Talk:Year 2000 problem/Archive 1

Blockbuster

Lol i remember when this happend "Items rented from Blockbuster prior to January 1st, 2000 and returned after January 1st were reportedly marked for astronomical late fees ($91250), as though the items were 100 years overdue." my dad was shocked on how much he owed.--EliteMike 02:17, 5 November 2006 (UTC)

First recorded mention

Earlier than any USENET newsgroup chatter was the book "Computers in Crisis" by Jerome and Marilyn Murray (Petrocelli, 1984), later reissued under a different title by McGrawHill in 1995.--4.252.171.51 20:46, 8 February 2007 (UTC)

"Event horizons"

It turned into a major fear that critical industries (electricity, financial, etc.) and government functions would stop working at 12:00 AM, January 1, 2000 and at other critical dates which were billed as "event horizons."

The article event horizon seems to have no relevance (it refers to the physics meaning - the point beyond which no information can reach the observer).

De-linking event horizon for this reason -- 201.50.249.78 16:19, 18 October 2006 (UTC)

Webpage

It so happens I maintained one of the best-known Y2K news sites, "Sanger's Review of Y2K News Reports"; see: http://web.archive.org/web/20000520011237/http://sangersreview.com/. I'm not embarrassed to admit it :-) , because I was always agnostic as to the extent of the impact. This really is a huge topic and this article really still doesn't do it justice. --Larry Sanger

Edward Yourdon and Edward Yardeni

I wonder if Edward Yourdon and Edward Yardeni had to fight it out over who was who in the Y2K world. Ortolan88

Oh yeah. Much confusion over the names.  :-) They were known as "the Eds." --Larry Sanger

Personal experience

Just a random little aside, but I was personally bit by the Y2K bug--I recieved a notice to show up for jury duty in January of 19100.

Misleading

The article currently states:

The Y2K problem mainly affected countries that follow the western calendar (Saudi Arabia does not). (This is very misleading and needs to be restated.)

I think this remark belongs here, so I'm moving it here. Someone should probably fix it (I don't know what is meant, myself). -- Timwi 20:54, 9 Feb 2004 (UTC)

supermarket chain????

"There were a number of minor problems that occurred even before the beginning of the year 2000. One such example was a supermarket chain in the midwestern United States. When a cash register encountered a credit card that had an expiration date that was after the year 2000, it created a serious error in the computer systems running the cash register. The error caused the computer system to shut down all the cash registers throughout the entire supermarket chain. This was used by experts to illustrate the need for businesses to study whether or not a Y2K bug could cripple them as well."

This sounds like an urban legend. If it is true, then we need to name which supermarket chain it was, what the date was that the incident began, and what city it started in. What date did the shutdown occur? Kingturtle 04:06, 19 Sep 2004 (UTC)

I second this -- the lack of a specific brand name marks this as an urban legend. --Jfruh 02:41, 24 Jun 2005 (UTC)

Looks like an urban legend to me, too. I live in the Midwest, and I didn't hear anything about this. I'm removing it until some sources can be brought foreward. Mance 22:39, 21 December 2005 (UTC)

It sounds highly implausible to me. I'm a cashier myself, and have, on several occassions, scanned a card that turned out to be expired. All that happened was the computer didn't accept it, and said "invalid expiration date". Surely the Y2K bug would simply cause a credit card with an expiration of, say, January 2000 to be read as expired (Interpreting 1/00 as January 1900). Nik42 04:27, 22 December 2005 (UTC)

Trivia

I had heard that at one prominent new year celebration (perhaps in Japan) that a countdown clock displayed 3 seconds to go 2, 1, "Game Over" - will try and verify this!

Meat

"As the decade progressed, more and more companies experienced problems and lost money due to erroneous date data. As another example, meat-processing companies incorrectly destroyed large amounts of good meat because the computerized inventory system identified the meat as expired."

This can't have been more than a few weeks (or months if frozen) before 2000, because meat doesn't last that long, though the article implies it was in the mid-1990s. Ben Finn 18:15, 30 Apr 2005 (UTC)

It was canned corned beef with along shelf life. None was destroyed, see below for the urban legend. It doesn't appear in the BBC archive so it is most likely another bit of misinformation and urban legend. Most of the stories reported here never appeared in CNN or BBC archives. And the story about Japan with an alarm sounding and a series of radiation monitors stopping working was reported here as: a power plant had to shut down.

--Richard Arthur Norton (1958- ) 19:06, 31 January 2006 (UTC)

Miscellaneous

I don't have time to do it myself, but methinks this article could do with a bit of a tidy-up to organize it into better chronological order. It goes over events before 2000, then after 2000, then before, then after, etc. unnecessarily. Ben Finn 18:30, 30 Apr 2005 (UTC)

Disasters avoided!!!

significant disasters such as nuclear reactor meltdowns or plane crashes were avoided

This sounds too sensational to be accurate. If I were to say "I avoided being trampled by a herd of elephants", this implies I was actually in danger of being trampled by a herd of elephants, and hides the fact that I live in Wisconsin where there is no danger of being trampled. CyborgTosser (Only half the battle) 17:19, 24 May 2005 (UTC)

Year 10000?

See Talk:Year 10,000 problem

Leap Year

Another related problem for the year 2000 was that it was a leap year even though years ending in "00" are normally not leap years. (A year is a leap year if it is divisible by 4 unless it is both divisible by 100 and not divisible by 400.) Fortunately, like Y2K, most programs were fixed in time.

I don't get that. Why would the programmers bother putting in the centennial exception if they didn't even bother with four-digit years? Nik42 04:24, 22 December 2005 (UTC)

I think the idea is that this is a unrelated (and mutually exclusive) problem. Some implementations had only two-digit years (the classic Y2K bug), while other implementations had four-digit years but handled leap-years badly. --Saforrest 02:24, 31 December 2005 (UTC)

Most people assume that a leap year occurs every four years (I did until I started doing Y2K remediation work), so most programs used very simple logic to determine whether it was a leap year - namely "can I divide the year by 4 with no remainder". Numerous variations existed depending on peoples understanding of leap years, some of which would work correctly for Y2K and some would not. In some cases they would work for Y2K but not for 2100 (see - we were starting to look ahead!)

"Factoids"

Why is there a section in here of "factoids?" Technically, a factoid means something that is spurious or almost certainly untrue; such things have no place in an encyclopedia. If they are all confirmed true and are just useless trivia, then calling the statements factoids is incorrect. --Mance 22:48, 21 December 2005 (UTC)

Look at the article on factoids -- the alternate (and in my experience, more usual) meaning is "A true but insignificant fact." However, since the term is deprecatory, I agree that it is a poor choice for a section heading. 208.20.251.27 16:55, 9 January 2006 (UTC)

My computer

had some problems running certain programs right. we had windows 95, and when we tried to view the ammount of space left in our C drive it kept telling us that we had 100% free space, but showing us numbers indicating that our drive was 150% full. when we took it in to get fixed (in August of 00, it kept running alright till then) they told us it was Y2K problems. Now I dont expect a section called "Pellaken's Computer" but surley we can add something like "some personal computers encoutered minor problems on Y2K without actually crashing" or something, couldent we? Pellaken 11:24, 29 December 2005 (UTC)

Nonevent

The original form stated: The public largely, but perhaps wrongly, felt y2k was a nonevent.

I then proceeded to look up nonevent, which stated, " Jump to: navigation, search

A non-event is an anticipated or highly publicized event that either does not occur, or simply turns out to be very anticlimactic or boring. Non-events are disappointing, as they are often much hyped prior to their occurrence." and listed Y2k as a nonevent.

Including "perhaps wrongly" does not fit with this definition and creates a contradiction in this encyclopedia.

I slightly disagree with you. The "perhaps wrongly" statement implies that there actually were problems, but the public regarded it as a nonevent. I believe that, at least in this case, the phrase "perhaps inaccurately" is better and more relevant. The original phrase, however, is still not a contradiction. --Zachary Murray 18:11, 3 January 2006 (UTC)

The programming problem

Recently the section "The programming problem" was removed. I re-added it because I felt it's an important section. If anyone feels it should be removed, could you supply a reason here? – Mipadi 17:19, 20 January 2006 (UTC)

Sources, please

These need to be sourced: "U.S. spy satellites that were blinded briefly, or the national high-speed and airport rail systems of Norway that briefly shut down on December 31, 2000, a date that was not tested for." I found a source for the Japan reference. --Richard Arthur Norton (1958- ) 18:17, 31 January 2006 (UTC)

This also was removed: Some industries started experiencing related problems early in the 1990s as software began to process future dates past 1999. For example, in 1993, some people with financial loans that were due in 2000 received (incorrect) notices that they were 93 years past due. As the decade progressed, more and more companies experienced problems and lost money due to erroneous date data. As another example, meat-processing companies incorrectly destroyed large amounts of good meat because the computerized inventory system identified the meat as expired. There were, in fact, many such minor "horror stories" like these, which received much play in the press as 2000 approached.

The article about the corned beef says the food was never destroyed, it just alerted people there was another error in the software. All others are unsourced rumours. Its amusing to read this time capsule. I am suprised no one has corrected it in the past 5 years. This is why Wikipedia gets a bad reputation.

Sadly, this must go: "*In 1996, pallets of Marks & Spencer canned corned beef were scheduled for disposal by an inventory program. The program thought the cans to be 96 years past their expiration date, because the labels read "12-1-00" and the program misinterpreted this as December 1, 1900." No references found in BBC archive. If you find one, restore it. Otherwise assume its another urban legend.

--Richard Arthur Norton (1958- ) 19:04, 31 January 2006 (UTC)

NPOV

"In the end, significant disasters such as nuclear reactor meltdowns or plane crashes did not occur, but the number of non-critical Y2K errors encountered on January 1, 2000 was extensive. Due to the lack of disasters and the faulty "end of the world" expectations, the public largely, but perhaps inaccurately, regarded the Y2K passage as a non-event."

