Anyone mind telling me what the Y2K bug was?

[Reirden]

I'm curious to see how many people knew/ know what the Y2K bug was and why we had to prepare two years in advance for it.

After I get enough answers I will post how I think Titor is a hoax.

EDIT: Eeek. I just realized how ignorant that sounded....sorry.

 

[fragile]

I a nutshell, the y2K bug was two fold....

1) representation by code of '00' was usually meant to signify 1900 not 2000
2) Transactions that spanned the '99 to '00 divide where incorrectly calculated as have take -99 years !

This had the potential to cause havoc in almost any computurised system that used dates for caluculations, the sudden negative numbers that would ensue would generally cause chao, and/or infinite values to spread through systems.

Hence there was a reasonable fear that anything that compared numbers that where pre and post 31/12/99 would or could fail.

Many, many legacy systems where problematic to fix, many COBOL written systems did not store long versions of years, and hence had the problem, I picked up a LOT of work testing and fixing such systems from about 1998 onwards, reaching a climax in December 99 when I could literally name my price....

 

[TimeNot_0]

Simply put, from nuclear power plants and all other devices where computer would effect the world, those systems would not work.

This means that the world would blow up, because of these computer systems.
Now it is a 2038 problem, even as Windows OS's are only good according to Microsoft till some month and date in 2037.

Financial records, anything kept on computer had the potential to create havox, and all about the World would not function, including missiles that may launch, planes that would not fly, anything of the sort.

A temporary fix was applied to the Y2K problem, since that is all that could be done due to shortsightedness of the full current problem of then. Now it is another problem coming up in 2038.

 

[allenchadclark]

Basically a hoax.
The fact of the matter is that computers were not going to crash like most assumed.

 

[fragile]

A Hoax ?

Not from where I was sitting at the time !

Setting the time to 12/31/1999 11:59 on an unpatched system would cause all kinds of chaos, I was heavily involved in preventing the chaos that unpatched legacy systems would have caused, and I can tell you that had we (the company I was working for at the time) not started preparing in 1997, there would have been an awfull lot of public utility companies with serious problems.

I admit that much of the fever that was induced - ie you microwave oven might explode etc, was as a result of public and press mis-understandings, and in part frustration of 'techs' trying to get the message accross to people that that was a potentially huge issue.

I only had a couple of customers that where stopped in their tracks, one was an engineering company that was unable to bill anyone, because the ledger system could not age invoices that where produced prior to 1/1/2000 correctly - assuming that all invoices produced in '99 where actually negative 99 years old, and therefore not due for a while !

The second company simply could not operate its CAM systems, because it assumed that all of the drawings where beyond their expiration data.

There where many other examples that I heard of, but mostly it was nailed dues to reasonably forward thinking, and a lot of testing.

Oh - and some older machines could not cope with a date in 2000 at all, as they had never been designed to last for 20 years, and the authors of the bios's had long ago retired, so some patches where produced with varying results....