Talk:Year 2000 problem

From Wikipedia, the free encyclopedia
Jump to: navigation, search
WikiProject Time (Rated B-class, High-importance)
WikiProject icon This article is within the scope of WikiProject Time, a collaborative effort to improve the coverage of Time on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
B-Class article B  This article has been rated as B-Class on the project's quality scale.
 High  This article has been rated as High-importance on the project's importance scale.
 
WikiProject Computing / CompSci (Rated B-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
B-Class article B  This article has been rated as B-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
Taskforce icon
This article is supported by WikiProject Computer science.
 

simply Y2K?[edit]

The description says "or simply Y2K". Is this correct? I think is always necessary to include bug, problem or issue, by saying only Y2K you are not implying the bug but the year 2000 in general. right? — Preceding unsigned comment added by 128.79.69.223 (talk) 12:26, 11 December 2014 (UTC)

Yes. It's often descibed as Y2K. Try a Google search of only the term. Walter Görlitz (talk) 14:29, 11 December 2014 (UTC)

Documented Errors - On 31 December 2000 or 1 January 2001[edit]

I would like to add this subsection, but my change was reverted. The BBC news story is here: http://news.bbc.co.uk/1/hi/sci/tech/1101917.stm. These were examples due to mis-handling of the 2000 leap year, as 2000 was 366 days long not 365, so problems occurred on last day of 2000 or first day of 2001 (I assume this was becuase two systems were involved, one with the bug and one without). During testing for Y2K setting clocks to 31 December 2000 and to 1 January 2001 were considered as requirements, due to liklihood of problems with those two dates. As per FAQ in the BSI spec (reference 1):

Q13 What about "critical dates" other than 1 January 2000 which are related to the Year 2000 problem? A13 ... Another date which may cause problems is 31st December 2000 because it is the 366th day of the year, and in previous leap years some systems have failed when this date was encountered.

Yes it was really a leap year problem, but it was included by BSI as part of Y2K, so any documented errors were scenarios that should have been checked as part of any Y2K conformance testing. John a s (talk) 12:46, 31 December 2014 (UTC)