This assumes that there were extensive errors after Y2K. DO you think it should be modified and restored? --Richard Arthur Norton (1958- ) 19:21, 31 January 2006 (UTC)

Anecdotal stories

It looks like all these stories are apocryphal but here they are:

• Windshear alert systems at five major and four smaller US airports produced error messages for about two hours, according to the New York Times.
• A US military spy satellite was blinded for at least two hours and operated at reduced capacity for a couple of days when its ground station went dark, according to the New York Times. (I found a reference to this but the DOD denied the event occured)
• The power went out at Diego Garcia, a super-secret island owned by Britain and leased by the US military, according to the New York Times.
• An unidentified field office of what was then called the US Bureau of Alcohol, Tobacco and Firearms had an unspecified problem with its security system that required guards to be posted outside, according to the New York Times.
• An unnamed building in Omaha, Nebraska suffered a security system flaw that left all its doors open, according to the New York Times.
• Computers at four smaller US airports began displaying a date of 1900, with no apparent operational effect, according to the New York Times.
• A Swedish company's kidney dialysis machine failed to go through its automatic cleaning cycle, a flaw that was reportedly caught before it injured anyone, according to the New York Times.
• A large number of arraignments in Queens, New York courts were delayed because a computer system misdated police complaint forms, according to the New York Times.
• There were unspecified "glitches," reportedly none of them critical, at an unspecified number of nuclear power plants in six US states, according to the New York Times.
• A US video store computer charged a customer a late fee of more than $90,000 when its computer flipped to 1900, according to the New York Times. Another unnamed company had a computer system that began automatically inventing and processing phantom orders, a glitch that was caught before anything shipped, also according to the New York Times.
• The full extent of problems will never be known, because many companies wanted to keep them secret for liability reasons. The New York Times reported that one consulting group that tackled Y2K problems for businesses received about 400 confidential reports of Y2K problems by Jan. 4, 2000.
• Wikipedia had lots of erroneous information posted since January 1, 2000.

How do you attribute and problem to Y2K without weeks of investigation? At work we had some data that was introduced as year "00" and it took months to find where the error was coming from. It turned out to be barcode reader software embedded within another instrument.

Has anyone else noticed like I had that the New York Times reported all of these events?--WaltCip 13:25, 2 October 2006 (UTC)

Y2K? What Y2K?

Sometimes being young and only catching bits and pieces of a potential disaster, is actually quite helpfull. The first time I heard of the Y2K bug was about June 1999. Every member of our family resorted to "fixing" the problem. At the time our home PC was a 1994 i486 running Windows 95. So I decided it wouldn't be such a loss if the system would crash, so I left it running. From 31 December 1999, 22:30 to 1 January 2000, 01:15 I left my Computer on. Absolutley nothing happened. So, in my honest opinion, I think the entire Y2K-bug was a hoax. A simple sales pitch for the big PC companies to make more money. Everyone I know spent lots of cash on 'fixing' their systems, while ours just simply walked into the year 2000. Problem? I doubt it.

But what about the applications (not the operating system?) Depending on what application (word processing, custom written software, ???) you might have been running & depending what it was doing (looking years into the future?), that application could have had problems. If the application didn't deal with dates (many business oriented applications do deal with dates) then clearly there wouldn't be a problem.

No question that a single, stand-alone, home-based PC would be far less likely to have problems. The potential for problems was much greater in businesses where data (often containing dates) is passed around between many applications.

The fact that many vendors took advantage of the fear factor and sold unnecessary "repairs" is unfortunate, but a common aspect of human nature. DEddy 00:38, 13 May 2006 (UTC)

I was working for a major bank on Y2K remediation work during this time and believe me there were issues. Unfortunately we were damned if we did fix the problems (problem, what problem?) and would really have been damned if we didn't.

Just from the testing I carried out on various systems the problems varied from overnight batches flat out not working (so no data, no bank accounts, total chaos), through quality of data issues (values being wrong by millions), interoperability issues (2 systems might both cope with Y2K dates but get confused when communicating) to numerous smaller problems. Fixing all of these issues took months (if not years) of work by tens of thousands of IT professionals.

If we had left it until the problems occurred then they would have taken weeks if not longer to resolve. Would anyone have been happy not having access to their money for days or weeks, and not knowing whether it would be right when they did see it? Apart from that the stock markets and other trading mechanisms would have collapsed leading to a global slump. So yes there really were problems - many of which we had to fix as soon as any Y2K dates turned up in the systems.

It wasn't all doom and gloom - there were other systems that behaved perfectly and had no problems at all. However we still had to go through rigorous testing to confirm they were OK.

And although many IT people earned plenty of money from this time, they also had to give up a huge amount of time including the chance to see the millenium in with a party (and alchohol) and were instead stuck in offices all night.

--Simongv 15:35, 7 July 2006 (UTC)

Y2K was indeed very real. I spent 18 horrible months hacking through badly-written or overly-patched COBOL code in the late nineties to expand all those two-digit years to four digits. To make things worse, this was "empty" expenditure as far as the company was concerned - no new features were being added to the system, so they were spending money just to maintain the status quo.

Before we started coding, we tested how much of a problem Y2K really would be by rolling the system date forward on a test machine and trying to run our applications. They barely functioned at all, even modules that we would not expect to be date-dependent.

We did the job right, though. When the year 2000 rolled around, only two minor faults occurred in a system of over 3000 programs.

The idea that Y2K was a hype probably results from all the rubbish that was talked about "planes falling out of the sky", and the over-promotion of the BIOS problem by people with a financial interest in selling BIOS solutions. The real issue was always about business applications, primarily in the financial sector.

Mr Barndoor 13:47, 11 September 2006 (UTC)

Barndoor. Gee... glad to know someone knows the potential problems from Y2K were in fact real.

Having something of a stake in what Y2K was or wasn't—unfortunately I typically label it as "Lessons NOT Learned" —I'm of the belief that most "civilians" [someone not intimately/painfully familiar with what software applications are like] do actually believe Y2K was a hoax.

This one is my main concern. Y2K was real enough, but the media hyped it up to something it was not. The problem was taken seriously (and then some), and the problems fixed, so Y2K pretty much came and went uneventfully, leading the public to believe it was a hoax. So, when the Y2038 problem, or other problems derived from the common use of arbitrary constraints, come knocking, people are likely to discount it as a rehash of the media hype around Y2K, especially since we can't give them an explanation that is easy to grasp by the decimal minded. Hence a higher likelyhood of people ignoring the problem, and the problem not being fixed.

Y2038 is far more pervasive, and much harder to fix in some regards. And, unlike the Y2K problem, it actually affects the normal way of representing time (or timing) on a Posix system. I recall someone half-jokingly referring to "19 Jan 2038" as "Armageddon". While we have a comparatively larger time frame between awareness of the event and its occurence, many companies I've dealt with are at best reluctant to do anything about this problem before it occurs, citing Y2K as "evidence" that it will be harmless. This, in turn, has led some embittered people who spent excessive amounts of overtime, including New Year's Eve itself, fixing Y2K, to state that they'll not bother next time around.

If we can find some sources, or if this is one of those points that don't really need sourcing, a note should be made that many feel that the media hype around Y2K, and the subsequent public reaction, has significantly damaged the drive to fix the Y2038 problem. As it stands, I don't feel that this article properly distinguishes between the claims made by the professionals in the field, and the media hype, possibly exacerbating the problem. (Using WP to source claims is becoming rather common, even in cases where primary sources are quoted that show WP to be wrong.) Zuiram 22:52, 28 October 2006 (UTC)

My perspective is from having:

(1) been a programmer (primarily in large financials services organizations),

(2) sold software "inventory" tools.

Three interesting points:

(1) in 1994 (just as I was beginning to market a new software inventory discovery tool) I learned that just in IBM mainframe space, no more than 10% of such sites had even purchased a software configuration management tool. I spoke with allegedly professional software people who did not make a disctinction between a library management tool and software configuration management.

(2) sometime in 1996 Capers Jones (see Function Point ... there is a Wiki page, I just can't get the embedded link to work) sent me a list of the 440+ software languages he monitored (point being: there are literally hundreds of software language in active use, not just COBOL, C & Java),

(3) eventually in our Y2K inventory adventures we settled on the rule-of-thumb observation that it was typical to find 50% of the things in "Production Libraries" to be unclaimed junk. Scary. DEddy 15:19, 11 September 2006 (UTC)

Y3K?

What about Y3K? Wouldn't the same "problem" happen in the year 3000?If so start prepping!(joke)Bon Scott 03:59, 29 May 2006 (UTC)

I'd worry more about 2030. A lot of software (including excel) was changed so that any 2 digit date of 30 or lower was assumed to be after the year 2000, and any greater than 30 was assumed to be 19nn. The logic went that "it was OK because these systems won't be around by then..." - pretty much the same words people used about the possibility of a Y2K bug!

--Simongv 15:49, 7 July 2006 (UTC)

No, actually the next problem is in 2038 i.e. the year 2038 problem. So it would be a Y2K38. Voortle 14:39, 3 August 2006 (UTC)

Nope - there's definitely some 2030 problems because that was the development guideline where I was working, though you could be right about 2038 for Excel etc. Y2K work was SO boring that I've tried hard to forget all about it! :-) --Simongv 12:19, 11 August 2006 (UTC)

I was talking last night to a former colleague. He was complaining that his new employers want him to go back to COBOL for the next year or so to help re-do their Y2K work properly. They had taken the approach of leaving the years as two digits and assuming that any year less than 10 has a "20" prefix, while any year of 10 or above has a "19" prefix. They had to go so low because their data has dates of birth as early as 1910. They had expected a replacement system to be in place before problems occurred. The replacement system hasn't happened for various reasons and will take years to develop. They're now getting problems because of the need to calculate future dates. The only real solution is to change to four-digit years. So it ain't over yet! Mr Barndoor 08:52, 6 October 2006 (UTC)

Or they could just change to a native representation of the data. Scarily enough, the Norwegian tax code, it's legal specification, is a Cobol program. I wish people would start migrating these systems to something more maintainable. A time_t of 64 bits pushes this problem about 4 billion years into the future, at which point the only thing it can interfere with are our attempts to cope with the sun going red giant ;) Zuiram 22:56, 28 October 2006 (UTC)

I've come across a bug in an application I use where dates of birth are now being assumed to be "19xx" if 50 or later, "20xx" if 49 or earlier. OK as long as no people on the database are over 56 years old... Yes, y2k problems are still with us and will be until 2050 at least (and will probably reappear in 2100). Many of the fixes put in were of the sticking-plaster variety, the old 2-character year field still lives on.

