W3C home > Mailing lists > Public > www-tag@w3.org > March 2012

Re: Classification of ISSUE-57 change proposals

From: Roy T. Fielding <fielding@gbiv.com>
Date: Fri, 30 Mar 2012 22:34:11 +0200
Cc: Noah Mendelsohn <nrm@arcanedomain.com>, Jeni Tennison <jeni@jenitennison.com>, "www-tag@w3.org" <www-tag@w3.org>
Message-Id: <2A188DA4-CBF6-48A0-89B1-CC4932D85E90@gbiv.com>
To: Jonathan A Rees <rees@mumble.net>
On Mar 30, 2012, at 5:56 PM, Jonathan A Rees wrote:

> Hmm... I agree in spirit, but as an aside I quibble with the phrasing
> of the last part, and since terminology is really screwing us over I
> want to hit what you say hard. "Representation" remember is a war
> zone, with Roy using it one way (see REST and HTTPbis) and TimBL using
> it another. It is an instrument of propaganda. In Roy's view a
> description could very well be a representation; in TimBL's that is
> not enough. AWWW sits on the fence, not really taking a position,
> while httpRange-14 has been understood as saying all representations
> (or at least any representation that is retrieved) are to be content,

That is not true.  httpRange-14 says that if the resource owner cares
to make a distinction, here is how they can make that distinction.
It does not say the distinction is necessary.  The assumption that
200 implies IR is pure speculation, which is fine if you care to
speculate, and can be ignored by those who do not speculate.

> so the scene is very confused. So be very careful of terminology. (In
> my opinion Roy has been inconsistent in agreeing to httpRange-14 and
> then turning around and putting an incompatible definition in
> HTTPbis.) I tend to give Roy priority in use of "representation",
> since REST is so well known and HTTPbis will have such prominent
> standing, and use "content" or "instance" in TimBL's preferred sense
> (i.e. encoding or shared properties, see note "generic resources and
> web metadata" which I annoyingly always trot out here, for what I
> think is meant). So here I think it would be better for you to say
> "content" rather than "representation".

Only if you prefer to be incorrect.  HTTPbis defines how HTTP works.
If you make imaginary assumptions about the state of the universe and
then expect those assumptions to be enforced by HTTP, then you will
be disappointed by HTTPbis.  Prepare to be disappointed.  HTTP does
not prevent resource owners from behaving in a way that fits this
imaginary universe, nor does it prevent them from behaving in ways
that do not fit in that universe.  It merely defines the interface.

Of course, it would help if folks stopped insisting that their's
is the only imaginary universe.  The Web is defined by human interaction,
not by your rules.  If the interaction seems weird, that's your problem.
Likewise, HTTP interactions involve URIs, resources, representations,
and *methods*.  If your model of the Web does not account for methods,
then it is hopelessly broken anyway.

....Roy
Received on Friday, 30 March 2012 20:34:38 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:33:14 UTC