Re: Issue 455 closed: Representation header and SOAP processing model

> Oh, in my recollection the rule *2. below was discussed
> as one of the approaches and dismissed in favor of the
> sticky role. Therefore the closing email [1] seems to
> be wrong.

I think that was my main point.  I do think I would go a bit further than 
you in correcting it to be clear that, if the role is not "sticky", you 
should not in general reinsert.  What was our exact agreement on this.  If 
it's "next" MAY you reinsert, or only if it's "sticky".  I think we're in 
basic agreement on the general direction, but I think it's important to be 
precise about the details.

--------------------------------------
Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142
1-617-693-4036
--------------------------------------