Exile 16:54, 20 December 2006 (UTC)

Y2K Overview

§ [This is a description that I wrote some years ago.]

In the 70's banks first realized the problem when 30 year loans had glitches. It wasn't until 1995 and 1996 that the federal government began to change their systems. 1997 and 1998 brought in the larger cities and universities.

In 1999 many businesses and government agencies belatedly upgraded their systems only to find that it takes at least 5 years to upgrade such large systems. Through 1999 the problems began to escalate. There was disinformation spread that Y2K would hit like a hammer at the strike of midnight in 2000. Most reasoned experts saw that the real problem would take years to develop. Dick Mills predicted rolling blackouts in the summer of 2000, as the increased use of electricity would tax an already fragile system. When California had it's blackouts, 30% of the utilities were down. This had never happened before.

___________________ From my archive:

Power engineer and Y2K analyst Dick Mills makes the following predictions regarding the availability in the United States of electrical power in the year 2000:

Prepare for blackouts in the first days of January 2000, lasting up to 72 hours. Prepare for shortages of power in the warm summer months of 2000.

Mills predicts that carefully regulated power shortages eventually affecting most U.S electricity customers is much more likely in the year 2000 than uncontrolled localized power blackouts. The odds of power shortages are particularly high during the summer of 2000, especially if the summer is hot in most areas of the United States. Power shortages result when power generation "margins" dip below 0%. Margins of 15-30% are standard for the industry in the United States, meaning that generation capacity is usually 15-30% larger than demand. _____________________

A small digression is needed here. An argument is presented that says that there are always software problems. The counter to this is that all these agencies were forced to upgrade at the same time thus concentrating the problems.

Consider this metaphor. A person is in a room filled to their neck in [excretement]. A brick is thrown at the person's head. The person ducks, lifts his head and declares that the brick never hit him. I say that his soiled face was caused by the brick.

The economy began to dive in March of 2000, way before the 2001 attack. Massive layoffs of tens of thousands occurred every month. 2000 brought the most bankruptcies here and around the world.

In 1998 insurance companies declared that Y2K problems would not be covered. This is one of the reasons that would inhibit disclosure.

There was also concern about the Herstatt risk, liquidity problems and the flight of capital. Our economic system depends a great deal on confidence.

The vast majority of problems were and are financial. There are thousands of businesses that cannot bill correctly and/or have no idea what is their profit or loss.

You can expect another 2 years of economic downturn as the final waves wash across the battered beach.

71.34.249.93 14:20, 6 January 2007 (UTC) spider

______________________________________________________________________

Y2K - 2002 in Review

This is a small sampling of stories that I collected through 2002. I picked those that would withstand the critic's blade. Stories from 2000 and 2001 are too numerous to post.

January 03, 2002 Bank Austria says Austrian ATMs back in operation

A manager at Europay Austria, the company that runs the ATM machines, told Reuters the breakdown was not due to the euro changeover or the large number of ATM transactions.

[These ATM's died on New Years 2002] ________________________________________

February 02, 2002 Carik Floral Wireservice files Chapter 11

January 2001: Carik President Carleen Heckendorf tells Floral Management in an interview that "2000 was a difficult year for Carik" and that "it has taken a great deal longer than it should to locate and repair [Y2K] problems." ________________________________________

February 02, 2002 Teddy bear maker prepares for second attempt at ERP rollout

The company has reason to be cautious. Three years ago, a Y2k-related migration from its homegrown distribution, financial and customer service systems to packaged ERP applications hit a brick wall. Saunders said the problems were severe enough for Russ Berrie to take many of the new applications off-line. ________________________________________

February 16, 2002 Enron mishandled billing

The dizzying pace of Enron Corp.'s deal-making overwhelmed its ability to track simple billing payments and some retail energy clients were overcharged tens of millions of dollars, a problem that could have contributed to the now-bankrupt company's distorted revenue picture, former employees said.

One of the biggest clients, insurance giant Kaiser Permanente, paid $30 million more than it should have because Enron's computer system was so scattered and poorly managed,

"We . . . found millions of dollars in billing errors,"

In November 2000, as billing problems mounted, Kaiser began complaining. ________________________________________

March 06, 2002 MI - Payroll glitches irritate union

There are about 100 vacant payroll audit positions, city officials acknowledge. Furthermore, the city uses a 25- year-old payroll system. In 2000, police sued the city for ongoing problems with the biweekly paychecks. The state also fined the city $1,200 for failing to turn over time records. ________________________________________

March 07, 2002 Owens Corning drops Y2K lawsuit

Owens Corning, based here, will drop a two-year-old lawsuit seeking to recover expenses associated with its Y2K computer conversions. ________________________________________ March 13, 2002 CN - Y2K computer slip-up sends double car tax bills to 22,000

HARTFORD, Conn. (AP) _ A Y2K computer glitch in the state Department of Motor Vehicles led to some 22,000 people being double-billed for motor vehicle taxes. ________________________________________ March 31, 2002 FL - Faulty Florida system sent benefits to inmates, the deceased

In a report released this month, auditor general William O. Monroe said the unemployment compensation agency had not corrected problems identified in a previous audit. In fiscal 2000, the agency issued 278 payments totaling $104,691 to persons shown as deceased by the Office of Vital Statistics, the auditors said. ________________________________________ March 31, 2002 HI - Inventory miscount causes $5.8 million overage at Honolulu utility

HONOLULU -- Auditors looking at the city's finances recently found that two sewer valve repair kits valued at $290 each were mistakenly counted nearly 10,000 times each over the past three years.

The discrepancy caused the sewer fund's inventory balance to be inflated by about $5.8 million. ________________________________________ April 12, 2002 Nigeria - For Computers, a Time to Remove the Layers of Mystery

THE orchestrated campaign to remedy the so called year 2000 (Y2K) problem hit Nigeria like a tornado from the seas. ________________________________________ April 13, 2002 VA - Transit troubles blamed on outdated accounting

The transit agency is struggling to manage a $50 million budget with a 25-year-old computer accounting system that doesn't provide current financial data and requires some manual work. In addition, the agency's longtime chief financial officer, the only one who knew the system well, retired more than a year ago. ________________________________________ April 16, 2002 Let's stop wasting $78 billion a year

Greg Seyk, newly appointed CIO of VisionQuest, had only months to rid his organization of its Y2K bugs.

With just two months to go before the clock struck midnight on Dec. 31, 1999, Seyk didn't have time to deploy the upgrade, even though the payment prioritization function had been a critical selling point for the 53-year-old CIO. "We had to implement accounts payable, the general ledger, payroll and human resources to make sure they were Y2K compliant. It was no small feat," says Seyk, who is also a vice president of the private company. ________________________________________ April 17, 2002 PA - Computer glitches still not solved

O'Hare said the county should get new software because the current system - which was installed in late 1999 and updated last year - is unworkable. ________________________________________ April 17, 2002 NV - CLARK COUNTY SCHOOL DISTRICT: $22.3 million program 'just a pain'

Elementary schools switched to the SASIxp system in 2000-01. Data entered that school year resulted in errors in the newly released accountability report ________________________________________ May 02, 2002 MO - Jackson County tax assessments flawed

Jackson County cannot guarantee the accuracy of its $7.3 billion in real-estate and personal- property tax assessments because of its flawed system, the county's former acting assessment director said Monday.

[Earlier stories explained that this was a Y2K upgrade] ________________________________________ May 06, 2002 FL - Brevard waiting for cities overcompensated for traffic fines to pay up

The municipalities owe the clerk money because of a computer glitch that lasted more than two years. ________________________________________ May 21, 2002 Nestle's ERP Odyssey

Development work began in July 1998. The deadline for four of the modules was Y2K. The new systems would have to double as code fixes and be in place for the millennial change. Nestlé USA made the deadline. But its haste created almost as many problems as it solved.

By the beginning of 2000, the rollout had collapsed into chaos.

The time constraints necessitated by Y2K had put too much pressure on the people in charge of executing the changes. ________________________________________ May 22, 2002 MI - Detroit behind on back pay

DETROIT -- Pay due to Detroit city workers going back as far as three years won't be delivered to all those owed before year's end, city officials said Tuesday. ________________________________________ June 02, 2002 Port of Seattle v. Lexington Insurance

The State of Washington's Court of Appeals affirmed dismissal of the Port of Seattle's suit for coverage of Y2K costs on May 28. ________________________________________ June 10, 2002 MD - Balto. Co. police late on crime statistics

Bill Toohey, a police spokesman, said the delay started in late 1999 when the department bought new computer software to replace software that was not Y2K compliant. ________________________________________ June 24, 2002 FL - Commissioner silent on software

Included in that purchase was software for budget preparation tasks that never worked, which was only one of a series of problems with computer software purchased in 1999. ________________________________________ June 27, 2002 KS - Municipal Court judge resigns

He wrote that the change in computer software in 1999 "was a disaster." Municipal Court was "reluctant to change computer systems in the first place" because the court system worked with more than 600,000 files. But he, along with other Municipal Court officials, were told the change was necessary to be Y2K compliant. ________________________________________ July 01, 2002 Xerox admits to huge accounting error

Company's restatement says $6.4 billion was improperly reported over 5 years ________________________________________ July 05, 2002 FL - City upgrades locks, cameras

TITUSVILLE -- The Y2K bug seems like history now, but its ramifications are still felt at the Titusville police station more than 21/2 years after the new millennium began.

The computer bug that hit on Jan. 1, 2000 knocked out the police building's 15-year-old electronic key card system. So employees use metal keys to get in, and the receptionist issues passes to visitors and buzzes them through the locked door. ________________________________________ July 05, 2002 Pac Bell to Pay Fine for DSL Problem

Since 1999, about 800 customers have complained to the state Public Utilities Commission that Pac Bell billed them - sometimes twice - for DSL services they didn't order, didn't receive or wanted to drop. Under the settlement, the company acknowledged incorrectly billing between 30,000 and 70,000 customers. ________________________________________ July 23, 2002 WorldCom's troubles could become problem for customers

"It's really rare, really rare" that a customer's billing statements reflect the negotiated rates, she said: about 90 percent of her client's billing statements have mistakes. ABC audits heaps of clients' bills on a regular basis.

