RE: Last Call for C14N 1.1 (and updated WG notes)

 

> -----Original Message-----
> From: Joseph Reagle [mailto:reagle@mit.edu] 
> Sent: Thursday, 2006 December 21 10:27
> To: Grosso, Paul
> Cc: w3c-ietf-xmldsig@w3.org; www-xml-canonicalization-comments@w3.org
> Subject: Re: Last Call for C14N 1.1 (and updated WG notes)
> 
> On Wednesday 20 December 2006 12:29, Grosso, Paul wrote:
> > The XML Core WG announces the publication of and requests 
> > review of the Last Call WD of:
> 
> I have read through the document online but in no way feel 
> technically 
> competent to do anything other than ask questions and make tentative 
> suggestions.
> 1. Would it be useful at the very start of the document to 
> simply say that 
> changes from the version 1.0 are addressed in section 4?

I should have announced "diff" versions earlier, but I have
now just done that (previous email).

By the way, the entire section 4 is unchanged in C14N 1.1 from 
C14N 1.0.  Most of the changes are in section 2.4 (perhaps that
is what you meant).

> 2. I wonder what problem the XML declaration was posing for 
> the decision to be made to remove it?

I don't understand this comment.  If you are referring to 
section 4.1 No XML Declaration, that section (in fact, the 
entire chapter 4) is unchanged in C14N 1.1 from C14N 1.0.

> 3. My memory is fuzzy, but I hope to internationalization people will 
> comment on "new character model normalization" as I believe 
> that was a concern to them in the first round.

I don't understand this comment.  Are you referring to 
section 4.2 No Character Model Normalization?  If so,
that section is unchanged in C14N 1.1 from C14N 1.0.

> 4. "Note that an argument similar to this can be leveled 
> against the XML 
> canonicalization method": you are referring to version 1.1 or 1.0? 

That sentence (in fact, the entire chapter 4) is unchanged
between C14N 1.0 and the latest draft.

> 5. "The C14N-20000119 Canonical XML draft alternated": here you are 
> referring to the causation method that belonged to the XML 
> working group, 
> before the signature took it over. It would probably be helpful to 
> disambiguate and be really clear at all times about which 
> version is being 
> spoken of and provide context if appropriate. Also, the 
> casual reader might 
> also wonder how this affects exclusive canonicalization. (I 
> presume the 
> answer is not at all, but it might be worthwhile to provide an 
> understanding of the landscape.)

That sentence is unchanged between C14N 1.0 and the latest draft.

The XML Core WG was not involved in the writing of C14N 1.0.

Our charter for C14N 1.1 was to fix the problem in C14N 1.0
with inheritance of xml:id and xml:base.  We made a conscious
effort not to make changes in other areas of the spec.

> 
> I would like to ask that when it comes to interoperability and 
> implementations the XML signature applications be a necessary 
> part of that 
> process. That is, it would not be sufficient to show adoption and 
> interoperability between two XML applications that are not 
> used as part of signature processing.
> 

We've purposely given C14N 1.1 a long LC period so that it will
overlap with other related WGs and give them time to review it.
We haven't even considered CR period yet.  We will certainly
want XML signature implementations at that time.

paul

Received on Thursday, 21 December 2006 19:42:45 UTC