RE: RE: [Bug 4675] add text in section 5.3.3 to require that consumers and producers are required to implement at a minimum the uri scheme

Actually, Ginny, given John's last email on Issue 4675
(http://lists.w3.org/Archives/Public/public-sml/2007Dec/0059.html)
I'm now considering whether or not the whole notion of interoperability
might be broader than the topic discussed in the new 5.3.  It seems to
me that John is saying that interoperability for SML-IF interchange can
be achieved trivially, assuming a fully by-value-embedded SML-IF
document within "very tight bounds" (even when the SML model contains
non-sml.uri reference schemes)--conditions that are not covered in 5.3.
Outside of these conditions, we have no definition of interoperability,
since the URI RFC doesn't define such a notion.  On the other hand,
interoperability ultimately depends on the set of reference schemes the
consumers and producers mutually support.

(When does one's head stop spinning from this discussion?)   

Kirk Wilson, Ph.D.
Research Staff Member
CA Labs
603 823-7146
 

-----Original Message-----
From: public-sml-request@w3.org [mailto:public-sml-request@w3.org] On
Behalf Of Smith, Virginia (HP Software)
Sent: Thursday, December 06, 2007 11:12 AM
To: Wilson, Kirk D; public-sml@w3.org
Subject: RE: RE: [Bug 4675] add text in section 5.3.3 to require that
consumers and producers are required to implement at a minimum the uri
scheme


Kirk,

Yes, the current 5.3 would become 5.4.  We do overload the use of
'references' in the IF spec. There is an open bug suggesting that we
adjust some of these terms to avoid confusion (post LC).

I like the idea of changing the title to "interoperability". I'll make
that change assuming we don't get any responses disagreeing with this.

--
ginny


-----Original Message-----
From: Wilson, Kirk D [mailto:Kirk.Wilson@ca.com]
Sent: Thursday, December 06, 2007 4:33 AM
To: Smith, Virginia (HP Software); public-sml@w3.org
Subject: RE: RE: [Bug 4675] add text in section 5.3.3 to require that
consumers and producers are required to implement at a minimum the uri
scheme

Ginny, thank you for the detailed explanation.  It answered all my
questions.  I would still suggest changing the title of 5.3 to
"Interoperability" to avoid having too many sections that refer to
"references" and "interoperability" seems to be a more precise topic of
what is discussed in the section.

I assume that you intend that the current 5.3 section becomes 5.4.

Kirk Wilson, Ph.D.
Research Staff Member
CA Labs
603 823-7146


-----Original Message-----
From: Smith, Virginia (HP Software) [mailto:virginia.smith@hp.com]
Sent: Wednesday, December 05, 2007 8:24 PM
To: Wilson, Kirk D; public-sml@w3.org
Subject: RE: RE: [Bug 4675] add text in section 5.3.3 to require that
consumers and producers are required to implement at a minimum the uri
scheme

Kirk,

Yes, this means that a conformant SML-IF document MUST contain an SML
URI reference for every non-null reference. I thought it read better to
state the requirement in a normative section (as are all other
requirements) rather than only in the conformance section. And, yes that
is what the proposal intends - to remove 2 levels of conformance in
favor of only one - the one which provides "full interoperability" as
you mention. I believe that is the goal of SML-IF.

As far as proprietary schemas goes, a vendor can always claim to be
SML-complaint with such a scheme since SML allows this. The vendor would
expect that its SML-IF document, using a proprietary scheme, is not
generally interoperable so I don't think it should be 'blessed' by the
SML-IF spec. In fact, I believe that 2 levels of compliance opens up the
gate for 2 SML stacks - both conformant but not interoperable.

Good point about the producer. I'll have to think about that wording.

--
ginny

-----Original Message-----
From: public-sml-request@w3.org [mailto:public-sml-request@w3.org] On
Behalf Of Wilson, Kirk D
Sent: Wednesday, December 05, 2007 4:41 PM
To: public-sml@w3.org
Subject: RE: [Bug 4675] add text in section 5.3.3 to require that
consumers and producers are required to implement at a minimum the uri
scheme


Before I make a comment in Bugzilla, I would like to ask several
questions.

I really like the 5.1 conformance criteria section.  However, doesn't
the fact that section 5.3 states that the interchange set "MUST" contain
an SML URI scheme in each non-null SML reference and the conformance
criteria of SML-IF documents states that a conformant document MUST
adhere to stipulations in normative sections (I assume 5.3 is classified
as a "normative section") mean that a conformant SML-IF document MUST
contain an SML URI reference for every non-null SML reference?

If so, haven't we just subtly re-introduced Level 2 conformance as the
criterion of conformance?  Rather than directly stating it, we imply it
(which makes it a little more difficult to ferret out of the text).  Is
this what is intended??  If so, don't we reintroduce the debate we had
originally at the F2F for introducing the two "levels"--namely (as I
recall), vendors who wish to develop their own schemes don't want to be
declared non-conformant?

Maybe instead "levels" of conformance, perhaps defining document
conformance as in 5.1 and allowing an explicit concept of
"interoperability" will resolve the problem.  Section 5.3 becomes a
section on "Interoperability Criterion" (rather than "SML
References"--we seem to have a lot of "SML Reference sections).  The SML
URI scheme is required for each SML reference to achieve full
interoperability.  SML-IF documents can be conformant to the spec (in
terms of schema structure and requirements) but not fully interoperable
(e.g., if they contain an instance of the EPR scheme).

Also, a question on the definition of conforming SML-IF Producer: must
have a conforming SML-IF Producer be able to literally take *any
arbitrary* SML model and generate a conforming SML-IF document--even if
the SML model contains non-interoperable reference schemes???  That
seems to be calling on the SML-IF Producer to do the impossible.

Kirk Wilson, Ph.D.
Research Staff Member
CA Labs
603 823-7146


-----Original Message-----
From: public-sml-request@w3.org [mailto:public-sml-request@w3.org] On
Behalf Of bugzilla@wiggum.w3.org
Sent: Wednesday, December 05, 2007 4:27 PM
To: public-sml@w3.org
Subject: [Bug 4675] add text in section 5.3.3 to require that consumers
and producers are required to implement at a minimum the uri scheme


http://www.w3.org/Bugs/Public/show_bug.cgi?id=4675





------- Comment #18 from virginia.smith@hp.com  2007-12-05 21:27 -------
I don't think we should have 2 levels of compliance for SML-IF
documents. This does not help with regard to interoperability which is
SML-IF's primary goal.
Rather, this dilutes the interoperability that SML-IF brings to the
table.

I propose the following 2 changes:

=================
5.1 Conformance Criteria

A conforming SML-IF Document MUST adhere to this specification as
described in the normative sections.

A conforming SML-IF Producer MUST be able to generate a conforming
SML-IF Document from an SML model.

A conforming SML-IF Consumer MUST process a conforming SML-IF Document
using, in whole or part, semantics defined by this specification. It is
OPTIONAL that a conforming SML-IF Consumer process all elements defined
in this specification, but any element that is processed MUST be
processed in a manner that is consistent with the semantics defined
here.



================
Add new section - insert after section 5.2

5.3 SML References

An SML reference can contain multiple representations using different
reference schemes. SML-IF requires that all non-null SML references in
the interchange set MUST contain an SML URI scheme [SML 1.1]
representation. Any SML reference MAY also include other scheme
representations as well.

Received on Thursday, 6 December 2007 16:30:20 UTC