The year 2000 problem was about the change of date from December 31, 1999 to January 1, 2000. The leap year bug is dealt with in another location and the BBC is simply guilty of conflating the issues. Walter Görlitz (talk) 15:33, 31 December 2014 (UTC)
No single authority was in charge of the year 2000 problem, so there is no authoritative definition of what is or is not included. "The year 2000 problem was about the change of date from December 31, 1999 to January 1, 2000" is just the opinion of one Wikipedia editor. Having been involved in screening programs for year 2000 problems, I know from personal experience that finding authoritative statements about whether 2000 was or was not a leap year was difficult, and this knowledge was a prerequisite for screening programs. Our department finally found a copy of the Calendar (New Style) Act 1750 and decided that was authoritative enough. Jc3s5h (talk) 16:13, 31 December 2014 (UTC)
And the "critical dates" issue is just the opinion of one BBC editor. Check the lede of this article. It states what the purpose of this article is. Then check leap year#Algorithm to see if the problem you're describing doesn't belong there instead. Walter Görlitz (talk) 16:19, 31 December 2014 (UTC)
The leap year was a problem, as simple-minded programmers had it wrong sometimes. But there was never any confusion about the rules; the scheme from 1750 was well known, not disputed, and mostly correctly implemented, except by programmers who didn't care about the long run, which came up in the same year as the digit rollover problem. I can't believe that "finding authoritative statements" was the problem. Dicklyon (talk) 16:22, 31 December 2014 (UTC)
There was a problem finding authoritative statements. The Internet hadn't progressed very far in terms of distinguishing good sites from bad sites. Governments were largely absent from the Internet. Some programmers had university libraries at hand, but many didn't. And for US programmers, there was this problem: "[The Gregorian Calendar] is the official calendar of the United Kingdom, but not of the United States (which has no official calendar)." (Richards 2013, p. 585) It's tough to find a law that doesn't exist, especially when you don't know it doesn't exist. Jc3s5h (talk) 16:34, 31 December 2014 (UTC)
I don't know about laws, but the idea that the US might use a calendar different from what Europe used seems like a strawman. As one born on 29 Feb, I had nearly a half century to get used to the idea that 2000 was special in that I would not miss a birthday the way some did in 1900. This was well known; I used it when I wrote a day-of-week calculating program in 1971 (in Fortran). Dicklyon (talk) 17:15, 31 December 2014 (UTC)
Here are a ton of pre-2000 books about it, including many of the books that were being pitched to you for the purpose of working on the Y2K problem; you wouldn't need the internet to find these (though there was a very well developed book market on the internet, including abebooks.com, by 1996). Dicklyon (talk) 17:28, 31 December 2014 (UTC)
I hope you'll understand that I didn't keep a record of all the sources that had it wrong, and don't remember which ones had it it wrong. But considering the number of seemingly reliable corporations like IBM, Lotus, and Microsoft that had some incorrect leap year handling, you can strike all the computer-oriented books off your list, because the computer industry was not perceived as reliable. With many of the remaining books, it isn't obvious from the titles or on-line descriptions that they addressed the Y2K problem.
Of course, if one could find a European law (which was reliably translated into English, if necessary) that described the leap year calculation, it would be safe to assume the US would do the same. But it's a lot easier for a person in the US, not near a major library, to research US law than foreign law. Jc3s5h (talk) 17:46, 31 December 2014 (UTC)
I know Excel had it wrong, but that was a bug, or laziness, or overight, not due to any source being wrong. Can you actually find a single example of a source that had it wrong (with the aid of the modern internet), or are you just making that up? Dicklyon (talk) 17:57, 31 December 2014 (UTC)
A quick search reveals this Infoworld page which suggest there is a year 4882 problem, when a second leap day will be required. It's written as if that is an actual law. Since there is no such law, the entire article looses all its credibility. Jc3s5h (talk) 18:11, 31 December 2014 (UTC)
It looks to me more like it's written as a joke that someone slipped by their editor! Hilarious, though. Dicklyon (talk) 17:58, 1 January 2015 (UTC)
But that's about leap years, not about the rolling over of dates due to the way dates were stored. Please stop attempting to conflate the two issues. Walter Görlitz (talk) 17:03, 31 December 2014 (UTC)
The year 2000 problem involves not only the most widespread aspect of the problem, dates being stored with only two digits, but also the correction and testing process. Anybody with more than two brain cells would require testing to see if each purported correction handled leap years correctly. So leap year processing is part of the year 2000 problem. Jc3s5h (talk) 17:09, 31 December 2014 (UTC)
It involved the the transition from December 31, 1999, to January 1, 2000. That entry only mentions "In addition, some computer software did not take into account that the year 2000 was a leap year" as a single sentence, not making it the focus of the problem. [1], [2], [3], [4] and other do not even mention the leap year issue. 208.81.212.222 (talk) 18:47, 31 December 2014 (UTC)
It seems odd that we are now in 2015 and are debating what was or was not the year 2000 problem, but hey, happy new year to you! The article introduction includes the BSI definition of the year 2000 problem, which includes as rule 4 "Year 2000 must be recognized as a leap year". As this appears to be the issue with systems mentioned in the 5 January 2001 BBC news story, 31 December 2000 and 1 January 2001 issues should be included. One of the agreed authoritative sources for Y2K stories was the BBC, I think the consensus would be that this is enough justification to include it. BTW the only reason I came to this WP article was there was a 15-year anniversary piece yesterday: http://www.bbc.co.uk/news/magazine-30576670.John a s (talk) 16:04, 1 January 2015 (UTC)
It seems odd that we are now in 2015 and are debating what was or was not the year 2000 problem Not odd at all. The media was totally out-of-control in 1999 (fear sells... "if it bleeds it leads") & people expected the world to end. When bodies did not fall from the skies—as predicted or implied by outrageously inflated media "reports"—people lept to the conclusion that the entire Y2K issue was a hoax. File under LESSONS NOT LEARNED. That nothing bad happened because a lot of dedicated people worked really, really hard, was simply too preposterous a conclusion for people to grasp. Then post 2000-01-01 there was a legal gag order at companies so there was no opportunity for hard working project teams to stand up & take a much deserved bow. Bottom line? Folks who should have learned how complex information systems are & how totally dependant society is on these systems, learned nothing. Management thinks IT ripped them off by crying wolf. DEddy (talk) 21:19, 5 January 2015 (UTC)
The BBC news item doesn't make much sense. It clearly says that tills thought it was 1901, not 2001, which is a millennium bug problem, not a leap year algorithm problem. Also, trains were not programmed not to run on 1 January, so whether the software thought it was 31 December or 1 January should not have made them stop. Why should we assume the railway used two systems to keep track of dates?
This an area in which people are generally clueless. Ask people at random if 2100 will be a leap year and you might be surprised at the answers you get. Even the simple Julian rule (one leap year every four years) is misquoted: [5], the edit is reverted [6] and the first editor puts the error back in! [7].
Looking at contemporary discussions, a lot of programmers were misinformed, but eventually realised that they needed to look no further than the Royal Greenwich Observatory (or the United States Naval Observatory) for authoritative information. [8], [9], [10], [11], [12]. 156.61.250.250 (talk) 10:16, 7 January 2015 (UTC)

