|
|
|
|
|
|
|
is the avoidance of another nonnegotiable deadline date. The disadvantages are many. The bugs that may be introduced are particularly difficult to track down. The temporary patch mentality is perpetuated as well. This example makes the moniker of sliding window for this technique easily understood. |
|
|
|
|
|
|
|
|
The single advantage of the bridge is that it may solve the single immediate problem of interpreting "00" as 1900, where this situation is a problem. There are numerous disadvantages to the bridge, however: |
|
|
|
|
|
|
|
|
It does not solve the real problem. The real problem is representation of years by two-digit numbers. The bridge simply moves the problem around, without solving it. Files that contain two-digit years before the fix will still contain two-digit years after the fix. |
|
|
|
|
|
|
|
|
The bridge is specifically designed, in fact, to avoid addressing the problem of correcting files and record layouts. |
|
|
|
|
|
|
|
|
The bridge may introduce additional problems. If the date range of the application includes valid dates in, for example, 1935, these dates will be erroneously interpreted as 2035. |
|
|
|
|
|
|
|
|
The bridge is a characteristic product of the myopic management mentality that produced Y2K in the first place. Once the immediate Y2K bullet has been dodged, the matter will be forgotten, as "more important" matters take priority. A few years later, the same problem, now the "2020 bug," will come as a complete surprise to another batch of managers. Identification of the problem code will be much more difficult, because it has been complicated by millions of manually installed patches. Designing yet another patch will be more difficult, and installing the patches in a corpus of software that has had another 20 or 30 years to grow will dwarf the cost of Y2K. |
|
|
|
|
|
|
|
|
You should also note that the year 2000, as a deadline, is the same for everyone, so there is a sense that everyone should pool resources to deal with a common problem. As different organizations implement different bridge solutions, however, the horizons will be different, and fewer resources will be brought to bear on what will be perceived as local problems. |
|
|
|
|
|
|
|
|
Use of the bridge guarantees that you not only have to fix Y2K, but you also get to repeat the entire crisis on an even larger scale. |
|
|
|
|
|
|
|
|
There is no single offset value that works for all applications. There is not even a guarantee that the technique will work at all for a given application. |
|
|
|
|
|
|
|
|
The major cost of Y2K is not fixing the bug, it is identifying, locating, and testing the buggy code. As a solution technique, the bridge does nothing to avoid these costs. Having found the problem, however, the bridge does not fix the problem, but rather creates a more complicated problem, which will then have to be found again. The bridge incurs the bulk of the expense of a real solution, but does not provide one. |
|
|
|
|
|