Anyone know the next "heads up" date to watch for, for computer failures?

-- NeedtoKnow (NeedtoKnow@knowledge.com), January 27, 1999

Answers

2000-03-01 (leap year some programs/hardware don't recognize)

-- vbProg (vbProg@microsoftsucks.com), January 27, 1999.

... tomorrow... It's the one we forgot about that gets us. Other than that, March 1st could bring out defects as a result of month end reporting which reveal '99 problems. The next one I know about is April 1st, 1999 which would be the begining of the 1999/2000 fiscal year for Japan, Canada, and New York State (64 days away). Then comes the biggy: April 9, 1999.

For systems with '99 problems that store the date as the number of days in the year connected with the last two digits of the year this is a potential problem. This is because April 9 is the 99th day of the year. A system storing this information may show: 9999

April 9th is also the day that that a power grid test has been planned.

-- Reporter (foo@foo.bar), January 27, 1999.

For a nice chart format of "Bug Breakout Days" visit

Mr. Y2K Non-Exploitive Year 2000 Site

-- Other Lisa (LisaWard2@aol.com), January 27, 1999.

Personally, I think the "big" one is July 1st, 1999.

Earlier dates could be interesting but expect the .gov & .mil lid to be held on tightly.

Diane

-- Diane J. Squire (sacredspaces@yahoo.com), January 27, 1999.

I agree with Diane. July is the regulatory deadline for nuclear power plants to be compliant. July will also be when companies take a good hard luck at their vendors and decide which ones to drop and which ones to switch their business to.

A date to watch for computer failures, though? Most of the potential April failures won't be reported. Any actual July 1 failures (40+ states) probably won't be in the news either. What's to stop a state government from temporarily changing the end date of its fiscal year in its system from June 30, 2000 to December 31, 1999?

The next date for noticeable failures? My guess is the GPS rollover on August 22, 1999.

-- Kevin (mixesmusic@worldnet.att.net), January 27, 1999.

This coming Monday, February 1st. In a lot of companies they start their new fiscal year next week, and most of them aren't ready for what's going to happen. Many of their inter-office programs (billing, payroll, insurance, etc.) will begin to perform calculations and produce information relating to the year 2000. In most companies this may not be apparent to the customers or affect their service, but internal operations will start to get screwed up. The longer they wait to fix it, the more errors will begin to pile up, and the cumulative effects could lead to more serious problems.

-- (@@@.@), January 27, 1999.

Agree. Feb 1 wil be interesting in a lot of companies as the JAE bites down on the data. Probably very little public effect, but keep your ears open. A lot of month end processing is about to burp bad data.

Best guess about public crash will be July 1. I doubt April 9 will be more than a minor problem. The JAE will continue to grow and should really show up at the end of first quarter processing. Still, this is mostly internal to companies and won't be very public.

now, if all the GPS stations go down... very public, but you want to bet you are going to be hearing about "cyberterrorists" soon causing havoc in various systems?

-- RD. ->H (drherr@erols.com), January 27, 1999.

The first of the remaining "public-affecting" dates is just next Monday, 02/01/99. The first full month of any JAE processing will be completed and the results will start hitting a wide audience. So far, we've been treated to stories of bank acounts fluctuating, and some businesses having spot problems with computerized systems.

At work, our inventory management system is currently in week two of a "one week, routine maintenance update". Luckily a smart person PC- ized a rudimentary inventory issue-out and tracking system. But another week and we'll have real shortages beginning to affect production. We can track what's going out, like a merchant tracking his inventory off of a listing, but our managment system did all the replenishment ordering. And don't you just know, next week on 02/01/99 the first inventory run and order cycles are due to be done.

The first end-of-month transaction cycle is my guess for the first major trigger date, with a larger exposure cycle than what 01/01/99 has in many people's minds. And 02/01/99 will just start any news cycle resulting from a 02/01/99 problem. Mid-to-late Feb is when the billings and orders run for 02/01/99 will hit the mailboxes.

I can't wait until the first month's weather of 1999 is projected into 2000 for those folks with "energy consumption averaging" power bills. If there's anything that will get a bunch of DWGI's to become GI, it'll be one hundred year's of weather like Jan 1999 averaged into a single year's power bill. "Balanced Billing" indeed.

WW

-- Wildweasel (vtmldm@epix.net), January 27, 1999.

I would extend what others have said. At this point, there are no particular dates to watch, but rather classes of dates. The beginning of fiscal periods, particularly years and quarters, are important. So are the April date mentioned earlier and September 9, 1999. So is the GPS rollover date, although that is not a Y2K problem per se but another example of the family of rollover problems that Y2K also belongs to. Anyway, from here on in, any day could be the day that a system first tries to process a date on or after 1/1/2000 and thus be at risk for a Y2K failure.

Try to focus more on the events (failure v. sucessful processing of post-Y2K dates) rather than on the dates they occur. Trying to watch out for specific "trouble" dates is going to become increasingly difficult. A recent Gartner Group study estimates that only 8% of Y2K failures will actually occur on 1/1/2000. That leaves 92% of failures happening on some other, currently poorly understood, date.

-- Paul Neuhardt (neuhardt@ultranet.com), January 27, 1999.

Nobody's mentioned the "nines" dates. 9th April (99th day of '99) and 9th. September (9/9/99). Some hairy old codes may treat these as "STOP" codes, and a lot of non-computer types got into the habit of using them as a "don't know" code which may cause trouble when they suddenly get needed as real dates. I doubt these will be a big problem, but if either are in an embedded system ... who knows? I certainly find the choice of 9th April for electricity industry testing in the USA rather odd, unless it's deliberate to provide an excuse.

UK Government starts next FY on 6th. April, which probably has more significance than any other date in '99 for the UK. Payroll troubles at the end of that week and month? If there aren't major hiccups by the end of April, I'll actualy start feeling a bit safer.

-- Nigel Arnot (nra@maxwell.ph.kcl.ac.uk), January 28, 1999.

ANYTIME for jo anne effect problems. a large company that i am personally aware of, had a problem on jan 15 because that's when they atarted doing forecasts into 2000.

other than that, feb 28-march 1 for leap year, april 1, april 9.

have a nice day.

-- jocelyne slough (jonslough@tln.net), January 28, 1999.