References

  • E. G. Richards, "Calendars", in eds. S. E. Urban and P. K. Seidelmann Explanatory Supplement to the Astronomical Almanac 3rd ed. (Mill Valley CA: University Science Books, 2013).

Y2K causing erroneous reports, leading to abortion[edit]

In the cited story for the Sheffield/Downs Syndrome error, there is absolutely nothing mentioned about abortions being performed, so until an actual source for this can be cited, the reference for "two abortions carried out" ought to be removed. — Preceding unsigned comment added by 128.230.82.27 (talk) 18:24, 5 January 2015 (UTC)

The relevant sentence in the source is "Investigators in Sheffield admitted two terminations were carried out as a direct result of the mistaken test reports." In this context "terminations" is a synonym of "abortions". Jc3s5h (talk) 18:34, 5 January 2015 (UTC)

The "Opposing View" section[edit]

The section with the arguments against Y2K being a serious problem is poor. It cites two areas where Y2K was not a problem, but offers no opinion on those areas where it undoubtedly would have been a big problem. The "absence of Y2K problems occurring before 1 January 2000" is a fatuous point. I was a Y2K test manager at a large insurance company. We were fully aware that our drop dead date was in the autumn of 1998, when renewal notices for annual contracts extending beyond 31st December 1999 would first be processed. Nearly all of the effort at that company was during 1997 and 1998. 1999 was for final fine tuning. Other, similar companies were doing the same thing.

The argument that problems should have been fixed when they occurred might have applied to organisations with small and simple applications. Large insurance companies, most of whose applications are date related, could not have coped with the avalanche of problems bringing complex systems down. To be kind I'll say that was an ill-informed opinion. I pored over these applications in minute detail and know how they would have failed. I don't have much respect for the opinion of people who tell me I needn't have bothered. — Preceding unsigned comment added by Freddie Threepwood (talkcontribs) 22:26, 11 January 2015

In particular I think the "hoax of the century" claim by the Wall Street Journal has undue prominence. It was written by a columnist, James Taranto, not a technology or business journalist. The link is broken and the article is no longer available online. Such extreme, hysterical claims should surely have a source available that we can go and check to see what evidence he has for his assertions. If it is merely Taranto's personal opinion then that is of no more significance than the views of the guy sitting next to me in the pub. Jean de Luz (talk) 17:05, 15 January 2015 (UTC)

