🚀 go-pugleaf

sitze@nmsu.edu (Richard Sitze) writes:
> While I agree in principle, I maintain that byte ordering is
> non-obvious in certain circumstances...

While it may not be obvious that the correct solution is a canonical
data format (it is not even true in all circumstances), it is patently
obvious that such an option exists.  As a case in point, consider the
following excerpt:

	"...Nor does LOCUS convert network messages to a canonical byte
	order, in the belief that much message traffic will be amoung
	like machines..." The Locus Distributed System Architecture, G.
	Popek and B. Walker, MIT Press, 1985.

The decision as to how to communicate between machines with different
byte orders was made between 1980 and 1982, and the phrasing of the
quote suggests that the authors believed the canonical format so obvious
that they had to justify a different choice.

As a point of historical reference:

> This is analogous to advocating that all books be
> translated into English before they get put into the library, and when
> you check the book out it gets translated into your native tongue.
> The OBVIOUS solution is print (store) books in a variety of languages,
> and then check out the appropriate book...

This OBVIOUS solution that you mention trailed the other by many
centuries.  During the middle ages, the language of learning was Latin.
Important ancient books were translated into this canonical (of the
canons) language, and all important books and documents were written in
it.  It was only with the advent of the printing press and the rise of
vernacular tongues as vehicles of knowledge that the OBVIOUS solution
you mention became the norm.

--bob--