An excellent post! I found this to be the best post of all the ones I have read on the various Y2K threads and pages. It is this type of reasoned argument that I hoped to find on the internet. Even though I have a different opinion then the writer, his (her) arguments are very well presented.
I am also in the computer business (over 35 years), and even wrote some COBOL programs with only a two position year when I was with IBM. Fortunately, none of my COBOL programs are still in use, the various companies having switched to mini and micro computers.
Your comments about the possible fixes are accurate, assuming that your computer software still has two position years. Many (most) computer software was written within the last ten years, when hopefully responsible programmers and systems analysts took the year 2000 into account. We have a couple of systems that use two position dates, and we simply inserted a few lines of code into each program (a subroutine actually), that says if the length of the year is two, then, if the year is greater then 50, add 1900 else add 2000. This works for most logic, display, and printing functions (assuming it will fit on the screen or report), but you are correct when it comes to sorts and date keys.
We also found that most of our programs that used dates did not do date calculations. In fact, when we discussed having data entry operators key in 2000 instead of just 00 each time, it was decided that this would slow down the data entry time. Therefore, it was only systems like payroll and financial reporting that we had to change.
We fully expect that some other reports, or displays will not work correctly next year, but have decided that we can fix these as they occur since they are not critical applications. This is a lot more efficient then poring though mountains of code trying to find where problems might occur, and perhaps causing more problems by making a bad fix. Computer bugs occur each day as unexpected conditions occur, and they are fixed each day as well.
Your point is taken, however, that there is computer software that must be fixed, I guess we just disagree on the extent of the problem, and the amount of work to fix.
With regard to embedded systems and chips, I am not an electrical engineer, but many of my friends are, since they work at NORTEL in my hometown. They have assured me that unless the chip or system has a method to manually input the year in case of power or battery failure, the chip/system CANNOT be date sensitive. If you accept this argument, it seems to me that those that need checking are relatively few, and can be identified very easily if they have a manual date input function. Someone mentioned EPROMS as perhaps being a problem, but from what I can find out, it is not possible to burn year logic into an EPROM.
I think that all computer gliches can be fixed and handled as they happen, except for essential services such as hydro, water, etc. I take comfort in the fact that every utility ie. Ontario Hydro, and community that has done a complete Y2K test, has reported no serious problems. I think they are ready.
We agree totally on the problems that WILL happen because of the panic that is being spread about the Y2K possibilities. In my opinion, the stock markets will go down as people take their money out, there will be shortages of some products as people hoard etc. I have not participated in many of these discussions, because I am amazed at the vicious reactions I get when I express my opinions. Even though they will not admit it, I believe that many people actually hope there will be problems, either because they sell goods or services to the Y2K market, or have invested in these companies, or have already bought generators, dried food, etc, and do not want to believe that all their preparations were not required.
Anyways, your post is very lucid and well presented. I think we only differ on the scope and impact of the problem. Here is an article written by computer programmers that I also found very reasoned.
ghsport.com
If you have time to read it, I would be very interested in your comments.
Jim
|