No-one has objected to, or passed any comment, on my criticism of the "opposing view" section. I will therefore update it to qualify the unsubstantiated claims. There are no links available to back up remarkable claims (ie that there would have only a few minor mistakes, and the ludicrous assertion that "fix on failure" would have been generally "efficient and cost-effective"). I will also remove the reference to the Wall Street Journal article, which is ill-informed polemic from a controversialist rather than a commentator. Jean de Luz (talk) 17:50, 18 January 2015 (UTC)

Credibility[edit]

The guy writes as though he knows—experienced—what he's talking about. Given that Y2K is considered a hoax, why not include a well expressed blog? Or are blogs off limits for Wikipedia? DEddy (talk) 00:44, 14 January 2015 (UTC)

See WP:RS, particularly WP:USERGENERATED. If you can show that the author of the blog has written a book on the subject or at least on the subject of software development, then it could be included. Many well-known technology authors have blogs of this sort. Walter Görlitz (talk) 07:00, 14 January 2015 (UTC)

I understand Wikipedia's stance, but it's frustrating in this case. I wrote the blog and I certainly do know what I'm talking about. During Y2K I was involved in high level planning and also dealing with the low level bits & bytes. I haven't seen a retrospective book with anything worthwhile to say about Y2K. The people who were there and understood the issue moved on to do other things, as I did. Journalists and academics who were appallingly ignorant then moved in to clean up. If a good book about Y2K is ever written then the author would have to speak to people like me, and use sources like the documents that I've retained. The Wall Street Journal article "hoax of the century", which is apparently a credible source, is laughable nonsense. It amazes me how confident people can be that they are right and I am wrong on this subject when they have no relevant experience. I've never seen anything credible from one of the sceptics, but Y2K has become one of these topics where people know they can sound off without getting challenged, except by people like me. We're just the guys who did the job. We're not "credible". I restrained myself in the blog I wrote. Comments added to the blog from people I know who are very prominent and credible in the software testing profession were blunter. One of those was a Y2K manager for Nokia. Another worked at a major North American utility company. Like me, they know what the consequences would have been if we'd not tackled the problem. Jean de Luz (talk) 13:02, 14 January 2015 (UTC)

