The main cause of the problem, as I see it, was that programmers believed (and still do for all I know) that they were creating throw-away products with a very short shelf-life. A programmer producing code in, say, the early 80s wouldn't have worried about a possible problem that might have arisen in 2000. The idea would have seemed absurd. Who would still be running the same code in 20 years??
However, organisations tend to think that if something isn't broken, then don't fix it. Or more accurately, they think that if something isn't broken too much, then please bodge a solution that will do for the next year or so.
When I was working in computing, most companies that I came into contact with were running with ancient systems that had been bodged, re-bodged and re-re-bodged to the nth degree.
That's what caused the problem. Whose fault was it? Take your pick.