See also: IRC log
resolution: minutes for 6/5 call are approved.
Kumar: my AI is related to Pratul's. Not done yet.
msm: Will have AIs done before f2f
Kumar: The existing text is correct. I have mentioned this in comment# 7.
msm: the existing text looks correct, ok to close this bug.
Pratul: we need to mention in the bug that comment# 5 no longer holds.
John: Henry looked at an old draft that really had the word 'containing' and since we fixed that in the LC draft, we should resolve this bug as 'fixed'.
resolution: mark as fixed.
Pratul: proposal in comment#
14
... does this preclude identifying references using
sml:ref=true ?
msm: no
... PSVI will still have the attribute value if the attribute
is present in the instance.
resolution: fix per comment# 14, mark decided+editorial
<johnarwe> LC text: 4.1.1 SML Reference
<johnarwe> An element information item in an SML model instance document is as an SML reference if and only if it has an attribute information item for which all of the following is true:
<johnarwe> 1. Its [local name] is ref
<johnarwe> 2. Its [namespace name] is http://www.w3.org/2008/03/sml
<johnarwe> 3. Its [normalized value], after whitespace normalization using collapse following schema rules, is either "true" or "1".
<johnarwe> This mechanism enables schema-less identification of SML reference, i.e., SML references can be identified without relying on the Post Schema Validation Infoset. [XML Schema Structures]
<johnarwe> Although its normative definition allows several syntaxes to be used to identify an SML reference, for the sake of brevity and consistency, the rest of this specification uses sml:ref="true" to denote an SML reference in examples and text.
john: The proposed text would replace the 'implementation-defined' part of the current text. Thus it removes the requirement for SML validators (which is fine, if the proposal is accepted), but it also removes the requirement for other SML processors (making it harder for users of those implementations to choose processors that will interoperate in practice).
<MSM> [Speaking for myself, I would like (a) to allow non-validating consumers to use either the base infoset or the PSVI, (b) to say so explicitly and clearly, not indirectly, (c) to say so normatively.]
<MSM> [... and (d) to require non-validating consumers to document which they use, as part of their claim of conformance]
<scribe> ACTION: john to open another bug about non-validating consumers. [recorded in http://www.w3.org/2008/06/12-sml-minutes.html#action01]
<trackbot> Created ACTION-194 - Open another bug about non-validating consumers. [on John Arwe - due 2008-06-19].
resolution: fix per comment# 14, mark decided+editorial
ginny: not ready to make a decision on this bug till I see msm's test case.
<MSM> [text in comment #5 works for me]
<MSM> I wonder if 'no element-specific constraints' would work?
<MSM> or 'no (target-* or acyclic) constraints'
... long discussion about the proposed text.
resolution: remove needsReview, resolve as fixed
<scribe> ACTION: virginia to open a new bug for defining the term 'SML constraints' [recorded in http://www.w3.org/2008/06/12-sml-minutes.html#action03]
<trackbot> Created ACTION-195 - Open a new bug for defining the term 'SML constraints' [on Virginia Smith - due 2008-06-19].
ginny: will re-fix the bug to use smlfn prefix
resolution: mark editorial again
Kumar: I had sent my findings about 2557 to public-sml. John had responded to it.
John: The biggest reason I remember was 2557 allows one alias per document, sml-if needs multiple.
Pratul: Kumar, can you add your findings to the bug?
Kumar: yes
<johnarwe> http://lists.w3.org/Archives/Public/public-sml/2008Apr/0153.html
john: ... talks about findings in
his email.
... everyone should read the emails sent by John & Kumar
and then they can decide if they want to do deeper research by
reading the rfc as well. Once everyone has done the research,
we can decide on this bug.
sandy: if relative reference is an empty string, we can skip 2.a as well.
modified proposal:
Note: If the relative reference is an empty string or if it consists of only a
fragment component then steps 2.a and 2.b are skipped because the document containing the
SML reference is the target document.
<johnarwe> apologies, need to leave promptly today for another call
<johnarwe> reminder to PD and MSM, mtg in 1 hr
sandy: we cannot deduce the conclusion in the proposed non-normative note using the normative text therefore that text should be normative.
<Jim> need to leave also
msm: I think that text should remain non-normative.
<MSM> johnarwe, ack
<MSM> sandy, the other bug on which I think this one depends is 5542
Pratul: Sandy can you add a summary of today's discussion to the bug?