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: [relax-ng] Minutes for RELAX NG TC Telecon 2002-06-20


Minutes for a RELAX NG teleconference held Thursday, 20 June 2002 at 10:00
am EDT (UTC -04:00)

Attending

David
James
John
Josh
Kohsuke
Makoto
Mike

Not attending

Norm, with regrets

Agenda

1. externalRef => external in the compact syntax?

Approved. Everyone was in favor of this change.

John Cowan suggested at an earlier meeting that it was inconsistent to leave
externalRef as is because parentRef had changed to parent in the compact
syntax.

2. notAllowed => ? in compact syntax?

James: We have one remaining mixed case or camelback keyword in the compact
syntax: notAllowed. Should we change that too?
John: It is so rarely used anyway
Kohsuke: I think it would be out of line to change it
James: Let's leave it as is

Resolved to close this issue with no action.

3. Should we use () like XQuery instead of the keyword empty in the compact
syntax?

James: I think this worth giving a hearing
Kohsuke: I think it is an improvement to have empty
Mike: I prefer empty
John: Why say something when you don't have anything to say?
James: I think the consensus is to leave this as is

Resolved to close this issue with no action.

4. Should we allow {} as a shorthand for { text } with attributes in the
compact syntax?

John: I always use <text/> anyway
James: { } suggests something is missing [but what?]
Mike: I thought it might be a nice analogue to the XML syntax but it's okay
as is
Kohsuke: It's not much of a savings
James: This saves typing some characters in the XML syntax

Resolved to close this issue with no action.

5. Should quotes be optional for a parameter value that is an NMTOKEN?

John: This is a looseness that a compact syntax processor should
accept...[but] this is a marginal feature

A vote was taken:

James: abstain => later changed to no
Makoto: yes, no problem with this
Kohsuke: yes, no problem
David: yes
Josh: no
Mike: no
John: no

Resolved to close this issue with no action.

6. Should we introduce an alternate string literal syntax, like Python's """
or '''?

James: My two bits...allows for richer documentation that just ##
Kohsuke: Can it contain XML fragments?
James: That would be a nightmare
John: presentation/HTML syntax convinces me
Mike: I think this would be a good thing
Josh: I like it
Makoto: I agree
James: OK we have three options
  1 triple quotes
  2 embedded [XML?]
  3 shell style, extra syntax
John: Will we have triple apostrophes as well
James: I think so
Mike: I think the Python syntax is good
David: [Python syntax] makes sense to me
John: We won't have another [dorky] hack

Approved the use of """ Python-style strings

James: Now for the XML syntax particulars
David: I think that processing [XML syntax in """] should be done externally
James: Anything goes between """ and """
John: I like what James said, "The compact syntax represents the semantics
but not the syntax of RELAX NG's XML syntax." I think you should add this to
the compact spec

ACTION: James to add this verbiage to the spec for the compact syntax

Vote [to allow XML syntax in """ strings]

James: in favor
Makoto: abstain
Kohsuke: either way
David: yes
John: yes
Mike: yes
Josh: yes

Approved allowing XML [unprocessed?] in """ strings

Disallow newlines?

John: This is good for error recovery
James: How will \x{d} behave?

ACTION: James to make proposal on how \x{d} will work in """

John: I like allowing escaped newlines
Makoto: Is this harder to implement
John: It is better if we don't have special hacks
James: Should we:
  a. disallow both [ escaped and literal newlines in """]
  b. disallow literal, allow escaped
  c. allow both

John: Given Roberts' rules of order, chair should vote last
James: [oh all right]
David: But given the time difference, aren't you always voting first? <=
quote of the week

Vote on newline options

Makoto: b but I can live with the other
Kohsuke: a but I don't want b
David: b, but I can live with a and c
John: a, can live with a, but can't live c
Mike: abstain
Josh: b, but I dislike c, don't like a
James: I'm happy to allow both

Vote on allowing newlines in singly quoted strings

Makoto: Yes
Kohsuke: No
David: Yes
John: No
Josh: No
James: Yes
Mike: Abstain

James: The status quo is to just allow both. I think we need to cogitate on
this some more before making a decision.

7. Using cc:java annotation, as Kohsuke suggests

Kohsuke: Normally assume that annotation documents what goes before
James: [annotates] the parent
Kohsuke: But this does not work out in RelaxNGCC
John: [why not make it] well documented how to add new annotations to Trang?
Makoto: Do we have enough experience [with use cases?] to do this?

We did not resolve this yet.

James: We will leave item 8 [conformance clause] for a couple of weeks

Makoto: I have corrected the errata...should I ask Norm to put it on the web
site?
Mike: I don't there is any reason to put it off

James: We will have a committee spec in two weeks



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


Powered by eList eXpress LLC