Re: Why is [message id] required for requests but not for other messages?

Prasad Yendluri wrote:

> That was one of the issues that was raised in
> http://www.w3.org/2002/ws/addr/lc-issues/#lc1 also.
> TC was not prepared to make it even a SHOULD requirement in the
> context of  discussion of that issue :)
> Now that 86 is closed with no action also, I am not holding up hopes
> but I think having [message id] in all messages is very useful.

I'm sympathetic, in that the current semantics provides neither the
benefit of having [message id] everywhere nor the flexibility of having
it truly optional.

On the other hand, I've always seen "very useful" as OPTIONAL, with
REQUIRED meaning "base functionality just won't work without it".

So properly echoing [message id] /if it's there/ should be REQUIRED. 
Including it in the first place should be OPTIONAL.  Or more precisely,
it should be OPTIONAL by default.  If an endpoint wants to require
[message id] on every message, it should be able to advertise this and
still conform, but the default assumption should be that a missing
[message id] is a non-event.

>
> David Hull wrote:
>
>> If [message id] is to be leveraged for uses other than correlation,
>> particularly duplicate elimination and security, wouldn't those
>> considerations apply at least equally well to non request/reply
>> interactions?  If not, what is the basis for requiring [message id] for
>> requests but not for other types of message?
>>
>>  
>>
>
>