W3C

XML Schema Patterns for Databinding WG

7 Feb 2006

Agenda

See also: IRC log

Attendees

Present

Jon Calladine (BT)
Paul Downey (BT)
Anthony (Tony) Julian (HL7)
Yves Lafon (W3C)
Ajith Ranabahu (WSO2)
Regrets
Paul Biron (HL7)
Chair
pauld
Scribe
tonyj

Contents


Administrivia

minutes approved from 24th January

Sekhar Vajjhala (Sun) has left the Working Group.

Fabio Vitali of the University of Bologna has joined the Working Group.

F2F planning

pauld: can people submit issues to help planning the agenda for the F2F

status of documents

<Yves> http://www.w3.org/2002/ws/databinding/roadmap.html

<Yves> http://www.w3.org/2002/ws/databinding/edcopy/basic/basic.html

yves: oulines the roadmap, aimed at our working group, but may be of use to others

pauld: thanks Yves, I will try to update and maintain this as we go along

<Yves> the basic document is a first-cut at the organization of the final output.

pauld: two ways of working - take the input document and remove / change patterns based upon issues
... or only add patterns under control of issues

ISSUE-10 mapping of names

<scribe> ACTION: pdowney to propose text for ISSUE-10 mapping of names [recorded in http://www.w3.org/2006/02/07-databinding-minutes.html#action01]

<trackbot> Created ACTION-15 - Propose text for ISSUE-10 mapping of names [on Paul Downey - due 2006-02-14].

ISSUE-7: xsi:nil and minOccurs=0

jonc: want to advise schema authors and toolkit developers
... explicit use of nillable is key
... authors dependent upon language interpertation of empty/null values
... authors should specify nillable for nillable items

jonc: Pete Hendry cited use-case of complexType with attributes where minOccurs is difficult

toolkits should make sure that the developer can explicitly state minoccurs and nillable

various schema combinations may carry semantics which we can clarify

should have a truth table to describe behavior of minoccurs and nillable

tony will dicuss this with PaulB

<pauld> ACTION: ajulian to talk with Paul Biron on why nillable=true and minOccurs=0 is to be avoided [recorded in http://www.w3.org/2006/02/07-databinding-minutes.html#action02]

<trackbot> Sorry, couldn't find user - ajulian

ISSUE-13 and ISSUE-14

<pauld> lines up these two new (related) issues

<pauld> will revisit this after some thought

Should we use this example to make this a testable assertion?

<pauld> http://www.itu.int/ITU-T/studygroups/com17/languages/X694pdf

<pauld> ACTION: yves to investigate referencing ITU-T Rec. X.694 (01/2004) [recorded in http://www.w3.org/2006/02/07-databinding-minutes.html#action03]

<trackbot> Created ACTION-16 - Investigate referencing ITU-T Rec. X.694 (01/2004) [on Yves Lafon - due 2006-02-14].

Yves was asked to investigate the state of this document and reuse of the patterns

ISSUE-8: Using patterns to constrain numerical types

Tony will query PaulB about constraint on numeric data types

Summary of Action Items

[NEW] ACTION: ajulian to talk with Paul Biron on why nillable=true and minOccurs=0 is to be avoided [recorded in http://www.w3.org/2006/02/07-databinding-minutes.html#action02]
[NEW] ACTION: pdowney to propose text for ISSUE-10 mapping of names [recorded in http://www.w3.org/2006/02/07-databinding-minutes.html#action01]
[NEW] ACTION: yves to investigate referencing ITU-T Rec. X.694 (01/2004) [recorded in http://www.w3.org/2006/02/07-databinding-minutes.html#action03]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2006/02/22 18:45:54 $