Packer said billing is his biggest problem with WorldCom. It takes up to three months to fix an error, he said. ________________________________________

July 26, 2002 GE - Properties will be reassessed

Assessed values have failed to increase since 1999 partly because of outdated and non-Y2K-compliant computer software in the tax assessor's office, Mr. Reece said. ________________________________________ July 31, 2002 Firm run by Massport CEO failed to thrive

HR Logic was hurt by the purchase of a larger and problem- ridden competitor in 1999 and the failure to efficiently upgrade and integrate its computer systems in 2001.

But the software crashed the system, according to several former managers, wiping out crucial data that had to be retyped and forcing company employees to temporarily write paper checks for the thousands of employees it serviced. ________________________________________ August 02, 2002 $800,000 Y2K Lawsuit

Documents which until recently were sealed by the federal court in N.D. Iowa, reveal that Keane Inc. paid a total of $800,000 to six hospital plaintiffs to settle lawsuits over alleged Y2K problems in its predecessor's (Source Data Systems) software. ________________________________________ August 9, 2002 FL - Glitch misses drivers' offenses

Some bad drivers may have gotten a free ride because of a Pinellas computer glitch that kept the outcome of 70,000 traffic cases from the eyes of state driver's license officials.

The problem was that pre-2000 data could not be sent on computers reprogrammed to avert Y2K bugs. ________________________________________ August 9, 2002 FL - Glitch misses drivers' offenses

The problem was that pre-2000 data could not be sent on computers reprogrammed to avert Y2K bugs. ________________________________________ August 10, 2002 CO - Detroit Finance System Sputters after Overhaul

"DRMS is not a boondoggle, it's just a very high-cost system that still has deficiencies,"

[An earlier story tells of a Y2K upgrade] ________________________________________ August 17, 2002 NY - A sad fact of regional life

The history of the phones hasn't been one happy chapter after another. They broke down in the summer of 1999, and repairs were difficult and parts elusive. But repaired they were, only to collapse again as one of the rare victims of the dreaded Y2K computer changeover. ________________________________________ August 25, 2002 GE - Assessing the damage

Tax officials have said the property values escalated so much this year mainly because of computer equipment that was not Y2K compliant ________________________________________ September 07, 2002. OR - Water Bureau hiring to fix billing

The city bought the system from Houston-based Severn Trent in 1997 with an eye toward having it in place well before the Y2K turnover. Implementation was delayed many times. Then, when the system finally went live in February 2000, problems surfaced with approximately 40,000 accounts. ________________________________________ September 18, 2002 WA - County spends $4M on computer upgrade

The county has no plans to restart efforts to update and merge its two major financial systems. Faced with mounting costs and delays, the county froze the project in 2000, having spent more than $40 million but still failing to get it off the ground. ________________________________________ October 14, 2002 MD - Stats reveal violent crimes up to highest in four years

The 2001 figures were supposed to be released months ago but were delayed due to a 2000 computer system that slowed data entry, police spokesman Bill Toohey said.

"We never recovered from Y2K," Toohey said. ________________________________________ October 14, 2002 FL - Computer gives 186 inmates extra bucks

It started in January 2000, when the inmate banking system went awry. ________________________________________ October 16, 2002. FL - Brevard commissioners rip clerk of court's upkeep

He inherited a glitchy computer system when he was elected in 2000 that was supposed to be faster, better and Y2K compliant. ________________________________________ October 20, 2002 TX - Hospital software hits a snag

The sluggish $75 million push to upgrade the Harris County Hospital District's outdated computer network has hit another snag -- new software to improve patient processing and billing won't be ready until June.

The Per-Se problem is just the latest glitch in a district computer upgrade that started in 1999. ________________________________________ November 06, 2002 GE - Paying for the past

Information Systems and Networks Corp. [ISN], a company with which the city is likely to soon settle a high-profile, Y2K-related lawsuit for $1 million. ________________________________________ November 09, 2002 CO - GarCo treasurer candidates clash over '99 audit problems

"We were in the process of preparing for Y2K and had just started using new computer software,"

the treasurer's six-month report was overstated by more than $1.5 million ________________________________________ November 10, 2002 MI - Water billing sytem to be replaced in massive technology overhaul

The problems in water billing, which first surfaced in 1999 shortly after installation, included late and inaccurate bills and estimated bills.

The council approved the purchases to get the city Y2K compliant ________________________________________ November 14, 2002 MT - State must scrap failing DOR computer system

What began as an honest mistake of signing off prematurely for the new computer system in 1999 (under the pressure of Y2K compliance), then became a runaway software disaster, and has now evolved into departmental cover-up as it fails to be able to conduct its primary mission: accurately collect, account for and audit Montana's taxes. ________________________________________ December 13, 2002 NY - State moves to restore emergency phones on Adirondack highway

The roughly 70-mile stretch of Interstate 87, the Adirondack Northway, has been without roadside emergency phone service since April, a remnant of Y2K computer glitches.

spider ___________________________________________________________________________

Accounting problems

Companies that restated their 1999-2001 losses, investigated by the SEC or had bookkeeping/accounting errors in this same period.

Kmart Swisslife MyTravel Flagstaff Unified School District. Xerox Quest Cutter & Buck Bristol-Meyers Mirant Rock-Tenn Massachusetts General Hospital Washington State Democratic Party U.S. government Merck Reliant Flow International NH - Monadnock Regional School District Rent-to-Own Network Associates Nvidia UK - Automotive Precision One.Tel Robotic Vision State of Massachusetts Pump Maker Performance Food Minuteman International Inc. AU - Department of Land & Water Conservation and the Rural Assistance Homestore MI - HARTLAND TOWNSHIP GenCorp State of Illinois Take-Two PG&E Measurement Specialties Inc. Computer Associates International Inc. Sun Microsystems IBM Hayes Lemmerz International Inc. PNC Financial Services Group Inc. Near North National Insurance Brokerage Inc. Enron World.com Raining Data Almost Family Inc. Oregon State University WA - Division of Developmental Disability Guardian iT Baltimore Technologies Japan WHIRLPOOL Anadarko warns Pennzoil-Quaker Electric City Tyco International Cendant PNC Bank General Electric State of Delaware Waste Management Sunbeam Rite Aid Lucent Forest Service Hayes Sykes Borden Chemicals and Plastics Government of Fiji Chronimed Cole Meyers Department 56 ConAgra Foods Inc. AOL Time Warner Inc. FL - city of Brooksville MO - Johnson County Agriculture Department GarCo Brazoria SC - Division of Motor Vehicles IRS UK - Schroders SA - Siltek Dollar General MO - Jackson County CMGI WA - King County Mattson Safety-Kleen AT&T

spider

=

Not March 2000, the turn down of the DJIA was January 14, 2000. At this writing, Aug 26, 2006, the DJIA is still below the high of 11,722. Factor in the inflation and it's still way down. Other true facts - NSA's internal computer system failed in January 2000 and had to be rebuilt from the ground up. Many large companies lost track of their Profit and Loss and failed. The economic losses were enormous, exactly as predicted but never connected to Y2K computer problems.

Y2K Marketing

Does anyone else remember products like universal remotes, computer cables, and wristwatches(the kind without the date) having 2yk compliant stickers on them?

I remember that. I saw a bunch of ridiculous Y2K compliant stickers on things that didn't even store dates. Of course, those stickers were nonsense, as nothing could possibly happen to them as they didn't even store the date. Voortle 14:37, 3 August 2006 (UTC)

Article title

The title of the article should reflect the popular term used; shouldn't it? I believe Y2K bug would be more appropriate as it's used more often; or something other than what's used now.

It was not a bug. In the context of using 2 digit years in software systems, it was an intentionally accepted design decision. The other stuff (end of the world, 2nd coming, etc.) was "just" out-of-control hype. DEddy 22:33, 30 June 2006 (UTC)

No, I'd say that in most cases it was a bug. People just used 2 digit years because thats what others had, thats how the system already worked or it was easier than typing all four and often because it saved storage space. Very rarely would someone have thought "I'll use 2 digits to save space, but it does mean this system won't work after 1999". --Simongv 16:12, 7 July 2006 (UTC)

Real Time vs Date/Time Systems

Not sure where this would fit in but it's an important distinction.

The reason that most safety critical software (Air Traffic Control, Nuclear Reactor software, traffic lights, lifts etc) didn't experience problems is that they are based on real time software. This means they don't care (or even know in most cases) what the current date is, they just know that in n seconds they need to take an action and track when n seconds has passed.

Which is where the Y2038 problem and such come in; hopefully, we'll be able to convince people that it isn't just a rehash of the media hype around Y2K. Notably, though, some of this software uses sub-second timing intervals, and are designed to deal with wraparound. Zuiram 23:02, 28 October 2006 (UTC)

On the other hand business, and especially banking & insurance, software deals in fixed dates (opening dates, termination dates, payment dates etc) and so have to know what the current date is and be able to accurately carry out date based calculations.

Probably needs some cleanup and more examples, will look into it if I get time before anyone else does.

--Simongv 15:47, 7 July 2006 (UTC)

Background/citation needed

ok... what qualifies for a "citation"... an image of the original email? DEddy 02:46, 25 July 2006 (UTC)

Who...?

Who actually believed that? Load of BS, if you ask me. I never believed it. Although the problem made for a good episode of Family Guy. --66.218.17.94 04:43, 3 August 2006 (UTC)

Good to see you've looked into it carefully. Imagine a program that stores the year you were born and the current year, both in two digits, and substracts your year of birth from the current year for your age. Here's the output...

YearOfBirth YearNow YourAge

 85       98     13
 85       99     14
 85       00    -85
 85       01    -84

...so as the year crossed into 2000 you become a negative age. This may seem trivial, but if it was in a system used to calculate how long a bond has been outstanding and so what its value is it gives totally incorrect financial figures. If the figure from the calculation isn't output directly but instead is then processed through another 10 or 20 steps afterwards you end up with a totally incorrect result which is very hard to track down without going through each step manually.

--Simongv 12:16, 11 August 2006 (UTC)

Artwork

It would be interesting to get a cite on the two-digit artwork dates. --Daniel C. Boyer 15:17, 12 October 2006 (UTC)

Hijacking?

