Year 2000 Anomalies

greenspun.com : LUSENET : Millennium Salons : One Thread

On Saturday, October 17th, 1998, Mitchell Barnes wrote:

"And my Eudora sort function is snafuing, not doing date sorts, sending sorts, etc. completely correct. And I think some posts are just going into electronic heaven, I never see them again at least. I've noticed this over the past few days when going into my own personal send archives and some posts are missing. Weird, never have had probs with Eudora at all prior to this. - mb

He sent a copy of that to Cynthia Beal, who replied:

We need to start a thread NOW called "anomalies". I'm worried. I just had 35 billion dollars and a lot of odd change added to my accounting program on Thursday - trashed all the input for the day and had to run to the backup. Ruined my data. This has never happened before.

I've got a lot of work this weekend, but will get going on this soon. The wildest stuff is going on here. We've got a reporter for a major international glossy women's magazine who's recently become, shall we say, "vibrantly aware" of the y2k problem. She's doing a story on the Women of Y2k this weekend, focusing on the Eugene area (probably because we have a college class, government getting active, utility running full bore at y2k, and a host of other women involved in the issue of preparedness). I'm inviting every woman I know who's been working Y2k over the last year...

Wild, wild, wild.

cynthia


So this is a thread for anomalies... Those little things that happen in the getting wild, wild, wilder all the time world of computing. If you have any happen where you are, just put 'em here. And tell your friends. And have them tell their friends to put their computing anomalies in the Millennium Salons forum in the Anomalies thread. And we'll get a collection going. And next thing we know, it'll be a category. And then, who knows? It might turn into a whole forum/salon of its own. This should be quite a year for anomalies, no?

Bill


Wednesday, October 21, 1998

Not long after letting a few people know about this thread, I received a few "cautionary" notes about its real value, probably best exemplified by my chief task master, Harlan Smith. Here are his comments:

Bill,

I don't recommend trying to do this.

There will be millions of such things and they will contribute to the noise level, rather than the signal level.

The objective in any communication effort should be to increase the signal-to-noise level. This will probably degrade it.

Harlan

As always, Harlan's comments to me along these lines have an initial tendency to grate. But he wasn't the only one who said essentially that, so I was trapped into having to think about it. I did, and, among other things, replied that I'd seek to put a "filter" on this to make as sure as possible the people reporting anomalies are as sure as they can be they're actually y2k-related: not just garden variety glitches that could be; not examples of computer-related problems that could be "previews" of what we may be in for ("A microcontroller broke down and the machinery went haywire. Is this what it will be like?"); not reports of things that got screwy with the code while people were working to remediate it. (As another person who works on y2k projects pointed out, lots of those things happen in any software project; they go with the territory.)

After saying I would "refine" this introduction this way, one of the other people who sent a cautionary note, Derald Grimwood, had this to say, and supplied the following suggestion for a "rating scale" to apply to postings:

OK, I think your redefinition sounds good. Perhaps you could scale or rate them like this.

A = Confirmed Y2K with Evidence
B = Stongly Suspect Y2K with some supporting evidence
C = Possible Y2K
D = Not Known if Y2K but noteworthy
E = Y2K Rumour of the week

Derald

That scale or rating system's a pretty good idea. So if you've got a year 2000 anomaly to report, please use it and make rough note of where your report fits into it, okay?

And one other point regarding the purpose of this thread, and how that relates to what Harlan had to say about the "signal-to-noise level." I agree with that. So please bear in mind that ultimately, this thread's intended to provide one more "tool" for helping make it easier for people to believe there could be problems arising. For better or worse, most people seem to have an, "I'm from Missouri" attitude about a lot of things, not to mention the ever-popular, "If it ain't broke don't fix it," point of view. In the minds of people who live by those things, and apply them to y2k, all the projections about what "might" or "could" or "probably will" happen can't hold a candle to examples of things that actually did happen. And that's what this thread is for.

Okay... That's that.

~ Photo and forum software courtesy of Phillip Greenspun (home) and office ~



-- Bill (billdale@lakesnet.net), October 18, 1998

Answers

Response to Anomalies

Where my wife works they have just installed MS Word from the Office '97 Proffesional package. When she's typing in the date, the logic of the program, tries to prompt her with 'it's own version' of the date.

And this is a snapshot of what it comes up with: http://www.cairns.net.au/~sharefin/2098.gif

The year = 2098 One hundred years into the future. ^o-o^

-- Nick Laird (sharefin@cairns.net.au), October 18, 1998.


The following note from Hawaii came across from the Lowell, MA, listserver on October 20th:

"I just received a last minute opportunity to talk on our Kauai public radio station in regards to Y2K. I am very literate on the subject, but they specificaly want me to focus on events that have already happened. Can anyone help me with some info tidbits on verified Y2K oriented problems, glitchs, etc, that have already occurred?

Kauai would be grateful.

K."

