Minutes of Transport Binding Task Force, 20 November 2001.

Attendees

1. Review of draft Transport Binding Framework document 

Quick poll on status of document:
GD - good direction; distinction between feature and MEP seems artificial
SW - preamble is good; needs more distinction between what is provided by
binding and SOAP; the 'deleted' text should go into SOAP Part 2
NM - good; concerned regarding implication that one-way is mandatory, wants
to include request-response; paragraph appearing above "Goals" section is
problematic
MJ - good; MEP/feature distinction needs clarification
HM - nothing beyond what has already been stated
YL - agree with GD; suggests stating that the MEP of a binding is a type of
feature
OH - good
MH - good; shares SW's concern about deleted paragraph; does every binding
have to support one-way?

Resolution of issues drawn from status comments:
(a) MEP/feature distinction
Agreed we should make MEPs a feature. Suggested wording changes -
explicitly state that MEPs are features, check instances of the phrase
"features and MEPs" and where appropriate change to "features".
Agreed that we should probably have a section describing MEPs (in general).
Unresolved discussion as to the relation of MEPs to faults.

(b) One-way emphasis
NM explained that his text was intended to reflect SOAP 1.1 which indicates
that SOAP messages are fundamentally one-way.  Did not reach a resolution
of the question whether a binding MUST specify at least one MEP. Later
discussion included a decision whether a binding (i) must support one-way
and possibly other MEPs (this is the sense of the text as written), and
(ii) must support at least one MEP. Several people indicated they "did not
mind" (i), no-one spoke in favour of (ii). Hence, we agreed to keep the
existing text.

(c) Deleted text
The deleted text included descriptions of properties and features. We
agreed these do not belong in the Introduction but do belong in Adjuncts,
and the deletion reflected this division rather than a decision to remove
property descriptions altogether.

(d) Distinction between binding and SOAP provided features
Discussion as to whether MEPs describe higher level abstractions or
mechanisms native to lower level transport bindings. SW concerned we had
previously made such a distinction but it was now lost in the document. SW
agreed this concern is not fatal, especially in light of the MEP/feature
wording changes agreed to earlier.

(e) The paragraph before "Goals"
The scribe did not manage to record this discussion. However, the TBTF did
agree to leave the text as is, except for (i) removing the paragraph break
between the paragraph in question and its predecessor, and (ii) inserting
an ed-note regarding a possible clarification


2. f2f topics

With regard the document.
SW volunteered to make edits to the existing document by 10p BST. NM will
send him text for the ed note regarding the paragraph before "Goals". The
TBTF then has until 5p PST to send any email comments on the doc before DF
will make it available for the f2f.

Agreed that the document represents the TBTF's proposal for section 5 (TB
Framework section). Agreed that the TBTF does not want to see this text
incorporated into a published version without changes to Part 2 of the
spec, including the HTTP binding; the rationale being that the proposed
section 5 and Adjuncts should not be published while they are still not
well aligned