> You asked, "Since we had virtually no GPS problems (except for some cars in Japan), do you feel more confident now that the embedded chip "problem" will not be serious"
>> Yes and no.
So, overall, I do feel better about the Y2K embedded system problem, but I am still very apprehensive. Final results from the GPS rollover aren't in yet, but the preliminary results do look good.
There's a very large difference between the Y2K and GPS problems I haven't seen stressed.
Y2k was the result of a design tradeoff decision 40
years or so ago, and we are faced with the lack of date
format standards, late start in converting, and large
amount of old code to be fixed.
GPS rollover is very different. The 1024 count week
field was in the initial GPS transmitted data format,
and only an incompetent programmer would fail to
design in rollover handling into his design. Granted
that some of the designs were not designed properly.
The fact that some units were not designed correctly
was known for a long time, and affected users had
plenty of time to determine if their equipment was
certified rollover ready, and replace it if it wasn't.
Even where GPS was used to provide timestamps to
large mainframes, it was generally an external box,
easy to replace. Very different from searching millions
of lines of source code.
The fact that GPS rollover went very well does not
indicate that Y2K will also go very well; it, of course,
surely doesn't indicate that Y2K will NOT go well.
There's no connection between the two events. |
