W3C home > Mailing lists > Public > public-xmlsec-maintwg@w3.org > June 2007

Re: DNAMES Section4.4.4.htm

From: Thomas Roessler <tlr@w3.org>
Date: Mon, 25 Jun 2007 10:00:08 -0700
To: Konrad Lanz <Konrad.Lanz@iaik.tugraz.at>
Cc: XMLSec <public-xmlsec-maintwg@w3.org>
Message-ID: <20070625170008.GR3883@raktajino.does-not-exist.org>

On 2007-06-20 20:16:39 +0200, Konrad Lanz wrote:

> * The lower case "should" should be a "MAY".

So updated in the Editor's Draft.

> * We should consider referencing RFC 2253 and also RFC 4514 (after checking 
> if they are no collateral effects).

So updated in the Editor's Draft.

> * Do we run into an "internationalization problem" when
> referencing the grammar of RFC 2253 / RFC 4514 ?

Thinking about this more, I'm inclined to believe that we don't run
into that problem -- XML is defined in terms of characters, the
grammar is defined in terms of a particular representation of these
characters, and if there is no UTF-8 encoded representation of the
string that was present in the DNAME, then the dotted-decimal form
can be used.

> I took a shot at the current red-line document and did some word
> smithing (please refer to the attachment).

Whooops, I had missed that before editing the draft.

>      * Since a XML document logically consists of characters, not
>	 octets, the resulting Unicode string is finally encoded
>	 according to the character encoding used for producing
>	 the physical representation of the XML document[DEL: .
>	 :DEL] [INS: implying that [8]character references could
>	 be intruduced. Due to [9]immediate expansion these are
>	 transparent in a canonicalized XML content or on an
>	 application layer.  Note that implementations MAY just as
>	 well rely on escaping such characters as allowed in
>	 section 2.4 of RFC 2253 [[10]LDAP-DN]. :INS] 

Ugh, that seems to make things more complicated than they need be.

> @@@Konrad to Thomas: if DNAMES are usewd in content that is not
> canonicalized before signed, this is useful and also
> implementations may also decide to escape all characters beyond
> ASCII @@@ @@@Konrad to Thomas: Some word smithing from your side
> may be required here. @@@ @@@Question to all: Is it possible that
> characters are introduced via [11]immediate expansion that cannot
> be represented in UTF-8 and if so would these be . @@@

I'd actually suggest that we move the "Since a XML document..." part
out of the bullet points (and therefore out of the scope of the
MAY), since (a) it doesn't actually introduce another conformance
requirement, and (b) it applies whether or not the augmented
processing rules are used.  I've made that change in the Editor's
Draft.

Cheers,
-- 
Thomas Roessler, W3C  <tlr@w3.org>
Received on Monday, 25 June 2007 17:00:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:22:00 GMT