Re: 12, 14.43: resource arguments and conneg

To begin with, I have yet to understand what, if anything, is actually
wrong with the current wording.  Maybe I'm dense, but I haven't clued
into the problem you seem to see (doesn't mean the problem isn't real,
just that I haven't seen it).  Somehow I always expect a response
to vary on the URI, Method and entity; it doesn't help to state the
obvious if it makes the non-obvious inscrutable.  At most, the current
words take certain things for granted, that I expect most would take
for granted.  And there hasn't been a great cry of confusion from
others reading the current text.

Now that aside, your suggested additions add not one, but three terms
to the terminology morass.  The definitions are long, and opaque, and
circular (defining terms in terms of items in this document is a Bad
Idea IMHO; we've worked hard to try to get the current terminology
definitions non-circular.)  For example, your definition of required
variation header requires understanding of the entire document (to
understand the MUST requirements).  This is worse than any disease the
current text has.

I read your rewrite of Vary, and find it opaque.

Again, if there is a problem that causes trouble in implementation
we have time between proposed and draft standard to work further
on the words.

			- Jim

Received on Sunday, 2 June 1996 17:27:54 UTC