W3C home > Mailing lists > Public > public-grddl-wg@w3.org > April 2007

RE: #xmlbase [1234]

From: Chimezie Ogbuji <ogbujic@ccf.org>
Date: Thu, 26 Apr 2007 17:19:43 -0400
To: "Clark, John" <CLARKJ2@ccf.org>
cc: "GRDDL Working Group" <public-grddl-wg@w3.org>
Message-ID: <1177622383.28278.97.camel@otherland>

The dialog on this is become rather philosophical, which leads me to
conclude that there needs to be better guidance than just saying "check
out this RFC specification and this XML base specification; all your
answers are there".

The crux of the problem, for me, is the idea that if I author a document
with an xml:base *I* embedded at the root element then that base URI
should be what empty relative URI references in a derived GRDDL result
should be resolved against - which is why I think (5.1.1) Base URI
embedded in content is relevant here.

This is not the case as an xml:base on the root element will only
associate a  base URI with the root element not the 'document' or
'document entity'.  So though I can control base resolution of relative
URIs within the original document, I *cannot* control the resolution of
URIs in the GRDDL result (when it does not specify it's own base URI).
This GRDDL result, however, is *supposed* to be a faithful rendition of
the source.  There goes that guarantee.

Without such a guarantee, the GRDDL result is not a faithful rendition
as it would be making statements about a URI that is not determined by
the author, unless the author is also the publisher, in which case I can
be certain about the URI from where the source is retrieved. 

Note that in a situation where the URI from where the source document is
resolved differs from the base URI of the root element / node in that
same source document, the issue-base-param resolution would conflict
with a faithful rendition as the 'source URI of the source document'
would be determined by the protocol not the author.

More below..

On Thu, 2007-04-26 at 14:47 -0400, Clark, John wrote:
> > 
> >    |  .----------------------------------------------------.  |
> >    |  |  .----------------------------------------------.  |  |
> >    |  |  |  .----------------------------------------.  |  |  |
> >    |  |  |  |  .----------------------------------.  |  |  |  |
> >    |  |  |  |  |       <relative-reference>       |  |  |  |  |
> >    |  |  |  |  `----------------------------------'  |  |  |  |
> >    |  |  |  | (5.1.1) Base URI embedded in content   |  |  |  |
> >    |  |  |  `----------------------------------------'  |  |  |
> >    |  |  | (5.1.2) Base URI of the encapsulating entity |  |  |
> >    |  |  |         (message, representation, or none)   |  |  |
> >    |  |  `----------------------------------------------'  |  |
> >    |  | (5.1.3) URI used to retrieve the entity            |  |
> >    |  `----------------------------------------------------'  |
> >    | (5.1.4) Default Base URI (application-dependent)         |
> >    `----------------------------------------------------------'
> > 
> I don't think that it makes sense to think of the GRDDL source document
> as an 'encapsulating entity' for a GRDDL result associated with that
> document.  

This is how I interpret the faithful rendition assurance.  If I say
'hello' in english, and someone translates it to 'hola' in spanish then
the translation is derived from 'hello' and the equivalent 'meaning' is
encapsulated within.  

> If this association held, what would be the context for
> encapsulating a GRDDL result within its GRDDL source document?  Would
> the context be the document, or would it be some element within that
> document?

It *should* be the document as that is what the GRDDL result is a
faithful rendition of.  Unfortunately this does not help the author
achieve his/her 'faithful rendition assurance'

>   XML Base provides us with (potentially) different base URIs
> for the document and for elements within that document; 

Yes, this is the main problem especially where there is this difference
and the difference has an impact on the GRDDL result.  

> this is defined
> in section 4.2[0].  If the context will be an element, do we choose the
> element where the GRDDL transformation for a particular GRDDL result is
> referenced?  

Since the document wont work, a better alternative is to have the
context be the outermost element that XML Base will permit the author to
append a base URI to: the root element.  The normative statements in the
specification consistently speak about the root node and descendants of
the root node only.  In particular, the general rule for GRDDL in XHTML
seems to suggest this:

..f or each a and link descendant element E whose relattribute[HTML4]
has transformation as one of its space separated values the resource
identified by the absolute form of the href attribute with respect to
the base IRI of E ..

If the xml:base is placed on the root, then the base IRI of E is
'inherited' from the base URI of the root Node (the value of xml:base),
*not* the base URI of the 'document' (as determined from the protocol).
Shouldn't URI resolution within a source document and URI resolution
within it's faithful rendition happen against the same base?

> This may provide different base URIs for portions of GRDDL
> results produced by GRDDL transformations in an XHTML document, so what
> do we do about XHTML documents? 

This ambiguity goes away if you consistently assume the root node in the
source document is the context for the GRDDL result: there is only *one*
root node.  For XHTML documents without any xml:base, this is straight
forward and we have test cases for this.  XHTML documents *with*
xml:base is a different matter.

>  Finally, since a GRDDL result is a
> faithful rendition of the source document, wouldn't that make it an
> alternative to the source document rather than encapsulated within the
> source document?

I don't read the faithful rendition this way.  The GRDDL result is
generated from content in the source document and makes statements about
data embedded in the source, so it is not an alternative.

> Because of my statement above, I think that we do get to 5.1.3, so we
> should use the retrieval URI of the GRDDL source document as the base
> URI of its GRDDL result(s).  I would not be opposed to the reading that
> Chime gives here, but I think some of the questions I raise might first
> need to be dealt with in the specification.

I don't think there is any ambiguity in the specification mostly because
of exactly how it calls out resolution WRT to baseURI's *within* the
content of the source document.  I also think an implicit assumption
that the base URI you use to resolve URI references in the source is the
same for resolution in the GRDDL result can be derived from the
expectation that the GRDDL result is a faithful rendition.

Chimezie Ogbuji
Lead Systems Analyst
Thoracic and Cardiovascular Surgery
Cleveland Clinic Foundation
9500 Euclid Avenue/ W26
Cleveland, Ohio 44195
Office: (216)444-8593


Cleveland Clinic is ranked one of the top 3 hospitals in
America by U.S.News & World Report. Visit us online at
http://www.clevelandclinic.org for a complete listing of
our services, staff and locations.

Confidentiality Note:  This message is intended for use
only by the individual or entity to which it is addressed
and may contain information that is privileged,
confidential, and exempt from disclosure under applicable
law.  If the reader of this message is not the intended
recipient or the employee or agent responsible for
delivering the message to the intended recipient, you are
hereby notified that any dissemination, distribution or
copying of this communication is strictly prohibited.  If
you have received this communication in error,  please
contact the sender immediately and destroy the material in
its entirety, whether electronic or hard copy.  Thank you.
Received on Thursday, 26 April 2007 21:20:01 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:11:49 GMT