[relax-ng] Minutes RELAX NG TC 2002-08-08

[Michael Fitzgerald <mike@wyeast.net>]

Minutes for a RELAX NG TC telecon held 8 August 2002 at 9:00 am EDT
(UTC -04:00)

The next meeting is planned for 22 August 2002.

Attending

James
Josh
Kohsuke
Makoto
Mike

Not attending

David
John
Norm

1. Conformance for RNC

James: Haven't written the conformance section yet. Promise to get to it by
next time.

2. Annotations for RNC

Makoto: About annotations to <choice> ... are there any elements that one
cannot annotate in the compact syntax?
James: There is one case: <except>...but you can put brackets around about
any following construct, star, etc., and it will work
Makoto: This is more complicated than I thought
James: Yes. We need a thorough review of A.1 in the compact spec
Makoto: Kohsuke, have you implemented the compact syntax yet?
Kohsuke: No but I have to...
Makoto: Are annotations the most difficult part of implementing the compact
syntax?
James: Yes

3. RelaxNGCC annotations >>: last remaining issue

James: the underlying problem is the annotations are largely unconstrained
in the XML syntax, except where <a:documentation> is concerned ... question
is what exactly are you annotating?
Kohsuke: in RelaxNGCC we are annotating transitions; for example, a <start>
element is a transition
Makoto: I am doing the compact syntax of Relaxer; I am concerned about
putting application specific information in a schema
James: It is as bad as using processing instructions in a schema...do we
want to add anything to the compact syntax to handle the RelaxNGCC case?
Kohsuke: Is there a way to check the version?
Makoto: Do we need a preamble?
James: We can do this with the metadata -- a preamble would be complicated
to handle
Mike: Is Kohsuke satisfied with his own solution to the problem?
Makoto: You can write your own syntax and translate it...if we don't have a
small solution, it won't be elegant
Kohsuke: I was thinking of a more Java centric syntax
James: Then it won't conform to the compact syntax...
James: Who is for maintaining the status quo, that is, not doing anything
extra for RelaxNGCC?
Makoto: Yes
Kohsuke: Yes
Mike: Yes
Josh: Yes
James: in favor

Resolved the last remaining issue by choosing to maintain the status quo and
do nothing...

4. Type assignment

Kohsuke: I spoke with Jonathan Robie yesterday about what RELAX NG plans to
do about type assignment [at Extreme in Montreal]
James: I think we can handle it with some standard annotations and a
separate spec
Kohsuke: Do we want to give names to types?
James: I am not convinced that we want to give names to types...that seems
like an XML Schema-centric approach
Makoto: I think we need a requirements document first
James: Yes...but I don't want to do something just because XML Schema is
doing it
Makoto: We can't solve all problems...

ACTION ITEMS

1. Careful review of appendix A.1 in the compact syntax spec by all
See
http://www.oasis-open.org/committees/relax-ng/compact-20020607.html#formal-s
yntax
2. James to write the conformance section
3. Deal with type assignment [on the list]

Mike: Do we have a timeline for completing the compact syntax?
James: I think we can go to committee spec at our next meeting [22 August],
if everyone does their job
Makoto: We can go to committee spec, only change it if we have to...
James: We are already in our sleep period I think...committee spec will be
our implementable draft