In reply, Tony Toews wrote:

"See my Real Life Incidents page at http://www.granite.ab.ca/year2000/incidents.htm"

Can't vouch for the validity of the referenced examples, but they look credible and legit. There is also a link there that leads to another site that collects these things.

Bill



-- Bill (windrow@mediaweb.gr), October 21, 1998.


From: Roleigh Martin

Kim Callis brought this to my attention. It's the first 1/1/1999 reported problem I've heard of all year (could not resist that literal joke).

--Roleigh

http:// search.washingtonpost.com/wp-srv/WAPO/19990101/V000828-010199- idx.html

Washington Post
Year 2000 Bug Felt Before '99 Dawns
By Mark Warbis
Associated Press Writer
Friday, January 1, 1999; 4:01 a.m. EST

BOISE, Idaho (AP) -- Thousands of people using outdated accounting software in their homes or businesses run the risk of being bitten by the Y2K bug a year earlier than expected.

Cougar Mountain Software Inc. of Boise rushed the newest version of its Act Plus accounting program to Lynn Electric on Thursday after the small Bluefield, W.Va., company tried to close its 1998 payroll.

It was using 3 1/2-year-old software unable to translate dates that included the year 2000.

"All the documents reverted to 1944," Cougar Mountain spokesman Dave Lakhani said. "They were unable to process their payroll and had to order the update to correct the problem."

Even with the lost man hours, potentially lost data and the hassle of trying again to close its books over the New Year's holiday weekend, Lynn Electric got off cheap. The software upgrade cost only $400.

But experts estimate larger businesses and those using customized software could face $50,000 to $100,000 expenses.

Vincent Hamm, president of Aim High Inc., a computer consulting firm in Golden, Colo., said he expects to hear similar tales of woe in coming days as accounting software users open 1999 financial calendars that typically extend 18 months -- into 2000.

"This is the first one that I've heard, but it makes perfect sense," he said. "Anytime that you've got something that's forward looking and it crosses that threshold, you've got a potential problem."

[snip]

-- Roleigh (Roleigh.Martin-1@tc.umn.edu), January 01, 1999.


Peter de Jager's list of links to actual Y2K failure incidents...

This page contains links to online news stories related to incidents of Y2K Bug Bytes. News sources are reporting expected and unexpected failures across the globe related to the Year 2000 computing problem. (From Roleigh Martin)

Year 2000 Bug Bytes

(From Roleigh Martin)

-- Bill (billdale@lakesnet.net), January 17, 1999.


From: "Rick Gold" miss_y2k@hotmail.com
Subject: Reports of Y2K failures on the rise
Date: Thu, 21 Jan 1999

Reports of Y2K failures on the rise - from other y2k lists

FROM WIRE REPORTS
http://www.joc.com/issues/990119/i1nsur/e23991.htm (online as of today's date)

The New Year promises to be one of increased computer problems attributed to the Year 2000 technology glitch, according to a report from a Y2K consultant.

With 2000 nearly upon us, a survey of 110 U.S. corporations, 12 federal, state and local government agencies, and 12 industrial sectors found that a large number of the respondents have experienced Year 2000-related failures, and nearly all of them -- 98% -- expect more such failures in 1999.

The quarterly survey was conducted by Cap Gemini America, a Y2K consulting and services provider.

The report found that every company is now developing contingency plans to prepare for expected failures.

In a similar study done in April, only 3% were putting together such plans. More and more firms expect Year 2000 issues to affect relationships with business partners, meaning that Y2K-compliant companies are now considering breaking off partnerships with those companies that are not making headway on the technology glitch.

The number of companies that are very likely, or potentially likely, to stop doing business with noncompliant suppliers rose to 69%, up from 60% in the third quarter.

The percentage of firms planning to incorporate Year 2000 compliance into marketing messages increased to 55% from 50% in October.

The study found that the percentage of companies with a process in place to validate renovated code prior to testing rose substantially in the past quarter from 16% to 62%.

"Testing is an essential and time-consuming element of an effective Year 2000 program," Jim Woodward, senior vice president of Cap Gemini America and head of its TransMillennium Services, said in a statement. "It's encouraging that nearly two-thirds of these large organizations are validating their renovated code before it enters the testing stage."

More than half of the respondents -- 55% -- have already experienced a Year 2000-related computer error.

This figure is up from 44% in October and 40% in July, according to Cap Gemini. Most failures caused processing disruptions of financial miscalculations.

Three quarters of the organizations -- 74% -- expect that more than half of their systems will be tested and compliant by Jan. 1, 1999, a percentage unchanged since October, according to the survey.

However, 92% of the companies report missing Year 2000 plan deadlines more frequently, up from 84% last July.

In regards to Y2K budgets, companies continue to underestimate their Y2K spending. Eighty-four percent classified their cost estimates as "too low," compared with 82% in the third quarter and 87% in the second quarter. The quarterly survey is conducted by Rubin Systems.

-- Bill (billdale@lakesnet.net), January 21, 1999.



Moderation questions? read the FAQ