Home » Technology » Scientific Computing » All I Wanted for Christmas Was an Abacus

All I Wanted for Christmas Was an Abacus

© Adam Hughes

I've put up a strong fight and weathered several months of the onslaught, but I've finally broken down: my wife and I went to the store the other day and purchased several gallons of water, enough canned tuna to feed the neighborhood cats for weeks, and we've even mapped out a nice warm spot in the basement to sleep should the New Year's Grinch decide to steal our electricity. Yes, we are now Y2K compliant, much to my dismay.

In my work, I've seen the day-to-day preparations for the coming "doom", and I've been shown how to safeguard several computer systems and software packages from the fall. So, from a rational point of view, it seems that there has been a lot of hoopla about something that will have little impact on our lives. However, it's that little nagging doubt somewhere in the recesses of my mind that has me scrambling right along with everyone else. And, as might be expected, these concerns don't apply solely to my personal life. There are issues to be addressed in the world of scientific computing, too, although I have a good bit more confidence in the precautions taken in this field than what might be happening in some segments of the commercial world.

In general, scientific code is pretty much bullet-proof when it comes to problems having to do with date recognition. This is primarily due to one simple fact: there are no dates involved in the algorithms used in these software packages. Of course, there are exceptions, the most notable of which is the dates attached to file handles. Some programs have input and output functions which may operate based on the date of existing files.

For instance, a program may automatically delete or erase a file that was created before a certain date. It's easy to imagine such a "scrubber" program working through a directory full of files looking for anything older than, say, 12/01/99. So if it saw a file dated 01/02/00, it's possible the file would be inadvertantly deleted.

This situation could potentially cause some problems, but it is a pretty rare case. Generally, a program either writes over old files or it doesn't, and this is something that researchers have to determine for each program they already use. The smart ones do careful bookkeeping and regularly back up their important data anyway.

While file dates could possibly cause some disruption to the computational scientist, the assigning of these dates is typically a function of the operating system of the computer on which the code runs. Because of this situation, and the lack of the use

The copyright of the article All I Wanted for Christmas Was an Abacus in Scientific Computing is owned by . Permission to republish All I Wanted for Christmas Was an Abacus in print or online must be granted by the author in writing.