Paul, I think you're mixing up the message:
> Can you explain why GM raised their Y2k-budget from $50 mil. to $500 mil
I am describing my experiences with embedded devices (as a developer). I have worked on many, many different kinds of these systems, ranging from PLCs, to CNCs, to process controllers, general firmware, etc. Of these systems, only a small percentage carry around the date. Of the systems that use a date, most do so only for display purposes and not for calculation. It is my opinion that a device that does not read the year field for purposes of calculation should function no differently on 1/1/2000 than on 1/1/1998. GM's budget notwithstanding (and I'm not talking at all about IT systems here, just embedded systems), it doesn't change the fact that this is my experience as a developer who has dealt with a fairly diverse range of embedded processing systems. And many of these systems, in fact, are either the same or similar to that which GM uses.
> Could you give me an estimate on the percentage of companies
Again, I'm relating my experiences. I suspect that if a department is tasked with checking Y2K compliance of embedded devices they will assume (as they should) worst-case and attempt to verify all devices will function. It is also worth noting that an IT department tasked with working in this area has - in general - little to no experience with firmware devices and embedded systems. Therefore, they have very little experience or knowledge with which to approach the issue.
> Could you give me one single explanation why you do believe
Yes. Management is assuming worst-case (again, as they should)... because management does not know enough about the technology which they've purchased or built (too many MBAs, not enough CS and EEs!).
> Have you ever tried to imagine how difficult it is to coordinate
Again, that's not what my message was about. I am relating my real world experiences with these systems.
Cheers! |
