RE: [dss] Compound operation Verify & Sign

["Edward Shallow" <ed.shallow@rogers.com>]

Trevor,

   You haven't addressed the embedded versus detached issue, where the
timestamp is detached from the signature. Would you add another response
element for the deatched archive timestamp ?

   Again, are Archive or Freshened TimeStamps in or out ? If in, what is the
FULL schema impact of this included support ?

Ed

-----Original Message-----
From: Trevor Perrin [mailto:trevp@trevp.net] 
Sent: November 4, 2003 2:17 PM
To: Edward Shallow; 'Nick Pope'
Subject: RE: [dss] Compound operation Verify & Sign

At 01:39 PM 11/4/2003 -0500, Edward Shallow wrote:
>Content-Transfer-Encoding: 7bit
>
>Disagree Nick.
>
>The best you can hope for is partial coverage of the common elements. I 
>also believe it is very likely that what you come up with for the "bucket"
>element definition will not suit the profile's needs. I am speaking 
>from experience as I am closley monitoring the ability of the DSS to 
>support an EPM profile extension. Unless you dig into to all the 
>profile's requirements as it pertains to returned 
>signatures/timestamps, you run the risk of introducing constructs which do
not serve the profile's objectives.

That's not terrible though - we can put an option in the core, such as an
<UpdatedSignature> output, and then if profiles don't want to use it, they
don't need to.

So I'm inclined to put a <RequestUpdatedSignature> / <UpdatedSignature> in
the next core draft, since this seems acceptable to people.  Let me know if
I'm reading this discussion wrong.



>Your first sentence below has not as yet been decided and is still 
>subject to debate.

What Nick describes concurs with the wd-04 spec, in fact all the specs since
wd-01, and some time before that.  This approach came out of discussions
with Rich Salz, see the "pizza topping" analogy in the post below.  It had
also been advocated by Tim Moses earlier.
http://lists.oasis-open.org/archives/dss/200309/msg00100.html

No-one objected at the time, as Rich and I produced several schemas based on
this approach.  There was also agreement on the next con-call on this
approach:
http://lists.oasis-open.org/archives/dss/200309/msg00110.html

So we can re-visit this (Ed, were you going to start an on-list discussion
about this?), but this is the direction we'd been going in for the last
several weeks.


Trevor