Can anyone tell me how Y2K is such a big deal?



Can anyone tell me how ISO/QS Y2K is such a big deal, besides the possible trouble of having the files stored into a computer database? What other problems does the year 2000 & beyond represent?
Jason M.


Fully vaccinated are you?
Well, you've chosen a significantly controversial topic. And I will try to address it in under 2,000,000 pages of text. ;)

There are already Y2K law suits. Most companies are being *Required* to state Y2K compliance by customers. Why? Well, it goes beyond "...the possibility of having files stored into a computer database.

Let's look at a couple of issues:

First, understand the concept of a 'Pivot Date'. That is where a device or software uses a 2 digit date field (typically) and at some point there is a 'break-point'. Everything before that date is the previous or next century. Microsoft is a great example of stupidity on a grand scale. Excel, Word and each other program is handled differently. Each have either a different 'reference' file (another idiot Microsoft 'solution') *or* pivot date (even the 'solutions' are different between a few programs). Let's say Excel has a 'pivot date' of 38. Excel would look at a date with the last two digits less than 38 as the next century. xx37 would indicate 2037 while xx39 would indicate 1939. If you have a date field and the person was born in 1934, the program would interpret the date as 2034.

Many microprocessors, for example, have date registers. Some are simple counters - typically 24 hour timers. But - some are continuous (have to be able to calculate *Leap Years*). A 24 timer is typically not a problem, but what about continuous counters? Some 'applications' use these timers, some don't - so a device with a noncompliant IC may work fine, but others which do use the date function (where it is continuous) may not work well.

Old Cobol code is problematic because there is so much of it (legacy included), but there are software evaluators (even for PC's, as of late) popping up everywhere. Those with an 'interest' in solving date problems by looking at each line of code by a HUMAN poo-poo them, but they are effective. But - do remember, third party software checks for possible data file problems, not program files. And they do not FIX any problems, but rather FLAG the possibility of a problem.

Well, I could go on and on but this will give you a brief flavour. Now - what's the big deal? Well, in part that's the problem. You can look at some applications and easily see what the effect would be. For example, look at a HR department where there is a range of workers between 18 years old and 65 years old and where the software contains data for 10 years giving it a 'sensitivity' of 60 years + 10 years = 70 years. And lets say there is a calculation which uses the field. Now - what do you think might happen?

The microprocessor issue is more vague because the producer has no idea what applications their microprocessors are used in (bought off the shelf, so to speak). And until recently folks incorporating the microprocessors into product didn't look (or few did) this as an issue (which it may or may not be). If the issue was never addressed, you may have no idea what will happen.

At this point, I suggest you hit the internet search and read some for details.

I don't see this as an issue where the US or the world will be thrown into turmoil on 1 January 2000 (or 1 January 2001), but I do expect minor problems to pop up over the next 10 years or more.

In short, you may believe that it is just an issue of "...possible trouble of having the files stored into a computer database..." but it has more to do with calculation functions and related functions. Say you know someone on social security and the code is based on a streight 00-->99 to indiate 19xx. Well, as we turn over to 1 January 2000, all entries (such as 2003) will be assumed to be 1903 (not 2003). Now, let's say the software does a calculation based upon a birth date - what could happen? Well, we can't know unless we try each possibility (oh, my - maybe Design of Experiments could be harnessed) because of the number of possibilities with consideration to the possibility of different calculations (in addition to the differences in dates for each of the people). Well, you can predict some possible outcomes - but how much you can predict depends in large part upon how thoroughly the 'problem code' is evaluated.

Y2K is thoroughly misunderstood - a friend of mine was talking to a programmer who was bragging about getting 'rich' working on Y2K problems. The programmer told him there are 2 programs which will 'evaluate and fix' Y2K programs on PCs (he supposedly stressed they fixed the programs then selves). Ummm, well, no there isn't. There are several programs which will look for data files (say Excel spread sheets) and flag 'possible' problems. But - third party programs will not (as my friend was led to believe and as far as I know) take the Excel program and 'fix' the actual program. So even many of those 'fixing' the problem have their heads up their butts.

So - there are a lot of possible problems.

Again, though, I don't see the world stopping and folks going hungry (any more than they do now). There will be bugs.

BTW - the Macintosh is the only computer and OS which has been compliant since (get this) 1984!

But this does NOT address Mac software Y2K issues. Most is - some is not. None of the Microsoft stuff is Y2K compliant for ANY platform it writes for. They are providing 'fixes' but that does NOT address the core issue - garbage software.

Now how confused are you?

[This message has been edited by Marc Smith (edited 01-15-99).]
Top Bottom