[bt-spec] Re: [business-transaction] Email vote - Issues 87 and 108 -voting endsTues May 30]

[William Cox <william.cox@bea.com>]

Peter -

What I meant was that

(1) while (whilst?) the conformance statement is very clear, it is unlikely to
result in the simple set of conformance categories that customers would seem to
want.  (The grouping into roles is simpler, and closer to a useful conformance
model.)

(2) a simple list of roles supported gives a 2**n effect on the number of
possible 'conformance statements.' The XA example is not to the point -- the
existing text suggests that you count the roles and state them; XA defines a
smaller number of roles.  What I'm saying is that stating conformance by listing
roles becomes less useful with more roles.

(3) By 'next stage' I did mean before proposing for an OASIS standards.

(4) I also take exception to the text  "An implementation may implement some
roles and
relationships in accordance with this specification, while providing the
(approximate) functionality of other roles in some other manner."  This is too
wooly to be in a formal conformance section. For brevity, I'll not go down this
route in this email.  How do you specify that you conform (approximately) to the
spec?!

In case it wasn't clear from my vote, this text is fine for the committee
draft.  It MUST be revisited before proceeding to the "next stage" (i.e., an
OASIS specification).

bill cox

Peter Furniss wrote:

> No problem.
>
> By "next stage", do you mean it needs sorting out before committee spec
> (i.e. before leaving Newcastle), or before proposing for OASIS standard ?
>
> Not sure what you mean by "sort of" conformance. This is meant to state that
> conformance is role-based - so something could support the Participant role
> (and Enroller) but no others, and would be able to take an effective an
> useful place in a transaction that had someone else's software at the top of
> the tree. Exactly analagous to claims of conformance to XA by a database,
> which is a different statement to a claim of conformance by a TM.
>
> Peter
>
> > -----Original Message-----
> > From: William Cox [mailto:william.cox@bea.com]
> > Sent: 30 April 2002 04:57
> > To: Peter Furniss
> > Subject: [Fwd: Re: [business-transaction] Email vote - Issues 87 and 108
> > - voting endsTues May 30]
> >
> >
> >
> > Sorry, peter, didn't get you on the message before hitting send.
> >
> > bill
> >
> > -------- Original Message --------
> > Subject: Re: [business-transaction] Email vote - Issues 87 and 108 -
> > voting endsTues May 30
> > Date: Mon, 29 Apr 2002 23:56:11 -0400
> > From: William Cox <william.cox@bea.com>
> > To: zpope@pobox.com
> > CC: "OASIS BTP (Main List)" <business-transaction@lists.oasis-open.org>
> > References: <BMEDJOFBGLHKOHLBGALLAEJLCNAA.zpope@pobox.com>
> >
> > I vote YES on issue 87, with the following note:  This proposed
> > conformance
> > statement is OK for a committee draft, but is not ready for the next
> > stage.
> > Specifying "sort of" conformance has no useful place.
> >
> > I vote YES on issue 108, with the following note:  Text should be added
> > to clarify
> > that (as in the problem statement) this isn't an issue with Atoms.
> >
> > bill cox
> >

Attachment: william.cox.vcf
Description: Card for William Cox