W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > October 2003

MIME type regostration notes

From: Graham Klyne <gk@ninebynine.org>
Date: Fri, 31 Oct 2003 11:03:38 +0000
Message-Id: <>
To: Brian McBride <bwm@hplb.hpl.hp.com>, rdf core <w3c-rdfcore-wg@w3.org>
Cc: me@aaronsw.com

At 16:25 30/10/03 +0000, Brian McBride wrote:
>ACTION 2002-04-05#2 Aaron
>   complete rdf mime type registration

Here's the current situation as I see it.

The current published draft is [1], dated 11-Sep-2003.

This was released following Aaron's request for review [2], and personal 
feedback from me following informal WG discussion [3].

As far as I can tell, the latest draft was not announced on the ietf-types 
mailing list [4], so (unsurprisingly there has been no comment).

I have re-reviewed the latest version (just now), and note the following 
minor points:

(a)  I think this is OK, but I note that a charset parameter is noted, with 
use per application/xml.  I mention this because there has been some recent 
discussion [5] about the undesirability of charset parameters on XML MIME 
types.  I think that deferring to application/xml is the right thing to do, 
but if this is likely to be an issue, we might add a note to the effect 
that use of the charset parameter is discouraged.

For now, I suggest:  do nothing.

(b)  Small matter.  There the registration says:

       Encoding considerations:

          Same as charset parameter of application/xml.

maybe it would be more appropriate to say just:
       Encoding considerations:

          Same as application/xml.

(c) I note that the document title for citation [3] is incorrect -- it 
should be "RDF Semantics".  This may be an out-of-date xml2rfc BibXML 
module ... I note that the current version of BibXML from xml.resource.org 
has the correct name for the latest draft release.  (BTW, these citation 
files for W3C documents are automagically generated from W3C's own RDF 
published details.)

(d) RFC2119 is mentioned in the reference list, but never actually cited in 
the document.

These are all nits which can probably be sorted out in the publication process.

Item (c) raises a question of timing.  Should we hold off requesting 
publication until the final RDF drafts are published?  I think not:  RFC 
publication is a lengthy process, and updating the references to the final 
versions should be easy enough, and should be clearly seen as merely 
editorial changes.  It might be worth adding a "Note to RFC editor" to 
check for the final published document details.  (This is common practice 
when publishing sets of RFCs.)

So where from here?  I suggest:
(1) remind folks on the ietf-types list [4] that a revised document is here 
for review.
(2) assuming no adverse feedback, request IESG approval for RFC publication.


     Also, HTML at

[2] http://eikenes.alvestrand.no/pipermail/ietf-types/2003-July/000073.html

[3] http://eikenes.alvestrand.no/pipermail/ietf-types/2003-July/000075.html

[4] http://eikenes.alvestrand.no/pipermail/ietf-types/

[5] http://eikenes.alvestrand.no/pipermail/ietf-types/2003-October/000092.html
     et seq

Graham Klyne
For email:
Received on Friday, 31 October 2003 06:36:09 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:54:08 UTC