ISSUE: Description of ImmediateDestination

(This should be the last of the promised series of issues against the WD text)

The current description of ImmediateDestination reads

>The identifier of the immediate destination of an outbound message.
>(NOTE: the URI supplied MAY be the identifier of a single destination
>SOAP node, or MAY be the identifier of a multicast group, which itself
>consists of zero or more destination nodes. Whether multicast is
>supported is binding-dependent. This MEP specification provides no
>standard means for representing a multicast group, except to require
>that the group as a whole be designated by a URI.)

0) The first sentence says the ImmediateDestination is the immediate destination (ImmediateSender has
the same problem)
1) The material about supporting multicast should be merged with the material at the top of section 2.2.
2) There is no need to introduce the undefined term "multicast group".  All we're saying is that a given
address may designate any number of recipients, as opposed to exactly one as one might guess.
3) As far as I know, no MEP specification provides any standard means of representing any particular kind
of address.  Perhaps the last sentence is trying to say that in some bindings, you can't
tell whether a given address designates a single receiver or more than one.  If so, this is exactly the
same as saying that any address used by that binding may designate 0 .. N receivers (if you can't tell,
you have to assume 0 .. N).
The interesting case is when the sender can tell when a binding limits receivers to exactly one (leaving
aside whether there can be more than one SOAP node attached to that receiver), either because the binding
does so for all addresses, or it provides some means of determining that a given address designates a
single receiver.

Proposal:

Insert a sentence in the first paragraph of section 2.2 reading:

Implementations MAY support sending to zero or more receivers, or MAY limit receivers to exactly
one either for all values of ImmediateDestination or for some distinguishable subset.

Change the description of ImmediateDestination to read:

Identifies, in a binding-specified way, the set of receivers for the message.

Comment:

I agree with a previous comment that the first part of the proposal is redundant in that even without it
implementations are free to be unicast.  Nonetheless, I believe it's still worth making explicit.

Received on Tuesday, 5 September 2006 16:40:36 UTC