It looks like someone hijacked and trolled the introduction to the article. I'll try fixing it, I guess. 72.225.142.168 22:08, 3 December 2006 (UTC)

Rumours

These are unsourced, please don't read them

Rumors

These are unreferenced anecdotes and urban legends.

• One theory has it that the Federal Reserve increased the money supply in 1999 to compensate for anticipated hoarding by a frightened populace. The populace, however, was not frightened, and the flood of new money fueled a stock market high tide that went out on January 14, 2000 when the Dow Jones Industrials fell from the all-time peak.
• Speculatively, the Y2K spending on information infrastructure caused a slowdown in information technology spending in 2000 and 2001 and may eventually lead to higher productivity in future years.
• Univision news reported that on the evening of December 31, 1999, a couple in Peru had committed suicide, for fears of what Y2K would bring.
• A few (but not many) computer systems did actually fail on January 1, although some of those did so on a yearly basis. An almost amusing postscript to the Y2K problem was the fact that a number of computers not set up for leap years actually failed the following February 29.
• Contrary to widespread warnings that personal computers should be powered down during the Y2K moment, some computers (e.g. those running Microsoft's Windows 95 operating system) handled the change properly if they were running, but needed to have the time and date reset if they were not.
• Items rented from Blockbuster prior to January 1, 2000 and returned after January 1st were reportedly marked for astronomical late fees ($91250), as though the items were 100 years overdue.
• In Germany the coordination system of the fire brigade in Berlin showed the error "named pipe closed", but the system kept working. A reboot of the affected machines rendered the system unusable. Emergency coordination had to be done via mobile phones.
• In the United States, it is purported that a series of callboxes failed along Interstate 87/the Adirondack Northway in New York [1]
• In Windsor, Ontario, the Variable Message Sign at the Cleary Auditorium above its driveway posted the date and time as "January 1, 1900". This was fixed by the end of the day.
• In Argentina, people received government tax collection notices dated for the year 19100, as well as 00
• In the months leading up to 2000, there were numerous reports, some substantiated, others not, of people withdrawing significant amounts of money and burying it in their backyards or elsewhere, due to the fears of massive failures at financial institutions. Many of the cases saw the money dug up and stolen.[2]
• Apple Computer's website was edited at the stroke of midnight to change the "19" prefix to "20", and displayed the date as January 1, 20100 above a prominent Y2K statement.

Y2K

Shouldn't this article be at "Y2K problem"? WP:COMMONNAME says yes. Seriously, does anybody actually think "Year 2000 problem" is the more commonly used name for this subject? 71.203.209.0 10:21, 24 February 2007 (UTC)

Was the cost justified/Opposing View and citation to eSchoolNews

In the section talking about the 'oposing view' to whether the cost was justified, it says "By September 1, 1999 only 28 percent of US schools had achieved compliance for mission critical systems, and a government report predicted that "Y2K failures could very well plague the computers used by schools to manage payrolls, student records, online curricula, and building safety systems" and gives citation 28, which is a link to an eSchoolnews.com article.

However that article actually says, "Citing a Department of Education survey of more than 3,500 school districts and local education agencies, the report says only 28 percent indicated that their mission-critical systems are Y2K compliant. That number is expected to jump to 72 percent by Oct. 1 and 98 percent by Jan. 1. "

Surely this citation and reference is only valid if you can determine how many schools did achieve Y2K compliance by 1-Jan 2000? If the number did rise to 98% that might suggest this citation is being taken out of context? —Preceding unsigned comment added by 77.99.164.88 (talk) 02:40, 3 May 2009 (UTC)

Opposing/Support paragraphs

This section is not written much like a encyclopedia, i suggest a rewrite and change of headings to: Costs involved, opposition and support Kernel geek (talk) 18:36, 13 January 2009 (UTC)

Photograph requested

Someone has placed the photo requested template above. What kind of picture would we use for an article on Y2K? A picture of a computer not knowing what the date is? Voortle 20:44, 18 March 2007 (UTC)

A screenshot of an error message? 142.59.172.187 11:14, 10 June 2007 (UTC)

There must something that could be added - a large article without a single picture isn't normally desirable. Richard001 04:26, 29 July 2007 (UTC)

I put together a quick illustration. Intended for aesthetics only. — xDanielx ^T/_C\^R 11:12, 10 December 2007 (UTC)

The picture of the gasping man, a hourglass, a computer and a bug? Why is it worthy of this article?116.15.90.76 (talk) 18:30, 28 December 2007 (UTC)

The purpose was aesthetic, not explanatory. I figured there wasn't much need for an explanatory image, but if you can come up with one, feel free to replace the current image. — xDanielx ^T/_C\^R 03:39, 29 December 2007 (UTC)

Complacency

In 1989 I was working on a computer system destined for use by the UK government. My role was as an independent quality consultant. The system represented dates internally in two-digit format (despite being programmed in C), and would fail if the system date was changed to anything after 2000. I repeatedly raised this point, much to the annoyance of the project manager, who even threatened to have me removed from the project! His argument? The system was warranted for ten years, and the developing company would have no liability when it failed.

The system, as installed, did continue to work after the critical date, so evidently some of the investment in tracking down and fixing Y2K bugs bore some fruit! I was no longer involved, and the company that had written the original software went bust in 1990. --King Hildebrand 16:58, 20 April 2007 (UTC)

Jargon

This article is waaaaay too technical. How are lay people unfamiliar with computer languages, etc expected to understand this.

I'll cut it down then. 76.27.163.35 17:10, 1 June 2007 (UTC)

Since Y2K was a technical problem, this is unavoidable. Mr Barndoor 11:12, 8 June 2007 (UTC)

       ^ I agree fully. —Preceding unsigned comment added by 76.165.238.151 (talk) 21:33, 14 April 2009 (UTC)

One Question

"It caused widespread concern that critical industries (such as electricity or finance) and government functions would cease operating at exactly midnight, January 1, 2000,..." What time zone was this supposed to happen? --68.183.43.30 02:03, 31 May 2007 (UTC)

Whichever time zone the computer was using. The failures occurred (and the prevented ones would have occurred) on a system-by-system basis, not all at once. Mr Barndoor 14:57, 31 May 2007 (UTC)

US Government Y2K Information Center

I suspect this falls under "original research" or "personal recollection", but I've not seen much documentation of the actual US operational facilities.

I was the external network architecture consultant to the Y2K Information Center, the operating agency under LTG (Ret) Peter Kind, which reported to John Koskinen in the Executive Office of the President. The Center's headquarters were at 1800 G St. NW, a few blocks from the White House, and had backup facilities in FEMA headquarters and a hardened site outside the immediate area. Online access by the public and most government agencies was hosted at two AT&T data centers, one in New York and the other in San Diego. Then under the Department of Justice, the Critical Infrastructure Protection Group was started in the Center offices. The facility was equipped to receive and disseminate reports at all security classification levels, although the great bulk of information was readily available.

As a bit of trivia, we did have an informal motto on a fair number of bulletin boards, "The Dark Ages were caused by the Y1K Problem." Hcberkowitz 03:00, 25 June 2007 (UTC)

Digimon in popular culture?

The article states "In the 2000 anime film, Digimon: Our War Game, the villainous Diaboromon is a personification of the Y2K bug. In an exaggeration of some expected events, he cuts off the phone networks and launches a nuclear missile at Tokyo." However, while the virus did indeed do this, I can't recall it being mentioned anywhere that the virus was related in any way to the Y2K, nor do we have any way of telling if it's the year 1999/2000 in the middle segment of the movie (which is divided in 3 segments across 3 different years). Should I remove it? Jaimeastorga2000 00:45, 15 August 2007 (UTC)

Exploding microwave oven anecdote

I recall a story, usually attributed to a name, of a guy who set the clock in a microwave oven past the year 2000 and it exploded. If this can be sourced, it seems to be a notable anecdote, with indication as to its veracity. We need only indicate that the anecdote existed. --Scottandrewhutchins 20:51, 13 September 2007 (UTC)

I have never seen a microwave, let alone any small appliance with a clock, ask for a year to be set. Even if one did, I find it highly unlikely that the input of an invalid string would cause physical damage - that could be compared to typing a date in the year 10000 into Excel and having your computer blow up.

Jb17kx (talk) 10:38, 18 November 2007 (UTC)

Wierd info

says some guy in iowa was charged video overdue 100 years cause of computer and penn. kid was charged overdue book.. please validate. 130.13.47.58 04:31, 1 November 2007 (UTC)guybethename??

Typical examples of the kind of bug anticipated are the rare occasions when a someone in the 105-6 years' old bracket gets a primary school induction notice, because the computers handling the population register worked to the least significant 2 decimal places. Another intended objective was to review the general coding of older systems, which eliminated a number of other bugs from, typîcally, the kind of linear systems predating object modelling. —Preceding unsigned comment added by 81.241.227.84 (talk) 13:22, 8 December 2009 (UTC)

(un) Documented errors

The list of unsourced incidents is getting larger...¿does anyone feels it's time to remove all unsourced incidents? (like "In the Solomon Islands, a ferry ticket was rejected for being 100 years out of date.")Seba5618 (talk) 03:02, 20 November 2007 (UTC)

Confusion - Literal vs. Popular Millenium

What does "It was also named the Millennium Bug because it was associated with the (popular, rather than literal) roll-over of the millennium." mean? What does a "popular millenium" and a "literal millenium" mean? --RaphaelBriand 16:57, 1 December 2007 (UTC)

Properly, the new millennium started on January 1st 2001 - this would be what is meant by the "literal millennium". 2000 was the last year of the old millennium. But most people treated January 1st 2000 as the start of the new millennium, because the first digit of the year changed from "1" to "2" - hence this would be what is meant by the "popular millennium". Mr Barndoor 11:08, 3 December 2007 (UTC)

Thanks --RaphaelBriand 20:05, 3 December 2007 (UTC)

I disagree on this "(popular, rather than literal)", if it is possible to disagree on such a thing. I mean it's poorly worded don't you think? Step through the definitions of each word. What you're effectively saying is that a thousand years would have passed by the start of 2001, but not by the start of 2000 - and we know this to be mathematically incorrect because time did not start in year 1001 or year 1000, but in fact far earlier than either of these dates. Or have I missed something here? Have we suddenly disproven the existence of God, and the dinosaurs? And I missed it!? Wow, the Earth is only 1006 years old, I would have never suspected this. The popular representation, where December 31st 1999 23:59:59 as the end of a period of a thousand years, is in fact literally correct, and no more or less so that the idea of the same one year later. What the person who wrote this is effectively saying is that anybody who chooses a cardinal system of numbering for years is doing it incorrectly, and that the method is invalid. I think there's a strong argument, if we must say that one is incorrect and not the other, for saying that the UNpopular method is not the literal method. As i said above, BOTH methods are in fact literal methods of numbering, but I would like to see some reason for for the implied invalidity of counting from zero? ? ? ? ? 125.236.211.165 05:36, 4 December 2007 (UTC)

In the Western calendar system, we assume a starting year of 1, not 0. The Millennium article gives a clear explanation of the popular versus literal millennium. Mr Barndoor 11:27, 4 December 2007 (UTC)

The majority popular approach was to treat the end of 1999 as the end of a millennium, and to hold millennium celebrations at midnight through December 31, 1999 to January 1, 2000, as per viewpoint 2. The cultural and psychological significance of the events listed above combined to cause celebrations to be observed one year earlier than the formal Gregorian date. This does not, of course, establish that insistence on the formal Gregorian date is "incorrect", though some view it as pedantic (as in the comment of Douglas Adams mentioned below).

