I am surprised that Paul Hyett (aj 30.4.98) should be confident that his building meets year 2000 requirements, as even the computer industry cannot agree what does and what does not comply.
Information Week 15-28 April carried an article which reported that the National Computing Centre (ncc) disagreed with the Computing Services and Software Association (cssa) over its published definition of compliance, which was considered too loose. bsi published its definition (pd 2000- 1) in January 1997, but the cssa says that 'it is too tight to be useful', meaning that its members are finding it difficult to prove compliance to their customers. As the problem has been caused by the computer industry, it is a matter of grave concern that it has not yet put its own house in order - why should building professionals, with only a passing knowledge of computing, be asked to place their pi insurance at risk by warranting things over which they have little control?
The problem is complex, and relates to the data held in programs as well as the software and firmware within the computers. Also, some embedded chips are used in a number of ways and will be compliant in one case, but not in another - I would never guarantee that anything is 'Y2K compliant' until after the millennium (and the following leap years) have passed. All we can say is that some tests have been carried out and passed.
The only thing which is certain is that nobody can be confident of whether their installation is or is not compliant.
Building Performance Group, London WC2