When is a Y2K bug not a Y2K bug? (depends on what "is" is)

greenspun.com : LUSENET : Grassroots Information Coordination Center (GICC) : One Thread

When is a Y2K bug not a Y2K bug?


Companies call their software 'compliant' if fixes are coming

Y2K bug problem? Watch your mouth!

Both Microsoft and Intuit, the makers, respectively, of the best-selling Internet Explorer Web browser and Quicken financial software, say their products are having problems dealing with the year 2000 changeover.

But don't dare call it a Y2K problem.

According to Mountain View-based Intuit, the latest versions of Quicken are "Y2K compliant."

The "Y2K compliant" software, one might think, would apparently be free of the Y2K bug.

That bug makes computers read the 1999-2000 rollover as a 1999-1900 time warp. Worldwide havoc, including mislaunched nuclear missiles and power outages, was supposed to occur at the stroke of midnight last Friday because of this flaw.

It never happened. But the Y2K critter does seem to be lingering in some consumers' computers. Unless, of course, you accept a company's definition of a Y2K bug.

Which seems to be: If it's in our program, it isn't a Y2K bug.

For instance, if a Quicken 2000 user enters a year 2000 transaction into a checkbook using the mouse to move between fields, a New Year check will end up being recorded as written in the year 1900. (plenty of time to clear the bank).

Users have to manually enter the full year 2000 into the check registry, not the usual way of entering the last two digits of the year.

In the Mac version of the program, a calendar scrolls from the year 1999 to the year 1900.

A Quicken official said a "patch" for the problems soon would be posted on the company's Web site.

The Web site (www.intuit.com/y2k) also lists other Y2K "issues" and "workarounds."

However, Intuit steadfastly maintains, the company's software is indeed Y2K compliant.

"It's a matter of semantics," said Intuit spokesperson Cecelia Denny.

According to Denny, there is no industry standard for the term "Y2K compliant."

"Each company is free to create their own definition of a Y2K bug problem," she added. In Intuit's opinion, therefore, a program is Y2K bug-free if there is a fix for a Y2K problem.

"Sure, that's a good enough definition," said Denny.

Monday, Microsoft announced there were some Y2K issues with some of its software, specifically its Hotmail free Web-based e-mail and its Explorer Web browsing software.

Explorer, for example, views some dated sites as being from the far-off future of 3900. Something to do with the "get date" command on some pages. Microsoft also pledges to put up patches to fix the problem (www.microsoft.com). cq However, said Microsoft spokesman Adam Sohn, "it's not a Y2K problem."

"Well, it does deal with a changeover to the year 2000, and it does deal with how it reads the changeover, so I guess you could, arguably, say its a Y2K problem," Sohn said. "But we're calling it an implementation problem, not a Y2K problem."


)2000 San Francisco Examiner Page C1

-- Lee Maloney (leemaloney@hotmail.com), January 14, 2000


I have Quicken 2000 on my PC and have discovered no such problems. The 95 version of Quicken was not compliant and would not accept Year 2000 dates. Once I upgraded to Quicken 2000 the problem stopped occurring.

As for the problem with MS Internet Explorer not reading dates properly on a few web pages, is this really worth getting all riled up about? How does this affect your life in a negative manner? Does it really matter to the gentleman who wrote this article that Bouncing Betty's house of Sleaze say's it was updated in 3900. Not exactly the end of the world. When Microsoft say it is an implementation problem, they very likely mean that the date submission code that these sites is using was not in place prior to Microsoft delivering their "Compliant" software. Are companies going to be blamed for not being able to accomodate the systems changes of other companies now? This sort of thin was quite common in 1999 and will still be common in 2020.

-- Lee Barrentine (Wrknman042Legacy@aol.com), January 14, 2000.

This is one of the phenomena which I predicted to my community Y2K group and anyone else who would listen: There would be an amazing number of glitches in Y2K which would be claimed to not be "caused" by the Y2K problem. And the denial would be truthful, to an extent. Perhaps the original Y2K problem did get fixed. But now the fix is the problem. Inevitable given the death march to get the fixes and replacements done on time, with too little time for testing. Happens all the time when new or modified systems are taken on-line, especially with too little time for thorough testing.

The point I have always been making, both before and after the rollover: What difference does it make whether the glitch is caused directly by the century date change issue, or by the repair or replacement to address the century date issue? Isn't a glitch still a glitch?

I am feeling increasingly vindicated about the accuracy of my predictions on this one!

-- Stefan Stackhouse (stefans@mindspring.com), January 14, 2000.

Moderation questions? read the FAQ