Uh, yeah. It mentions the popular one but it currently omits the literal one, or at least by name. What it does do is point out that a millennium is a millennium no matter which day you begin measuring from. 125.236.211.165 (talk) 00:19, 24 December 2007 (UTC)

The International Standardisation Organisation has recently clarified its date numbering system that the year before 1 was the year 0, before that -1 etc., so those who celebrated the millenium on 1 January 2000 were right after all. PatGallacher (talk) 13:16, 8 October 2008 (UTC)

Okay to scrap general references?

We have a solid pool of 26 inline citations (not counting duplicates), mostly from very reputable sources. I think general references (by that I mean, references without inline citations) are better than nothing in that they add to the credibility of the article and provide a starting point for further research, but with a strong assortment of inline citations I don't think the general refs really have much use. Would anyone object to nuking them? — xDanielx ^T/_C\^R 08:04, 10 December 2007 (UTC)

Guess I'll be bold and go ahead with it. Open to post-action review if anyone disagrees, of course. :) — xDanielx ^T/_C\^R 08:59, 13 December 2007 (UTC)

War of the worlds

I find that the inclusion of "War of the worlds" is of poor relevance, and may formulate an opinion (comparing the Y2K reaction to the hysteria following this show's broadcasting?). The only link seems to be that an english or us tv catastrophe show about the y2k bug was showed with a disclaimer to avoid a "war of the worlds" effect (see war of the worlds article). Anyway if none objects i'd delete it.82.224.109.80 (talk) 21:51, 26 January 2008 (UTC)

Microsoft Excel Date/Time Storage

I believe this statement is incorrect:

The popular spreadsheet Microsoft Excel stores a date as a number of day since an origin (often erroneously called a Julian date). A "Julian date" stored in a 16-bit integer will overflow after 65,536 days (approximately 179 years).

Current versions (and certainly version since Excel 97 i.e. predating Y2K) use a (32 bit) floating point number to repesent a date, whereby the integer part of the number represents the number of days since an arbitary start date which is actually the zeroth of January 1900. Using the floating point format naturally allows the system to accomodate time as the fraction of a day resolved down to the accuracy of the floating point representation of the number. It also allows date difference to be obtained simply using the subtraction operator and simplifies date comparison.Fizzackerly (talk) 13:52, 25 March 2008 (UTC)

Obviously, it also allows dates significantly further into the future than 179 years after 1900 to be represented!Fizzackerly (talk) 14:05, 25 March 2008 (UTC)

Y2K The Movie

Why is there nothing about this movie on here? [3] I find it a bit strange, given the warning that were put at the start of this film so not to panic people. Katana Geldar 11:38, 31 May 2008 (UTC)

Origin of the term

Are we sure about that story? I seem to remember references specifically to Y2K in the Movie Strange Days. It was released in the fall of 1995, so was probably in post production by the time the term was created. Anyone has insight on this? Observer31 (talk) 00:11, 12 August 2008 (UTC)

I swear that I've not heard of movie "Strange Days" & certainly didn't see it... I was way too busy. It was simply a lazy abbreviation "year 2000" is 9 keystrokes, while Y2K is just 3. Nothing more. DEddy (talk) 01:02, 12 August 2008 (UTC)

I am not 100% certain my affirmation, and until I can find a copy of the film and confirm I'm not going to change the article. If someone else has it and can look it up it would be good. I'm not trying to make a point, just be accurate :) Observer31 (talk) 04:17, 13 August 2008 (UTC)

You'll then have to analyze if there was cross fertilization between the movie & Peter de Jager's Year 2000 forum. I rather doubt it. I didn't know I'd coined the term in the year 2000 technical community until sometime in 1998, a good 3 years after I accidently coined it in 1995. DEddy (talk) 13:12, 13 August 2008 (UTC)

I've done some searching and found a few links. http://www.thedigitalbits.com/reviews/strangedays.html , and http://64.233.169.104/search?q=cache:yqxCCgqoTeMJ:www.infoshop.org/sf/Strange_Days_(film)+strange+days+y2k&hl=en&ct=clnk&cd=42&gl=us which seem to indicate that the movie did indeed contain the term Y2K. Observer31 (talk) 21:50, 13 August 2008 (UTC)

I didn't see anything in those TERTIARY sources that indicates they coined the term. A PR piece written after the fact is NOT a primary source. If you can dig up the original screenplay then I might be willing to reconsider. Do note that this wiki page is umpty-ump years old & the folks who own the movie have never stepped forward to claim origination rights. Also... there is the ever so slight <sarcasm> point that Y2K means the software challenge, NOT the end-of-the-world hype that the media blew it into.

I'd be more than willing to post a pdf image of the my email but I don't know how to. DEddy (talk) 00:37, 14 August 2008 (UTC)

I am well aware that these are not primary sources, which is why I used the wording "seem to indicate". I'm not going to change the article until I have solid evidence. Observer31 (talk) 04:15, 19 August 2008 (UTC)

99 years & 9/9/99

{{editsemiprotected}} I've worked on financial systems in Britain & Australia using COBOL as a programmer with over 25 years experience.

COBOL was used extensively in big business and many millions of programs had been written in this language dealing with various aspects of billing. 99 years would have been generated from date calculations using 2000 - 1999 because the result was stored in an unsigned field (no minus sign) thus 00 - 99 = -99, which in an unsigned field becomes 99. Therefore, billions of customers would've been billed for 99 years!

I fail to see how anyone in computing would think 9/9/99 would relate to 9999. I've never heard of this; so it must have been some stupid rumour. Dates were usually stored in yymmdd format and converted to the required format. Thus, 9/9/99 was actually stored as 990909; a far cry from 9999.

Malcolm Hill

The article presents the 9/9/99 problem as a theoretical concern, not as an actual occurrence or as a focus of remedial work. I think it's fair enough to let it stand as an example of the thinking that was going on around Y2K and other potentially problematic dates. Mr Barndoor (talk) 11:02, 2 September 2008 (UTC)

The editsemiprotected request doesn't have a clearly and agreed-upon change to make. Please provide or describe a specific change to make. —EncMstr (talk) 00:24, 13 September 2008 (UTC)

This is my first time in attempting to make a change, so please allow for my lack of protocol knowledge. I've made a few changes to the above. The article doesn't spell out that COBOL was used extensively in big business and that if the problem wasn't fixed, the business world would've collapsed under the weight of customer complaints. As I've never heard of the so called 9999 problem, I was trying to show that if anyone had heard of this, it was totally erroneous. I'm trying to set the record straight for future posterity as even today, people think it just a money making exercise. Malcolm Hill Molbrum (talk) 09:07, 16 September 2008 (UTC)

There were programs that used "99/99/99" or "9/9/99" and other values as a special magical dates that meant things like "never expires" or "not known". In fact, there still are such programs. In general, the impact of accidentally treating a date as magical would not matter much. Regards, Ben Aveling 06:50, 22 February 2009 (UTC)

Arbitrary Trivia

ZimZalaBim... thank you for stepping up & deleting that irrelevant noise. DEddy (talk) 20:21, 15 December 2008 (UTC)

Microsoft Zune 30GB Bitten by Leap Year Bug

There are stories that Microsoft's 30GB Zune MP3 player has been bitten by a leap year bug. Shades of Y2K.

I predict there will be additional leap year bug stories. DEddy (talk) 15:52, 2 January 2009 (UTC)

The first sentence

I think the term bug is confused with virus, and should be defined in the first sentence. Also, maybe the word should be plural because the problem was in more than one program. I read the article to see if the text message term Y2K was invented to wonder why Dec 25 wasn’t January 1 and why it got associated with the feared virus Coast to Coast AM was talking about in in the late '90s. From clicking the link to "computer bug" I was reminded bug does not mean virus. So, I think the first sentence should define bug, maybe in parentheses. --Chuck (talk) 03:35, 11 January 2009 (UTC)

Replies: (a) The very fact there is a wikilink to computer bug in the opening sentence, providing you the definition you seek, solves your concern. That is the very purpose of the link. (b) It should remain singular since the "problem" itself is a singular phenomenon (albeit with multiple components). --ZimZalaBim ^talk 20:13, 12 January 2009 (UTC)

Documented Errors

The last two entries in this section appear to a misunderstanding of the PERL programming language (perhaps others as well) rather than actual Y2K bugs. The perl localtime() function returns the number of years since 1900. Pre-Y2K this looks like it simply returns a short year (99 vs 1999) and is thus not Y2K compliant, however, it *is* a Y2K compliant function. The correct way to display the results obtained from localtime() is then "$YEAR + 1900" but many programmers assumed that localtime() returned the short year and as such wrote code as "19" + $YEAR resulting in the '19100' displayed in the main article. To state that these entries are 'Y2K-bug' related is in error, they are not caused by a Y2K-bug but rather a programmer failing to read and understand a well documented function.

Please consider removing the references.

