See also: IRC log
<trackbot> Date: 20 October 2009
<scribe> scribe: alewis
note: quorum not filled.
minutes not approved.
agenda review.
call for comments: none.
eric: issue filed on soap prefix issue in 3.4.5.
resolved: action 116
issue 117: changes to spec, action completed by phil.
resolved: action 117
<padams> issue 15
eric: discussion of issue 15?
phil: makes sense to make changes
<eric> http://lists.w3.org/Archives/Public/public-soap-jms/2009Oct/0012.html
eric: discussion of a number of
items related to wsdl.
... item 1: should we reference the uri scheme and set a link
the first time, and after that just do bibrefs?
phil: we should have a link the first time, then not?
amy: shouldn't we link first time then bibref?
eric: owl spec links first time, then subsequent uses are undecorated.
phil: what's the difference between the two?
eric: first time is link to bibliography, then undecorated. what we're doing wrong now is direct link, in middle of spec.
phil: should be changed, what's
right?
... first time of mention, what would we see?
amy: first link should be to bibliography; biblio should link to spec.
eric: don't do that now. just appears.
phil: first reference in binding
spec to uri spec links to biblio (section 1.1).
... third point links to uri spec.
eric: no, that points to the rfc
for uri.
... actually, we're talking about not the link to the soap/jms
uri spec, but to the uri spec.
... we eventually link to the uri rfc, in section 3.4.5.
... so we need to make a bibref much earlier, and then review
the spec to insure that we use the pattern: first mention
bibref, later mention undecorated.
phil: was confused, thinking this
was jms uri, but this is actually in reference to uri
spec.
... we need the first mention to point to rfc 3987, and later
mentions to be unadorned uri text.
<eric> (Amy was also confused.)
amy: yes, i was confused as
well.
... but eric wasn't confused. on the other hand, he wrote the
email.
eric: this raises a larger issue: we need to review the spec to insure that all references work this way.
<scribe> ACTION: review the spec for references, and propose resolution [recorded in http://www.w3.org/2009/10/20-soap-jms-minutes.html#action01]
<trackbot> Sorry, couldn't find user - review
<scribe> ACTION: eric review the spec for references, and propose resolution [recorded in http://www.w3.org/2009/10/20-soap-jms-minutes.html#action02]
<trackbot> Created ACTION-118 - Review the spec for references, and propose resolution [on Eric Johnson - due 2009-10-27].
eric: looking at assertions
around wsdl, they seem thin.
... normative statement 3001 doesn't seem to need to be
normative, given resolution of issue 15.
... statement 3002 says: most specific must take
precedence.
... essentially identical to 3001, which says propagates.
... in any event, there's a lot of overlap, and we need to
clarify.
... raise issue?
phil: yes
eric: allow multiple sibling instances of elements? no prohibition at present. need normative prohibition.
amy: agree
eric: 3003 also appears redundant with 3002. says uri trumps wsdl.
amy: recommend that single issue be raised for overlap of 3001-3002-3003.
eric: 3004 does not contain rfc 2119 keywords.
phil: sentence is incomplete.
eric: non-normative?
phil: oh, just a caption.
eric: weird to have normative
caption.
... says: soap action property is only exception that may
appear multiple places.
... doesn't soap have a means of defining soap action?
... we need to delegate to soap.
amy: but soap says only valid for http
eric: yes, but we specifically reject that in 3.4.3.
phil: soap action defined in wsdl would be defined in binding.
eric: normative statement (if
any) is that this particular property is not allowed to appear
in multiple places, which implies that this is a soap-jms
property, which is wrong.
... better to say that things may appear in blah-blah-blah
places.
phil: if we get rid of the last property, we don't need a table, because all the properties have the same restrictions.
eric: so we switch the wording,
drop the table. eric will open the issue.
... wsdl testing, after we make the suggested changes,
indicates that it would be straightforward to write test cases
that test the precedence rules etc.
... do we actually have normative statements about wsdl?
phil: in non-wsdl sections, we
were specific about consequences of failing assertion
... in wsdl, not clear that a fault should be raised.
... seems as though we say: "this is how to do it if you want
to; if you don't want to, we don't care."
... there are no faults to raise, really.
eric: the conformance target is
different.
... in other sections, you raise an error, because it's a
protocol error.
... in this case, we would accept or reject the document.
phil: "reject" mean raise an error? or just ignore?
eric: if generated incorrectly, that's an error, but can't mandate "no bugs". we can only specify that a malformed document must be rejected (or ignored?).
phil: would we require a conforming implementation to raise an error for a misplaced element? or would it just not be looking there?
amy: rants about wsdl, but can't remember and scribe at the same time.
eric: are you saying "it is defined for these locations and is undefined elsewhere?"
amy: yes, approximately.
eric: our normative statements are about precedence.
amy: and about where it has meaning if it appears?
eric: yes, but we don't say you can't put it elsewhere.
amy: yes.
... we don't define what it means if it appears elsewhere.
eric: our only concerns are
precedence and multiple appearance in single scope.
... any other ambiguities?
<scribe> ACTION: eric to try to determine what is normative and how to generate test cases. [recorded in http://www.w3.org/2009/10/20-soap-jms-minutes.html#action03]
<trackbot> Created ACTION-119 - Try to determine what is normative and how to generate test cases. [on Eric Johnson - due 2009-10-27].
amy: argh. that action should include "for wsdl".
eric: comments on FAQ?
... on testing?
... implementation updates?
no comments on the above.
other business?
eric: other obligations raise the
need for an interim chair, or otherwise handle his possible
absence.
... change in dst in the zone where zakim lives; note
difference between us and european changes.
trackbot, end telcon
This is scribe.perl Revision: 1.135 of Date: 2009/03/02 03:52:20 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: RRSAgent_Text_Format (score 1.00) Succeeded: s/115/116/ Succeeded: s/issue/action/ Found Scribe: alewis Inferring ScribeNick: alewis WARNING: No "Topic:" lines found. Default Present: +1.512.286.aaaa, +1.919.663.aabb, +1.650.213.aacc, alewis, eric, padams Present: +1.512.286.aaaa +1.919.663.aabb +1.650.213.aacc alewis eric padams WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Found Date: 20 Oct 2009 Guessing minutes URL: http://www.w3.org/2009/10/20-soap-jms-minutes.html People with action items: eric review WARNING: No "Topic: ..." lines found! Resulting HTML may have an empty (invalid) <ol>...</ol>. Explanation: "Topic: ..." lines are used to indicate the start of new discussion topics or agenda items, such as: <dbooth> Topic: Review of Amy's report[End of scribe.perl diagnostic output]