You're missing the point. It's not about you being a Y2K expert, that's not what the guideline states. You're not a recognized computer development or technology expert. Journalists are considered reliable sources since their information is fact-checked. Similarly tech authors go through the same vetting. Bloggers, traditionally, do not. Walter Görlitz (talk) 15:03, 14 January 2015 (UTC)
Journalists are considered reliable sources since their information is fact-checked. Ummmm... Errrrr.... Ahhhhh... Walter "fact checking for Journalists" is so 20th century. Maybe the word hasn't filtered through to the powers-that-be at Wikipedia, but this thing called the Internet has taken a huge chunk out of the newspaper publishing revenue model. Fact checkers were an early casualty.
Two personal Y2K experiences: around 1995 or so the scales fell from my eyes & I finally learned that just because someone could write well about a (technical) topic I was interested in, said absolutely nothing about their knowledge of the topic. They knew how to interview someone & write well. Second: when I was interviewed by the WSJ in August 1995, the piece almost didn't run since the journalist couldn't track my boss vacationing (no cell phone) in the UK for fact checking. Twenty years on, fact-checking in journalism—particularly on technical topics—is largely a fond memory. Personally I consider many allegedly technical journalists as seriously ill informed.
Look at the announcements in WSJ/NYTimes about the new z13 mainframe. Is that journalism or regurgitating a company press release? DEddy (talk) 23:06, 14 January 2015 (UTC)
I never said that the rules for RS are sufficiently modern, I was simply explaining it. I don't care how they create their content only that if Wikipedia is sued for publishing something that we can point to the source of that information.
Thanks for sharing. I don't actually care about your personal Y2K experiences and I'll do you the favour of not sharing mine with you. Walter Görlitz (talk) 05:28, 15 January 2015 (UTC)
I appreciate your being so dismissive of my professional sharing. You've missed something here... since there are (to the best of my knowledge) NO reliable professional recountings of Y2K all we have are our "personal" experiences. If you know of reliable sources for Y2K after action reviews, please list them. I'm currently unaware of any such published resources. DEddy (talk) 13:18, 15 January 2015 (UTC)
I'm sorry if I offended you. You missed the fact that I worked through the Y2K repairs. Walter Görlitz (talk) 14:56, 15 January 2015 (UTC)
WP:V states in part "Self-published expert sources may be considered reliable when produced by an established expert on the subject matter, whose work in the relevant field has previously been published by reliable third-party publications." Since the subject matter is computers, if the blog author had also published work in recognized computer magazines or journals, his comments could remain in Wikipedia. Jc3s5h (talk) 15:39, 14 January 2015 (UTC)
if the blog author had also published work in recognized computer magazines or journals, his comments could remain in Wikipedia This is too good to be true. The medium in which Wikipedia is "published" (am I allowed to say Wikipedia is a form of publishing?) is not considered to be a reliable form of publishing. Words fail me.
Let me get this straight. A technical journalist for the NYTimes is considered a "reliable source" and someone with 30 years hands-on technical experience with thoughts on a blog is NOT reliable? Do I have that right?
What about the 110 Y2K essays (I don't believe "blog" was a word in 1999) I wrote for the Westergaard site (there were at least a dozen or more contributors like myself)? The site is long since gone to the bit bucket in the sky. I think some of the site is in the Wayback Machine. Are those essays "reliable" or "hearsay?" I wrote them. They were "published" on the now defunct Westergaard site (wwww.y2ktimebomb.com), & I have the essays. DEddy (talk) 23:06, 14 January 2015 (UTC)
If by medium you mean paper vs. Internet, Internet sources are just as acceptable as paper sources. If by medium you mean sources that exercise control over what is published vs. blogs (or vanity presses) that allow paying customers to publish whatever they feel like, you're right, those are not acceptable. You are not a Y2K expert, you're a userid on the Internet that cannot vetted. Jc3s5h (talk) 17:54, 15 January 2015 (UTC)

I've had plenty of articles published in IT magazines and journals. Maybe I should get one of them to take the blog article. I know journalists are supposed to have their articles fact checked, but today I've been checking a couple of technology articles in the Mail on Sunday in which the journalist has distorted the figures by a factor of at least 10. C'est la vie. Jean de Luz (talk) 20:39, 14 January 2015 (UTC)

Jean de Luz I haven't seen a retrospective book with anything worthwhile to say about Y2K. I would agree. Here's an interesting sample... http://www.computersincrisis.com I wrote one final Y2K piece in 2000 for some sort of operations magazine. It was a membership magazine (members are told/warned they may be called upon to be interviewed) so as the "journalist" they gave me a list of people, titles, phone numbers & their companies. Objective of the piece was to give folks a richly deserved "atta boy!" for a job well done. I quickly slammed into a stone wall. Companies were under a gag order from the legal department to NOT talk to the press. Companies were fearful of sounding the all-clear too soon. Individuals could not be quoted & requests for quotes would not be considered by legal departments. I have hopes the working papers from that assignment will surface someday from my archives, but I'm not holding my breath.

It's a pity... silence plays to paranoia. It's my firm belief that most people believe that since bodies didn't fall from the sky at the stroke of midnight—as predicted by those reliable sources in journalism—the whole thing was a hoax. The fact that a lot of people worked really hard to avoid slamming into the iceberg (Yes I had a Titanic Y2K essay...) simply cannot be considered. Bottom line? No lessons learned.

Here's a cheery thought... personally I believe the approaching Retirement Brain Drain—Baby Boomers who built the mainframe systems—has the potential of being a significant challenge. For Y2K the Subject Matter Experts were around & eager to pitch in. Going forward those skill/knowledge will not be available. The mainframe systems are still here. I remember one of the standard "Y2K isn't an issue" excusses was that "We're going to replace that system real soon now." Color me skeptical. DEddy (talk) 23:06, 14 January 2015 (UTC)