Milledel (talk) 03:10, 22 January 2009 (UTC)

The fact that the coder needs to write "$YEAR + 1900" is a bug in perl - one that is inherited from C of course. The need to write "$MONTH + 1" is also a bug. Regards, Ben Aveling 06:37, 22 February 2009 (UTC)

I would disagree. Perl's localtime() uses C's localtime(), this is true. But if you look at the definitions of the structures that are used in C, they are clearly defined (in the glibc documentation):

    int tm_mon;    /* months since January (0 to 11) */
    int tm_year;   /* years since 1900 */

Now, you may not like that definition but it is most certainly NOT a bug in the C language, nor a bug in Perl. Nor is "The fact that the coder needs to write "$YEAR + 1900" " a bug in perl. A programmer has to actually understand the functions that he's dealing with and what values they return -- there is no bug in localtime(). —Preceding unsigned comment added by 65.40.72.134 (talk) 04:54, 3 April 2009 (UTC)

65.40.72.134 (talk) 04:55, 3 April 2009 (UTC)

Published Articles citing Y2K Bugs caught in advance

Jeff Furman, along with his team of Change Management Support Specialists at a leading brokerage house in New York City in the early 90’s, published several early articles in 1994, 1995, and 1996 proving the existence of the Y2K Bug by citing date-handling errors they found in advance in mainframe vendor software products.

They found the bugs by using Date Simulation software products to advance the date on their company’s mainframes to key Y2K test dates such as 12/31/1999, 01/01/2000, 02/29/2000, 03/01/2000, and 12/31/2000.

The 2 Date Simulation products they used to capture bugs in advance were: 1) “TicToc” from Isogon Inc., and 2) “HourGlass 2000” from Mainware Software Inc.

The 2 main articles where he cited his team’s Y2K findings were: 1) “Year 2000 Denial,” COMPUTERWORLD, October 24, 1994,

  (co-authored by Jeff Furman and Albert Marotta) and

2) "Y2K: Not Just An Applications Issue," Technical Support Magazine, July 1995,

    by Jeff Furman. 

On Dec. 20, 2006, at the request of the Project Management Institute (PMI's New York City Chapter) Jeff Furman gave a seminar and led a Q&A on the topic “Y2K Lessons Learned.”

JFNYC JFNYC (talk) 02:20, 1 February 2009 (UTC)

—Preceding unsigned comment added by JFNYC (talk • contribs) 02:11, 1 February 2009 (UTC) 

Historical Context

I think this article misses the mark in conveying the hysteria surrounding Y2K. The media reported all kinds of craziness -- planes would be falling out of the sky, banks not working, etc. Clearly none of that happened, and I think the article well explains the differing arguments of Why. But it misses the real panic of the time, by Dec 1999 it had mostly dispersed, but earlier when the first repors of this starting being disseminated it was all hell break loose. Vegaswiki (talk) 17:17, 11 February 2009 (UTC)

There are two conflicting threads in this Y2K wiki description: (1) Y2K was a very real software challenge, and (2) the media is driven by the old mantra of "if it bleeds, it leads." People have been gleefully predicting & selling "the end is nigh" for a looooong time. Y2K had absolutely NOTHING to do with the End Times, the end of the world, elevators in Jerusalem, or whatever whacko madness floats your boat. DEddy (talk) 17:54, 26 February 2009 (UTC)

decimal v binary

The Y2K problem is based on the idea that computer time (dates) are maintained internally in decimal and are compared on the last two digits; hence, the angst that year 00 (2000) would be seen as before year 99 (1999).

Internally, microcomputer operating systems use a binary clock based on 32-bit or 64-bit counts of seconds, with different systems starting their second zero at midnight 1 Jan 1915 (Mac) or 1970 (Unix) or 1980 (IBM/Microsoft); so, their internal binary date/time comparisons did not matter over the 1999--2000 decimal year transition. The real problem will be when those binary second counts wrap from the max count all 1's back to zero. The year 2107 will be the crucial point for MSDOS. Naaman Brown (talk) 14:03, 25 March 2009 (UTC)

I beg to differ. Regardless of how the operating system handles date/time, an APPLICATION has the ability to step in & do something different & potentially dangerous. DEddy (talk) 15:37, 25 March 2009 (UTC)

Popular Culture deletion

EncMster: Thank you! DEddy (talk) 01:47, 17 April 2009 (UTC)

Perhaps I am biased on this particular issue, but Y2K was about fixing a long standing BAD software practice that would have messed up a lot of information systems. It was NOT about the end-of-the-world. If a constituency needs to revel in the End Times predictions/mania, how about moving the "Popular Culture" to a new page, please? DEddy (talk) 01:35, 18 April 2009 (UTC)

True, it was about a software problem. But in the late 90s, not only was the problem itself a major issue, but the very phrase "Y2K" was a major part of pop culture. There were puns on the phrase too, like the Will Smith song "Will 2K" and the Beanie Baby "Ty2K." Had the Y2K problem never been, these pop icons of their days never would have existed. Wikipedia is not totally anti-trivia, and pop culture is perfectly acceptable. Xyz7890 (talk) 03:24, 19 April 2009 (UTC)

While I agree there is a place for pop culture in Wikipedia, what had happened to this article was a prime example why trivia is justifiably discouraged. It contained dozens of drive-by contributions enumerating every instance of a fictional character making some quip concerning year 2000. All but one or two of these were useless in this article.

A proper, useful in popular culture section would provide someone unfamiliar with the topic some insight as to what was widely or collectively perceived about the problem. Examples which come to mind would be:

• the number of jokes Jay Leno, Saturday Night Live, MADtv, David Letterman, etc. made about it afterward, and how the jokes were distributed over the years
• the quantity and timing of news articles about it
• the number of bomb shelters, safe rooms, and weapons bought in preparation
• etc.

—EncMstr (talk) 17:59, 19 April 2009 (UTC)

It is your opinion that most of these are useless. That is fine. Before we remove any, we should have a clear concensus that any particular ones do not belong, otherwise they will just reappear. Xyz7890 (talk) 18:13, 19 April 2009 (UTC)

It's 90% the case that if the section isn't present, trivia additions don't occur—at least for most articles. D.B. Cooper has a unique solution (D. B. Cooper in popular culture), though a few editors have protested. If there is genuine demand to add pop culture tidbits to this topic, perhaps the same arrangement would work here? —EncMstr (talk) 21:28, 19 April 2009 (UTC)

I would not call these "pop culture" additions useless. I would call them not relevant to the Y2K software issue. The fact that pop culture takes over is really a more long standing issue. When a phenomenon reaches a certain level it assumes a life of its own. The whole "end of the world" angle of many of these pop culture comments should be filed in a separate article... perhaps "Y2K Mania (not software)." Clearly lots of people do believe in the "end of the world" predictions & are eager to wrap their beliefs around what comes to hand. Software systems are complex enough without wrapping them in millennia old superstitions. DEddy (talk) 16:47, 26 April 2009 (UTC)

The article is about the Y2K problem. I can see how a Will Smith CD or a Chris Jericho gimmick would be related to an article about Y2K, but not about the specific problem.

Eugenespeed (talk) 09:53, 3 May 2009 (UTC)

JavaScript

I've added more information, tried to clear things up and added two refrences. I do however remember it like the original text said that JavaScript was changed from always returning the years from 1900 to this complex solution where the return values number of digits depends on which century it is. I can't however find any sources for my claim. Anyone willing, and able, is welcome to do so. A reference on Javascript 1.0 would probably be a good start but I can't find one. 83.253.140.185 (talk) 16:42, 7 June 2009 (UTC)

Y2G?

"helped along"? How so? This/today is the first time I've seen/heard of Y2G. Supporting evidence please. DEddy (talk) 20:32, 24 June 2009 (UTC)

Straight Dope and US Senate on Y2K - possible ref materal

Hello all,

The Straight Dope has recently published a column on the after effects of Y2K. See it here. In the column, Cecil Adams makes reference to a February 2000 report by a US Senate Y2K panel on the effects of the bug. The Wayback folks have archived the report here as well (warning: PDF); these could be interesting ref materals. Hope this helps. - Thanks, Hoshie 09:33, 20 July 2009 (UTC)

Basically an opinion piece, I'd say. It makes reference to other material, but there's a good argument for saying that the Senate report is also an opinion piece. Mr Barndoor (talk) 14:28, 20 July 2009 (UTC)

4 Digit Years on Forms

"Many banks have responded to the Y2K problem by forcing full 4-digit year entries on cheque forms, which helps to prevent the error from occurring in accounting environments." What does printing a 4 digit year on a check/form have to do with how SYSTEMS process dates? Have we learned nothing? DEddy (talk) 19:51, 9 August 2009 (UTC)

Professor Anderson

I listened to Professor Anderson's interview on the BBC (http://www.bbc.co.uk/programmes/b00mz53r ). Unfortunately the analysis he offered (minimal problems found at his university & no problems in Korea) is at such a high level as to be meaningless. There were plenty of reasons why there were no likely Y2K issues in an organization... someONE was smart enough to glance at the practice of using 2 digit years, instantly see it would be a nasty problem in 20 years & be in the position to decree the the use of 4 digit years. Perhaps Korea was in the social practice of using 4 digit years.

His suggested solution of "turn it off & turn it back on again..." sounds like he's only looking at hardware. It's not possible to just turn the information systems of a large organization off/on, since typically there are 100s/1000s of such applications & no one understands how data flows between them.

The potential damage done here is that it reinforces the widespread belief that Y2K was a hoax because there was not widespread chaos.

There are several comments to Anderson's interview that explain that Y2K issues were very real. DEddy (talk) 17:24, 4 October 2009 (UTC)

Y2K Scam?

Someone should include the viewpoint that the y2k scare fueled consumer consumption and many companies made significant profits. For example bunker manufacturers made a lot of profits, as they did during the nuclear atomic arms scare of the 1950s between US & Russia. —Preceding unsigned comment added by 24.80.121.28 (talk) 05:51, 16 November 2009 (UTC)

Y2K--the problem with the widespread (but by no means universal) use of 2 digit years in computerized information systems--was NOT a scam. The scam aspect piled on when other human factors kicked in. Selling the end of the world has been--and continues to be--a long established, successful business. In the media, the practice is known as "If it bleeds it leads." Bad news sells. The unavoidable technical fact is that there is no two digit number after 99. Period! Full Stop!! DEddy (talk) 19:59, 26 November 2009 (UTC)

Lessons NOT learned

What's the objection to the following two statements which were removed without comment?

