A great man once said, "That of which we can not speak we must pass over in silence." The authors would have done well to follow that advice. Even though their conclusion is correct.

As Rapporteur of the OSI Model for most its life, I should put a few things right. Although it would take a book to do it right. O, that's right I have written a book on what is next: Patterns in Network Architecture from Prentice Hall.

The OSI Model began as a productization of the Internet based on the recommendations of INWG, Cerf, et al. 1978. It was not just X.25. The lower 4 layers were reasonably solid and an advance over the current Internet. By 1983, we had determined that the upper 3 layers were a single layer and modified the standards to allow such an implementation by the clueful. Even so there were still fundamental flaws with the model (see Chapter 4 of Patterns).

I teach networking at Boston University and I do not cover the OSI Model in any detail (only in terms of what one can learn). The main thing we can learn from the OSI Model is not invite the legacy into any new effort.

Our knowledge has moved much beyond OSI. Although that knowledge is unfortunately not represented by the Internet, which has more in common with DOS than UNIX.

It is high time we moved beyond the world of the current legacy and quit throwing good money after bad.