Millenium bug!

There's actually two basic issues relating to the millenium bug:

1) The (obvious) problem of using the year "00" in the two digit year portion of a date
2) The problem of February 39th 2000. As a default, every 100 years is NOT a leap year, although every 400 is.

In reality, banks (I work for one myself) and corporations that are publicly listed have been told that they must submit their Y2K strategy to the Sotck Market for review.  In the case of Australia (where I am), all major banks and majority share held corporations have been told by the Australian Stock Exchange that they stand the chance of being de-listed if they don't submit their plans/strategies for review.

For yourself (assuming you're a home user) my advice is simply to run your home PC over the date(s) related to the issues.  For example, I switched my computer's date over to mid-December 1999 and ran it for three months.  This covered both the year "00" issue and the Feb 29th/March 1st issue.

To be honest, their is very little else you can do to prepare yourself.  You might complain about the fact that this form of testing will datestamp your files incorrectly etc etc but in reality the only way to truly test your computer against the Y2K issues is to actually put it through it yourself (I backed up first of course!).

Also, search the web for white papers.  You will find that a lot of people have written complete documents outlining the forseeable results of the dreaded "bug". Look especially at web sites owned by PC companies and computer anti-virus companies.  A good place to start is ICSA's (formerly NCSA) web site: http://www.icsa.com.  These guys deal with computer security issues for a living.

You are right, however, in referring to most of it as "doom and gloom hysteria".  Some people have watched too many "end of the world" sci-fi movies. You can do all the testing you like with all of the compliance checking software in the world and still not be fully prepared for the real thing.

Finally, don't put too much stock in worrying about banks and companies etc.  They have had plenty of time to deal with the issue.  Fortunately, we do not live in an entirely paperless society (yet!) so you will still have records such as bank statements, bills etc. to fall back on if things do go wrong.

I realise that this is not a "in a nutshell" but the Y2K issue is not small enough for a brief paragraph or two of words.

Thanks, korvan!  Nice explanation.  Well written..

RoseFire

don't phone anyone on the millenium switch over :), you will get a phone bill of 100 non stop on the phone :):), hang on I know someone like that...

100 years that is.

The first thing you need to be is rational - don't panic.
Yes the Year 2000 is a problem but is generally over-hyped. The reason why it is over-hyped is firstly the press, secondly to get people to act by creating an awareness, and lastly because some people see it as a money making opportunity.

Another general rule can also be if you do nothing to prepare yourself then you will have a problem.  This will be true and vary depending on situation. The "bug" effects computers but the real effect is the business or operations driven by those computers or electronic chips.  Think of it this way if your home pc hangs or falls over at midnight - so what it crashes often enough normally - just reboot - set the date and away you go.  But what happens if the chip that controls your central heating decides that it is now summer and not winter (just an example)- you wake up a little colder and reset the control and life goes on.  In both of these you could have been prepared and known what will happen and what you should do to rectify the problem, or even solved it before it happened.

The morale of the story is that the Year 2000 is a problem but you have an advantage - you know when it is going to happen.  The disadvantage is that it effects everywhere at roughly the same time.

You have to assess your situation and take rational action - this does not necessarily mean throw money at it.

The web has tons of info for you to sift through.  Some sites I can recommend are:-
http://www.cinderella.co.za/ - Zero Cost Basis
http://www.year2000.com/  - Peter de Jager's site
http://www.yahoo.com/Computers_and_Internet/Year_2000_Problem/
http://www.iee.org.uk/2000risk/ - Embedded systems
http://greenwich2000.com/ - Greenwich 2000
http://www.gmt-2000.com/ - Home of Check 2000
http://www.itmanagersjournal.com/cgi-bin/itm_search.pl?y2k

This should keep you busy and give you a variety of perspectives.  Of course if you need any further info just ask what exactly concerns you about the Year 2000.

Only 379 days left and counting.

In a nutshell - basically the problem began in the early days of computers when memory was an extremely valuable resource. Programmers had to be very stringent in their use of this resource so it became the practice to only use two bytes to store the year portion of a date in, and the first two digits would be assumed to be 19. Nobody foresaw the coming turn of the century.

As you can see it is very unpredictable what computers will do when the assumed 19 becomes a 20. Some computers will think the year is 1900, others will go back to the earliest date there CMOS allows for (often Jan 1 1980).

The problem can effect computers at three levels
   1) The RTC (Real Time Clock)which basically counts  oscilations of a crystal used for timing within the computer.
   2) The BIOS which sets the system date and time from the RTC
   3) Application software.

I have been told that virtually any computer made from 1996 onwards will not have a problem with the first two levels. But if you have any software which uses dates (particularly databases) then you had better ensure they are 2000 compliant, or expect your data to become corrupt.

Like some of the others who made comments I to believe that the problem has been deliberately exaggerated to suit some peoples ends (probably financially) and that the Y2K problem will only be a problem to those who are still using old equipment (386s, etc)and who cannot afford to update. As for the larger companies they have had plenty of time to do something about it. Insurance companies and banks would have become aware of the problem when they started entering data for policies or mortgages which would end after the year 2000.

