Re: [ws-rx] [NEW ISSUE] When to piggy-back RM headers

[Anish Karmarkar <Anish.Karmarkar@oracle.com>]

Gilbert Pilz wrote:
> I was wondering if we could use this conversation to clear up something 
> that has always bothered me. Suppose we group all SOAP messages into two 
> sets; those that have "something to do with" WS-RM and those that do 
> not. Messages in the first set include Sequence Lifecycle Messages (e.g. 
> CloseSequence), Sequence Traffic Messages (messages bearing a Sequence 
> header), and WS-RM fault messages. Messages are the later set include 
> all SOAP messages that aren't in the first set.
>  
> Of these two sets, which are candidates for being piggybacked? Both or 
> just the first set? Depending upon your implementation it may be that 
> the RMS and RMD only ever "see" messages in the first set, right?
>  

If one were to piggy-back seqack/ackreq on a message that is in the 2nd 
set, then by the above definition the act of piggy-backing would move 
the message from set 2 to set 1 -- so this should not be a problem for 
the RM implementations that see only messages in the 1st set.

-Anish
--