Re: Soap Message Canonicalization (SM-C14N)

I think so; the scope of this entire discussion *seems* to be "things
intermediaries do with the parts of messages which they aren't given
any authority over (by targeting)." If a block's semantics are
"rearrange the rest of the message by the phase of the moon," and
it's targeted at me, I'm free, no, expected to do so.

It would be nice if there were some framework to talk about how to
re-insert blocks, even if it's just some prose to reuse in a module

On Wed, Feb 20, 2002 at 11:34:23AM -0800, Henrik Frystyk Nielsen wrote:
> >The current spec does require header blocks to be forwarded in
> >the same 
> >order they were received: "Such relayed SOAP messages MUST contain all 
> >SOAP header blocks and the SOAP body from the original SOAP 
> >message, in 
> >the original order, except that SOAP header blocks targeted at 
> >the SOAP 
> >intermediary MUST be removed (such SOAP blocks are removed 
> >regardless of 
> >whether they were processed or ignored)." Unless we change this rule I 
> >am not convinced that we need sorting in the C14N algorithm.
> I think it was me who brought up the concern of oversimplifying the need
> for preserving order. The current requirement is very strict and I think
> will lead to problems in handling expectations of what intermediaries
> can do. What about saying something like the following... I think it is
> a compromise proposal that should clarify what intermediaries are
> allowed to do with respect to ordering:
> * A SOAP intermediary must not change the order of header blocks NOT
> targeted at it.
> * There are no ordering constraints for "re-inserted" header blocks.
> That is, if one uses "repeated header blocks" then an intermediary is
> NOT required to re-insert them in the same order as they were received
> nor necessarily in the same location in the SOAP message.
> * There are no restrictions on where a SOAP intermediary can insert
> additional header blocks. That is, an intermediary can insert blocks
> between any other blocks in a SOAP header.
> Hope this makes sense!
> Henrik

Mark Nottingham

Received on Wednesday, 20 February 2002 15:22:15 UTC