Finally if you wish to check your own machine out go to www.ontrack.com where you will find a free Y2K PC analysis tool which you can download and run over your computer. If you want to know more take a look at either the IBM, or MICROSOFT sites. Both have a lot of stuff on Y2K

Jason71,

I am posting this just in case you are not sure how things work here and in korvan's interest (uninvited).

In my opinion your question was answered very well by the first post from korvan.  It would be appropriate for you, in my opinion, to reject anray's answer and invite korvan to post an answer that you can approve.  It is not uncommon for new people here to misunderstand the line between answer and comment, and indeed this point plagues some of us who have posted quite a bit as well.

RoseFire

Korvan:
>2) The problem of February 39th 2000. As a default, every 100
>years is NOT a leap year, although every 400 is

Beside the fact that Feb 39th doesn't exists  :)  , fortunately, the big majority of date-validating routines simply checks if year is multiple of 4 to see if Feb must have 29 days. That's accomplished just by stating   IF (YR MOD 4)=0 THEN Feb=29

Moreover, programs which use last two digits, are still capable of manage it (in fact, to know if a number is multiple of 4, it's enough to check last 2 digits).

I said "fortunately" because, (as you well said) although beginnings of centuries are NOT leap years (1800, 1900), they are when are multiple of 400. Thus, the former check is still useful for Y2K, because 2000 WILL BE a leap year. Of course, it won't work for 2100, but by that time, ¿who cares?...  :)

Viking,

Thanks for pointing out the typo...

The entire issue of millenium "bug" breaks down to a simple question - does both the computer and the software used on it support the year 2000 as the actual year 2000?

In some (most even?) cases of older equipment and software, this is not the case and I have found in my experience that the best test is to actually jump the queue and set your PC's clock forward and see the results for yourself.  Re-setting the clock itself takes very little effort if things do go wrong, by which stage you may have a decent idea of what to expect and, therefore, be able to consider alternatives.

The case is point in my comment was that, as a default, some software has been written with only the first portion of the leap year "rule" in mind ie: every 100 years is not a leap year.

As to my own knowledge of this, I have the advantage of having previously worked with one of the most respected technical writers of our time, Jakub Kaminski (Tech Editor of virus bulletin), in his capacity as leading virus/bug technician at Vet Anti-Virus Software.  He has written an extensive white paper on the subject which was well received at the World Virus Bulletin Conference last year.

Really, the question is whether developers have really prepared themselves for the millenium "bug".  This is not always the case and, as in the case of the Thai Government for instance, some organisations are being forced to set their clocks back to the year 1972(?), this being the last year with an identical calendar.  This is the result of "hard-coded" dates in various electronic componentry, such as radar early-warning systems, and certain pieces of software, such as computer BIOSes.

Thanks to rosefire too :)

I'm not chasing the points here - I simply wanted to give JASON71 the best perspective on the issue as I have known it.  The best you can do, JASON71, is research as many white papers as you can... ;)

korvan, the points obscure the real problem which is that there are no rules for how things work here.  Conventions do exist infomally based on fairness and best serving the people here.   The boundries are very nebulous however.

If we don't tell  people when they clearly  break with the conventions, the will never know.  If you were to make the point that the you really answered first you end up feeling like and looking like you are interested just in points.  That is why I posted a note on your behalf.  Your writeup was very good.  So were the others.

In this case, it looks as though the poster has forgotten you anyway.  I heard that the questions eventually time out (1 month) and the credit will be given to anray since JASON71 has not replied.  I don't know if this is true, but if it is, it means the database at EE could be filled with questions with wrong answers that just timed out. I really don't know if this is the case.

RoseFire

Korvan:
You're right, of course; I've simply mentioned that fact almost as a curiosity.

Jason:
The fact is: our computer can be Y2K compliant, but ¿what about the software we have running into it? Our machine can well tolerate the shift from XX to XXI century with no problems at all, but that says nothing about a certain software, developed by who knows who; if that software evaluates differences between dates just only with last 2 digits of year, and took no other providence regarding that, that soft will surely have problems when be executed on Jan 1st, 2000.

If such a program is running in a bank, and it calculates interests for fixed time deposits, the most probable is that a deposit made on 30-Nov-99, with maturity date at 30-Jan-00, will receive a NEGATIVE interest, simply because 00 is LESS than 99.

And the same problem can be extrapolated to travel tickets, insurance policies, retirement documents, and an ENORMOUS "etc".

This is strange...I wanted to accept Korvan's answer but now I do not have this option? I think all the answers are excellent and refrained from giving points simply to learn more from the comments! Any ideas on how to resolve this points dilemma now that we have solved the millenium bug dilemma?
regards JASON71

You have to reject the current answer and when doing so leave a "comment"  ask korvan to submit an "answer" and you will get the option to accept his.

RoseFire

korvan CAN YOU submit an "answer" and I will get the option to accept yours?

Jason71

I doesn't matter which answer you accept or reject, as long as you, the questioner, are satisfied with the information you receive. :-)

anray, I wish more people here had that attitude.  This is about helping people, not scoring points. At least I would like to think so.

RoseFire

Sorry for taking so long...I wanted to see as many comments as possible. I wished I could give points to everyone.
Thanks Jason!