Since there has been a minor flurry of interest in the year 2000 issue on its 10 year anniversary, it is duly noted that to a noticeable extent reporting on the issue still tends to emphasize either the problem was not real or it was over dramatized to generate unneeded spending.

One facet seems to be accepted, that the "inventories" of systems were quickly discarded or allowed to become outdated, and organizations quickly returned to a state of blissful ignorance as to the extent and interconnectedness of their software portfolios.

DEddy (talk) 02:53, 6 January 2010 (UTC)

Is either of them sourced? Or are they simply observations you've made? If you have citations of people making these assertions then we can present them as something people have noticed. If you are making the observations yourself then they would be considered OR and wouldn't be allowed in the article. Padillah (talk) 12:56, 6 January 2010 (UTC)

What is a "source" or "observation" or "citation" in the context of what you're asking? I have "experienced" this personally, in conversations/interviews with journalists, and reading in online "magazines." DEddy (talk) 18:24, 6 January 2010 (UTC)

This commentary is what comes to hand first... http://www.infoworld.com/d/adventures-in-it/y2k-10-years-later-459?page=0,0 Would it be considered an acceptable source/observation/citation? DEddy (talk) 18:49, 6 January 2010 (UTC)

What I mean by sourced is, any thing represented by Wikipedia has to be frank statements. "COBOL is a computer-based language", stuff like that. Those can be defended, they can be sourced to the original work that made the statement. I can show you a book that has the above statement in it. It is verifiable and correct. We can add anything we want to an article but it has to be neutral and verifiable. The one thing we can't do is post our personal experience, because it should be obvious that each of us cannot experience the entirety of any given subject thus we cannot be the sole owner of information regarding a subject. Rules have been put in place to discourage posting original research to an article since it can't be defended (other than syaing "Did, too" like a 6 year-old). So, that is what I mean by "sourced" vs. "observation". If you have a source that plainly says "Reporting on the Y2K bug still emphasizes ..." than we can add it and attribute the statement to a source. If it's just something you've noticed than it would be considered OR and removed. As for the article you provided, it's a Reliable Source but it doesn't support what you are asking for. It's an example of yet another report about Y2K being overblown or not real, but it doesn't make that statement. That is a situation called WP:SYN|synthesizing information]], when you come to a conclusion based on sources rather than supported by sources. Padillah (talk) 20:29, 6 January 2010 (UTC)

Opposing Views Section - misleading citation

For the bullet point that states "The lack of Y2K-related problems in countries such as Italy ... None of these countries experienced any Y2K problems regarded as worth reporting:" The citation on the sentence "none of these countries experienced any Y2K problems" cites an article published in 1999. So that citation is misleading on that sentence. Is there any relevant citation for this claim?

--Steve3003 (talk) 14:29, 6 January 2010 (UTC)

Extremely poorly written article

Is no one here who believes that this article is extremely poorly written?

• First off all: what is so hard in pointing out that "if (80 < 10) then ..." yields false while "if (1980 < 2010) then ..." yields true? So every programmer and non programmer understands the implications instead of writing 3 or 4 screens which in fact don't explain what the problem is?
• Second: in typical program (e.g. in COBOL ) that is not using 4 digits for years per 1 million lines of code you have roughly 500 to 2000 line of code with a "faulty" comparison like noted above. The estimation ow server this problem is, is left up for the reader.
• Third: what is the problem with a leap year? Easy explained: if your year is e.g. 2003, which is not a leap year, 1st of February will be the same weekday like 1st of March (because February has 28 days). Now in a leap year like e.g. 2000 (1900 was not a leap year!!!) February has an extra day. So now in the leap year the weekdays from March on are one day off. That means systems that would shut down on sundays and on saturday (e.g. access control systems for secure areas in a chemical factory or something, or a simple elevator control) now shut down on fridays and saturdays (because they believe the friday is a saturday and the sunday is a monday). A system using 2 digits will obviously in year 2000 assume this is not a leap year because it will believe it is 1900.
• Fourth: no, no, no, the problem did not come from the desire to save memory space or storing space. Perhaps there was a desire to save space on disk in a database but that is doubtable. A typical program written in COBOL will just do some simple thing: e.g. lets calculate which "employee" has his birthday today and will retire, and will be shifted from employee to pension receiver. That might be just 100 lines in COBOL walking over 1 table in a data base and copying new retired people from the working population to the one receiving pensions. Such a program might read from a table that only has 2 digits for years but could always have had expanded them to 4 digits during loading of that record into memory. In Memory you always had only one record, which might be 2000 bytes big, 2 bytes more would not have been a problem. But programmers did not do that. The real reason for using 2 digits instead of 4 is very simple: programmers are people like you and me. In real live they only use 2 digits. So why should they consider to use 4 in programming? The argument it came from "space saving" is a newspaper argument and should not be in Wikipedia as a main argument, it is simply false.
  • Saving space and time were absolutely factors in using 2 digits rather than 4. Aside from the forms that used 2 digits (much faster to write 2 digits), "databases" were boxes of punch cards. Keypunching was both slow and expensive, so saving 2 digits was closely watched. There was no single reason to use 2 rather than 4 digits, but there were multiple factors... speed, space, habit, existing paper forms, etc. DEddy (talk) 00:51, 15 April 2010 (UTC)
• Fives: in the section about "Opposing view" you are not able to debunk any single of them? Hello?
• Six: There are lots of types of errors normal people can not imagine. E.g. I have a insurance contract running from 1988 till 2010. And for some reason the starting year is also encoded into the "contract id" and not only saved in records (in the database). So some parts of the program perhaps wont use external records (which might be 4 digits meanwhile because they got fixed in 1995) but they extract the supposed year from the "contract id". And again a "if (year_a < year_b) then ..." comparison might fail.

Sorry for being negative, but I'm a programmer. From a standpoint of a programmer this article completely fails to explain what the programming problem really is. Further more it is written as if the "Opposing views" had any solid base to be founded on, which neither of them has.

• • Sadly the real problem of 2 vs 4 digit years, has long since been drowned in the belief by most people that since nothing appeared to happen—"they" said bodies would fall from the sky at the stroke of midnight..."—therefore the whole thing was a hoax. DEddy (talk) 00:51, 15 April 2010 (UTC)

Aos1966 (talk) 23:40, 14 April 2010 (UTC)

The quote "I’m one of the culprits who created this problem. I used to write those programs back in the 1960s and 1970s, anunning through various mathematical exercises" contains a typo? I can't find "annunning" in any dictionary... is that supposed to be "running"? —Preceding unsigned comment added by 67.193.106.116 (talk) 01:47, 15 April 2010 (UTC)

Yes indeed, someone messed up the quote on the 13th of March 2010. I've now corrected it. Mr Barndoor (talk) 13:41, 19 April 2010 (UTC)

World ends in 2156!

Why do people always seem to overlook the fact that one byte can hold a value up to 255? The bug should be triggered, then, in 2156, not 2000. Am I missing something? 142.59.172.187 11:16, 10 June 2007 (UTC)

In some systems this might be the case, but it is not so common I think. There are several dates in the future that can cause software problems, like 2038, 2100, 2156 etc. The world will not end. -- BIL 14:51, 16 June 2007 (UTC)

The date problems will be those involving not storing 4 digit years at all [Y2K], incorrect interpreation of a user entered 2 digit years [potenitally always], simplified leap year calculations [Y1900 due to bug in Excell, but numerious programs will have it in 2100 if they are still in use], and overflow [Y2038] . Someone has also added a Y2007 problem to the see also, but that is unrelated [only dealing with times being off by an hour due to change in US daylight savings time start & end dates]. I'm not aware of any special significance to 2156. Jon 21:06, 22 August 2007 (UTC)

Like 142.59.172.187 said, the problem with 2156 is in storing dates in single bytes with 1900 as the base year. A byte can only hold 256 values (0-255) and after that (at 256), it will roll over back to 0. 1900 + 256 is 2156. Thats where this come from. — Jaxad0127 21:41, 22 August 2007 (UTC)

I strongly doubt they'll be any significant problems for 2156. The smallest number fields I've ever seen anyone actually use in DB2, Oracle, or M$ Access was a 16 bit small int even when it's known the data won't prossibly exceed 10. Jon 20:26, 5 November 2007 (UTC)

The worst affected programs were in Cobol, where the practice is to store pretty much everything in strings. So dates were typically a 6 character field, or three 2 character fields, one for day, one for month, one for year. That is, 1980 was stored as the character 8 and the character 0, not the numbers 8 and 0, nor the number 80. If someone did store the year as a single byte, there could still be issues, depending on whether that number was added to 1900, or appended. I've seen personally seen a program display the current year as "19100". You also need to consider signed integers, which don't have the same range as unsigned integers. For example, a signed char would typically go from -128 to +127. But I doubt there are many such programs, or that they will still be much of importance in 2018. Regards, Ben Aveling 06:29, 22 February 2009 (UTC)

I was a COBOL programmer for about 15 years beginning in the 1970 in financial services firms. One must ALWAYS prefix statements of "this was how it was done" with "it depends"... I don't remember ever seeing dates stored as character strings. What I remember seeing was lots of dates defined as 999999 (ddmmyy) DIGITS. It's been way too long now, but "storing pretty much everything as strings" is highly dubious. COBOL was not designed/intended to be a string processing language. It was field oriented... numeric & character fields. DEddy (talk) 03:15, 25 February 2010 (UTC)

The problem occours when the 8 bits are used to store two, 4 bit unsigned integers, which will hold 0-15 each. You can then use 4 bits for each part of the year mod 100 (i.e. for 1982 you would have 82 and store 8 in one 4 bit integer and 2 in the other). This seems like a fantastic waste to me. Having the simpler 8 bit signed integer and add a constant to it (i.e. for 1982 you would store 82 in the integer and add 1900) seems to me to be the better thing to do (in at least by far most cases). It wastes less storage compared to what it does and allows for a longer period of time before it breaks (meaning it is more likely that it will be replaced or fixed before it becomes a problem). —Preceding unsigned comment added by 80.167.145.223 (talk) 23:14, 24 February 2010 (UTC)