Re: Error in merlin-iaikTests-two.tar.gz?

Joseph Reagle <[email protected]> Fri, 24 May 2002 10:02:56 -0400
Newsgroups gmane.ietf.xmldsig
Organization W3C
Message-ID <[email protected]>
On Friday 24 May 2002 04:36, Christian Geuer-Pollmann wrote:
> I guess, NOW I got it. First of all, I can verify Merlin's signature now.
> Just to check that I understood it right:
>
> If I c14nize a document subset, I do only have to make this weird xml:*
> copying magic if the current element is an "orphan" node (to use this
> exclc14n lingo term), i.e. if the parent of the element is not in the
> document subset.
> If the parent of the current element IS in the document subset, I do NOT
> do the copying of all xml:* attrs, right?

Yes, *and* when you do copy all the xml:* from the ancestors because of a 
missing parent element you go *all* the way up to the root, not just up to 
the next ancestor element in the subset. Counter-intuitive but the spec is 
clear once you get over your intuitions -- I had the same problem! [1]

[1] http://mail.python.org/pipermail/xml-sig/2002-April/007665.html