We are told that y2k will not be much of a problem because key 1999 dates (Jan. 1, Apr. 1, and July 1) came and went without much obvious disruption.

My question for those of you who understand these things: Is it possible to do a quick, short-term fix on a Fiscal Year 2000 application by telling it to look no further ahead than Dec. 31, 1999?

In other words, is it possible to re-program the end-date so that an application which would normally look out to, say, June 30, 2000, looks no further than the end of this year, thereby avoiding (for now) any "00" year dates?

And if so, would this mean that it's possible that many programs were only temporarily patched rather than permanently fixed?

-- ace (x@y.z), July 20, 1999

Answers

For some systems, it's easy to fake the end date. For example, if your financial system figures out the end-of-quarter date or end-of-month date by looking it up in a table, all you have to do is go edit that table. What system might do that? Maybe one written by an idiot who couldn't write a routine to calc the last date of the financial periods. Maybe one written by a genius, who provided for maximum flexibility.

I never saw this kind of code except in poorly written systems, particularly when written by an end-user. In my experience, mainframe systems calc their dates real carefully, and you cannot bypass that calculation.

If it's manual, can it be worked around? You betcha. Will it be? Absolutely. There are some unemployment systems that are now using fake end-of-coverage dates that are within 1999, to achieve precisely what you're talking about. I'm reasonably sure (roughly as I'm reasonably sure that the earth rotates) that there are systems out there running on VERY short-term patches right now, while programmers work like crazy to get it fixed by the end of the year. No clue as to how many systems, or what their chances are of getting done in time.

-- bw (home@puget.sound), July 20, 1999.

My area of expertise, insurance, is in pretty good shape. Longer policies, for 3 years, had to be dealt with some time ago. One cannot hold the date on a policy, unlike other programs. The problems we see are with independent claim adjusters using old systems, and agents trying to interact with state non-compliant programs dealing with the MVD, CLUE, or MIB state departments. The number of "old timer" agencies that are not ready is staggering. We estimate that 25% will go under after 2000. Most of the programmers and people I know are clueless when they are told this. "Gee, thats rough". Gets alot rougher when you've had a car accident, and have to wait 6 months (or longer) while the claim is processed. Or you are a construction firm that can't get a certificate because your agent thought Y2K was bull. Now you have a firm with 75 unemployed artisans because some agent didn't upgrade his DOS x286.

-- Retroman (retro50@agapeis.net), July 20, 1999.

Some info on state unemployment insurance systems and the patch for January 1, 1999:

http://www.gao.gov/new.items/a299221t.pdf

[snip]

In addition, in June 1999, OMB reported that as of March 31, 1999, 27 states' unemployment insurance systems were compliant, 11 planned to be completed between April and June 1999, 10 planned to be completed between July and September, and 5 planned to be completed between October and December.

[snip]

As an example of the benefits that federal/state partnerships can provide is illustrated by the Department of Labor's unemployment services program. In September 1998, we reported that many State Employment Security Agencies were at risk of failure as early as January 1999 and urged the Department of Labor to initiate the development of realistic contingency plans to ensure continuity of core business processes in the event of Year 2000-induced failures. Just last month, we testified that four state agencies systems could have failed if systems in those states had not been programmed with an emergency patch in December 1998. This patch was developed by several of the state agencies and promoted to other state agencies by the Department of Labor.

[snip]

-- Linkmeister (link@librarian.edu), July 20, 1999.

ace, you may be right about temporary fixes. my husband has had phone calls all year from small and medium sized business wanting help to fix their Jo Anne problems, in most cases they had no idea that it was at all related to year 2000. they were told how to temporarily fix the problem by changing their forecast horizon to remain in 1999.

they were also told they had to get a permanent fix soon, and most did. but it's possible that some of them procrastinated, and we won't know for sure as these are not current clients.

-- jocelyne slough (jonslough@tln.net), July 22, 1999.