OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help

relax-ng message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Subject: Minutes for RELAX NG TC 2001-09-06

Minutes for the RELAX NG TC meeting held on Thursday, Septemeber 6, 2001, 10:30 am EDT (UTC -04:00)



Not attending


1. Context-dependent datatypes and defaulting. There seem to be 3 choices

a) Disallow defaults for context-dependent datatypes
b) Allow them, but don't require processors to check that the default 
matches the attribute pattern
c) Allow them, but require processors to check only that the default 
satisfies the lexical requirements of the datatype

James: we have a choice between a, b, and c...
Makoto: what about an empty context, that is, with no mapping?
James: it is is an error to access a context...
David: I like (a); there is no business case for context-dependent datatypes.
James: it makes no sense to have a default for ENTITY; QNames don't make a whole lot of sense either. This would require us to add a new method to the interface. (b) is easier -- we simply don't check them -- and it allows flexibility in the definition.
Norm: I like (a).
Makoto: I like (a) or (c).
Mike: Before the meeting I was leaning towards (c). I like (a) or (c) now.
Josh: I like (a).
David: (a) is the right choice from a business use case.
Kohsuke: I like (b), but (a) is okay.
James: I am ok with (a). I propose that we adopt (a).

No dissent for adopting (a).

2. Any other issues with the Compatibility and Datatype Guidelines specs

Makoto: what about disallowing some defaults?
Kohsuke: having it optional or required attribute value depends on case...
James: I have a vague feeling for default on optional...there is no way to make the default separate from the attribute.
David: sometimes I use external entities that allow partners to toggle between their choice of default values.
James: I take it as a basic principle that schemas should be human readable, and that we should avoid misleading human readers.
David: does smack of being a bug...
Norm: it seems like a strange thing to do
James: I want to disallow a:defaultValue on required attributes
Kohsuke: I also want to disallow this on required attributes.
David: I am ok with disallowing this.
Makoto: fine, happy to agree with disallowing this.
James: The consensus seems to be that there be no default on a required attribute.

There was no dissent.

3. Approving the specs for publication

James: let's take a poll on who thinks these two specs are ready to publish as committee documents, after the agreed changes:
David: Yes
Makoto: Yes
Kohsuke: Yes
Norm: Absolutely
Mike: Yes
Josh: Yes
James: Who's missing? Me. Yes.

ACTION ITEM for James: Write precisely what context-dependence means.

Further discussion:

David: Perhaps we should plan some logistics around the release of these specs?
James: I prefer to wait until after the 1.0 release which is only a few weeks away.
David: Fine.
Makoto: Did we decide on the use of NCName or Name?
James: We are trying to be compatible with XML 1.0 + Namespaces

4. Date of next meeting

The next RELAX NG telcon will take place at 10:30am EDT on Thursday, October 11, 2001, the day after the RELAX NG 0.9 comment period.


James: we should start writing FAQs during this five-week comment period. Let's do it on the mail list. I'll propose some FAQs for discussion.
David: Should we have someone coordinate?
James: It's all on SourceForge. Let's just do a brain dump now, and select final answers later.

 - Send draft answers to the list
 - You can add onto another person's answer (rather than writing over their answer)
 - We can reach consensus later
 - Answer as many questions as we can

James: who was the chap who did RELAXNGCC? [Daisuke Okajima]
Someone: He is a friend of Kohsuke's...

Many thanks to James for all his hard work!


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]

Powered by eList eXpress LLC