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 2001-07-12

Minutes for a meeting of the RELAX NG TC on Thursday, July 12, 2001 10:30am
EDT (UTC -05:00)


Mike F.

Not Attending

Mike S.

We took the agenda items out of order because Norm's time was limited.

Informal discussion of default attribute values:

Norm confessed that though DocBook has several of them he believes that
default attribute values are evil. He said he would like to remove them, but
it would take several revisions of DocBook to do so. He said that he was
going to use some sort of annotation to represent default attribute values

6. <element name="foo"/> v <attribute name="foo"/> (issue 48)

Members of the TC were OK with the asymmetry of element/attribute empty
declarations, that is:

<attribute name="blat"/>

is correct but

<element name="blort"/>

is not allowed. It must be stated this way:

<element name="blort"><empty/></element>

TC voted to close this issue with no action.

9. Syntax sugar for optional attributes (issue 50)

Norm said that he thought that to add sugar in order to remove a little
verbosity was unnecessary. Makoto said that he was happy not to introduce
the syntax.

TC voted to close this issue with no action.

News flash:


At this point James mentioned that preexisting issues should be permitted to
branch into new issues if the issue requires action, without a coalition of
three committee members behind it. There were no dissenting voices.

Now, back to our regularly scheduled programming:

1. Syntax of "global" attribute on "attribute" element (issue 5)

As James' mail explains, because you can inherit the ns attribute value from
the <name> element, the global attribute on <attribute> becomes unnecessary.
We can just use this syntax:

<attribute><name>blah</name> pattern </attribute>

See http://lists.oasis-open.org/archives/relax-ng/200107/msg00035.html

TC voted to close this issue. [I assume that this syntax will be reflected
in the spec. -mjf]

2. Adding the pattern facet to <text/> and <mixed/> (issue 41)

The TC discussed simply referencing Martin Duerst's document on character
collections (http://www.w3.org/TR/charcol/) rather than adding this facet.
James does not like Martin Duerst's preference to compactness over XML.
Makoto was satisfied with this approach of referencing the document. See

TC voted to close this issue and will add a non-binding reference to
Martin's document.

3. <except/> pattern (issue 42)

Currently we allow <data>, <value>, <choice>. Adding <list> would be
complicated. This was accepted after some discussion.

TC voted to close this issue, leaving the syntax as described in the current
(5 July 2001) spec.

4. <key>/<keyRef> element (issue 43)

As the spec says, we currently allow <data>, <value>, and <choice>. We
discussed what children were possible, such as a <choice> of <value> or
multiple <data> elements -- even could have the same value for type
attribute, as long as they had different <param>s.

Makoto asked that this be postponed another week to which James agreed.

5. Disallow empty strings as key/keyRef (issue 53)

Makoto agreed that empty strings are not very useful as keys. TC felt that
we need a bit more thought on this issue and deferred.

7. Overriding attributes (issue 52)

James stated that not having this is no worse than not having lax
validation. He said we can think about this after v1.0.

Issue closed for now.

8. nsName using prefix (issue 49)

See http://lists.oasis-open.org/archives/relax-ng/200107/msg00017.html.

Issue closed with no action.

10. Implicit grouping in <start> (issue 51)

See http://lists.oasis-open.org/archives/relax-ng/200107/msg00018.html.

The TC felt it was a good idea allow only one pattern as a child of <start>.

Issue closed.

11. Drop name attribute from <start> (issue 44)

Makoto felt that we should drop this as it is slightly easier. Kohsuke said
that he would like to get rid of it. James was neutral but was happy to go

See http://lists.oasis-open.org/archives/relax-ng/200106/msg00332.html.

TC closed the issue without action.

12. Fragment identifiers in href attribute (issue 46)

Makoto admitted he was not a fan of FIs, but since they are only used in
HTML, he was OK. James felt that the URI -- which allows FIs -- should be
considered a black box and that we should allow relevant RFCs to specify
their use. James stated that he thinks FIs are "deeply wrong" and Makoto
believes they are "bogus."

Nonetheless, TC decided to permit them, deferring to pertinent RFCs.

See http://lists.oasis-open.org/archives/relax-ng/200107/msg00000.html.

13. References to named patterns in external grammars (issue 54)

We deferred discussion on this as it needs more thought and discussion.

We discussed the spec for a time. Congratulations to James for all his hard
work there. Mike F. suggested expanding the definitions (something like XML
1.0 Second Edition did). This can be done without insulting the experts nor
leaving neophytes behind. Makoto said he did no like some of the line
breaks. James has been having a time with them. He is using nested tables
and offered to provide the source so that Makoto could play around with
them. Makoto suggested a line between two conditions and [someone else]
suggested an explicit AND.

Wy'east Communications     http://www.wyeast.net     mailto:mike@wyeast.net

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

Powered by eList eXpress LLC