Re: Comments on CPIM draft 03

"Peterson, Jon" <[email protected]> writes:

> Um... I'm kind of getting lost here. Are you arguing for or against having a
> subscription identifier? I thought above you were motivating it as something
> distinct from the transID? But here you seem to be saying that we only need
> one identifier, which is what I thought Mr. Ramsdell had proposed... his
> last message about this, I thought, proposed three identifiers as a
> straw-man, to show that these three identifiers were really being used for
> the same purpose in different contexts, and therefore having only one
> identifier would suffice. John, am I understanding you correctly?

My suggestion about three identifiers was, as you correctly observed,
just a straw-man.  It was suggested only to get people to carefully
describe the semantics of transaction IDs, and enumerate all their
roles.  I think the straw-man performed well, don't you?  In any
event, don't take my three identifier suggestion seriously.

It is clear that one of the roles of transaction IDs is to match a
request with it's responses.  Rather than using transaction IDs for
this purpose, why not simply state in CPIM the requirement that every
CPIM compliant protocol MUST provide the means to match a request with
it's responses, and then only use transaction IDs to specify other
requirements on CPIM compliant protocols?  This might make the
document easier to understand.

John



  [reminder: [email protected] for non-technical discussions, please]