W3C home > Mailing lists > Public > w3c-wai-er-ig@w3.org > October 2001

Re: Where does the EARL go?

From: Charles McCathieNevile <charles@w3.org>
Date: Fri, 19 Oct 2001 15:33:33 -0400 (EDT)
To: Jim Ley <jim@jibbering.com>
cc: <w3c-wai-er-ig@w3.org>
Message-ID: <Pine.LNX.4.30.0110191523060.27106-100000@tux.w3.org>
On Fri, 19 Oct 2001, Jim Ley wrote:

  How does that work with a web site, a unique earl report for each url on
  the site, how does that work with urls with query portions?  How can EARL
  exist unless this has already been decided, how EARL is used is much more
  important than what the tags look like.

EARL requires a URI to determine what an EARL assertion is about as part of
any valid piece of EARL.

Therefore it is not necessary that a page identify the EARL that is about it,
but it is necessary that we determine how to find EARL about a given piece of
content in a useful way. This is a problem that has been dealt with before,
in at least the following W3C Technologies

  Annotea
  PICS ratings bureaux
  P3P
  Photo-RDF

with a variety of solutions. The fact that a single piece of EARL can be
about any collection of resources or pages (including a subset of a page)
and the fact that it is liekly to come from multiple sources suggests to me
that we should look for a solution that doesn't rely on a well-known filename
scheme - the Annotea model looks promising to me. It is similar to the PICS
model, added to which it has built-in support for handling RDF content, and
could even cope with the difference between N3 and RDF-XML, but I am not sure
how it deals with multiple different servers.

I will take an action item to (again) go to the Annotea folks and ask them
about this scenario - this time including the question of using multiple
servers and scalability. There are other people working on RDF who are also
looking at different solutions for this problem in different contexts - I may
be lucky and stumble on some.

  > For now, all EARL is going to be served as text/plain or text/xml, so
  just
  > leave the "type" attribute off. It's only advisory.

  Why is EARL served as text/plain

For now all earl is written as N3 or RDF-XML, and neither of those has a
registered MIME type. Either we decide that we are going to use
text/x-something or that we are going to wait for RDF to have a registered
mime type, but "for now, all EARL ...etc". As we are aware, this needs to be
resolved, but this should be resolved in line with the approach taken by the
RDF groups as a whole.

Charles
Received on Friday, 19 October 2001 15:36:50 GMT

This archive was generated by hypermail 2.2.0 + w3c-0.30 : Thursday, 9 June 2005 12:10:39 GMT