RE: httpRange proposed text

> >> The proposed text offers no advice about *how* to guarantee the
> >> unambiguous identification characteristics of a URI.  It simply
affirms
> >> that this is the *purpose* of URIs, and recommends that people
choose
> >> URIs which do not conflict with that purpose.
> 
> There's a contradiction here: the proposed text either offers no
advice,
> *or* it advises that people choose URIs not to conflict with this
> purpose.  In the first case ("no advice") it doesn't belong in the
> architecture document, which exists to offer advice on the web
> architecture, what to do and what not to do.

There is no contradiction.  There are two independent issues:

A) URIs intended to be unambiguous identifiers, and people are advised
to respect that.
B) People are advised to avoid X, Y, Z because that would violate
principle "A".

Like you, I think it's a good idea to "dodge" B, at least until we have
something other than people's opinions to give credible advice.

However, I think it is of paramount importance to stress the advice on
issue "A".  This principle falls into the category of "don't even mess
with it."  It's the axiom upon which all other axioms rest.  It is even
more fundamental than the "don't create new schemes" advice.  

And, like Tim, I think that this fundamental axiom of web architecture
is all too easy for people (including myself) to forget.  In my opinion,
the proposed text should be the very first advice introduced in any web
architecture doc.  This text shouldn't in any way be controversial; I am
not aware of anyone who would disagree with it, and clearly it would be
silly to even talk about "B" without having agreement on the proposed
text.

So yes, I think it belongs in the web architecture doc, and in fact is
essential for setting context and establishing the most basic advice
under which web participants need to operate.

Received on Tuesday, 30 July 2002 16:42:27 UTC