W3C home > Mailing lists > Public > public-rdf-in-xhtml-tf@w3.org > September 2007

Re: Fine-tuning CURIEs

From: Niklas Lindström <lindstream@gmail.com>
Date: Thu, 13 Sep 2007 16:04:22 +0200
Message-ID: <cf8107640709130704w1fa69f8epf8a5e72618645f9@mail.gmail.com>
To: "Shane McCarron" <shane@aptest.com>, "Ivan Herman" <ivan@w3.org>, "Mark Birbeck" <mark.birbeck@formsplayer.com>, "W3C RDFa task force" <public-rdf-in-xhtml-tf@w3.org>

Hi all,

I just quickly read through this thread, so I'm sorry if my remarks
restates things you already mentioned.

First, this remark by Shane is really important:

> You ask if the DTD permits what you did above.  It *does*, but what you
> did above would have a really strange effect since it would put the "p"
> element into that namespace too.  [...]

So, e.g. this:

  <link rel="openid.server" xmlns="http://openid.net/"
     href="https://api.screenname.aol.com/auth/openidServer" />

means that it is not an xhtml:link element at all, but an
"openid:link" element. Any XML-aware technology would at best ignore

Futhermore, and to the main point, isn't the case that non-prefixed
values in @rel are *supposed* to be handled specially -- (some of it)
via preprocessing (GRDDL). The XHTML spec says [1] that LinkTypes
(which values of @rel are) have a bunch of *case-insensitive*
predefined values, and then:

    "Authors may wish to define additional link types not described in
this specification. If they do so, they should use a profile to cite
the conventions used to define the link types."

I figure this, depending on @profile, could mean many things,
translating different non-prefixed values into URIs in different
vocabularies (some in DC, some in OpenID etc). They can now also be
implemented via GRDDL. Can RDFa really get any closer without going
against this (admittedly complex) contract? My current opinion is that
they should be left uninterpreted (for (h)GRDDL to catch if at all).

Granted, it seems reasonable that any XHTML1.1+RDFa processor should
detect the predefined names (only!) and generate triples for them --
translating them into URIs by doing <http://www.w3.org/1999/xhtml> +
perhaps '#' or '/' + (lower-cased?) term. My interpretation of this
thread is that you're agreeing on this?

Therefore: no automatic triples from non-prefixed names other than the
predefined. I'd also like to suggest that non-prefixed names are
*only* allowed in @rel/@rev. This since these attributes where defined
before RDFa, and unlike @href and @src they aren't taken "verbatim",
but extended. I'd say RDFa implements the last note about LinkTypes:

    "Note that in a future version of this specification, the Working
Group expects to evolve this type from a simple name to a Qualified
Name (QName)."

I hastily wrote this though, so please correct me if I've missed or
confused anything.

(Note: the clarification Mark does of how the prefix handling is done
looks really good and necessary. Not the least since if not, at least
the presence of @xmlns:_ would mess  up bnodes..)

Best regards,

[1] <http://www.w3.org/TR/xhtml-modularization/abstraction.html#dt_LinkTypes>
Received on Thursday, 13 September 2007 14:04:27 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:01:52 UTC