W3C home > Mailing lists > Public > www-tag@w3.org > January 2003

Re: Less is more (WAS: Clarifying what a URL identifies (Four Uses of a URL))

From: Sandro Hawke <sandro@w3.org>
Date: Thu, 23 Jan 2003 14:45:33 -0500
Message-Id: <200301231945.h0NJjXn24656@wadimousa.hawke.org>
To: Miles Sabin <miles@milessabin.com>
cc: www-tag@w3.org


Miles Sabin wrote:
> Tim Bray wrote,
> > It is quite possible that the Web Architecture works *because* it
> > works around the intractable problems of meaning and only deals with
> > comparing identifiers and shuffling representations around; avoiding a
> > lot of problems that historically have been intractable.
> 
> I replied to the paragraph containing this sentence earlier, but I 
> missed it, until Len flagged it up.
> 
> I believe Tim is absolutely 100% correct. But look Ma, no spooky  
> abstract Resources ... just strings and representations.

Yeah, but it's more like strings and strings.  You invoke GET with a
URI string parameter and, if the operation is successful, you get back
a "MIME entity" string result.  URIs and MIME entities are both
character strings with certain syntactic and semantic constraints
imposed by RFCs [1].  The fact that a picture or a knowledge base or
an XML document can be encoded into a MIME entity does not mean we
should necessarily ignore that encoding step.

You can imagine a simplied web API:

    MIME_Entity e;
    PNG_Image i;

    e = web.get("http://www.w3.org/Icons/w3c_main");
    i = e.decode();       // run time type checking?

and perhaps there's a shortcut:

    i = web.getAndDecode("http://www.w3.org/Icons/w3c_main");

but I'm very hesitant to move directly to the shortcut.

I'd also like to see the term "representation" saved for talking about
relationships ("x is a representation of y", "y has representation x")
rather than classes ("x is a representation", "the type of x is:
Representation").  If you mean Representation as a subclass of
MIME_Entity, then how is it different?  If you mean Representation as
a subclass of Thing, then ... how is it different?

   -- sandro

[1]  RFC 2045: Multipurpose Internet Mail Extensions (MIME) Part One:
     Format of Internet Message Bodies. N. Freed,
     N. Borenstein. November 1996.  (Format: TXT=72932 bytes)
     (Obsoletes RFC1521, RFC1522, RFC1590) (Updated by RFC2184,
     RFC2231) (Status: DRAFT STANDARD) http://ietf.org/rfc/rfc2045.txt

     RFC 2396: Uniform Resource Identifiers (URI): Generic Syntax. T.
     Berners-Lee, R. Fielding, L. Masinter. August 1998. (Format:
     TXT=83639 bytes) (Updates RFC1808, RFC1738) (Status: DRAFT
     STANDARD) http://ietf.org/rfc/rfc2396.txt
Received on Thursday, 23 January 2003 14:47:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:47:15 GMT