W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > January to March 2001

XML Schema feedback

From: Visar Nimani <visar.nimani@oracle.com>
Date: Thu, 11 Jan 2001 14:55:30 -0800
Message-ID: <3A5E39E0.1976D81D@oracle.com>
To: www-xml-schema-comments@w3.org
CC: Jim Trezzo <jim.trezzo@oracle.com>
This feedback email is based on the online form...  Sending it as email
because I was told that the web-form I submitted somehow did not
register....


Available for public use:   as commercial product / closed source /
available now
Type of Imp:                    XML Schema Generator

Description:                    Given a SQL Query the XSU can retrieve
the data from the DB. and
                                       generate an XML doc it also can
generate the XML Schema
                                       to go with the doc.

State of Impl:                   Released, but still being worked on

========================
Feature coverage/utility
========================
Part1 (Structures)

none
none
none
none
none
none
none
full         --  Schema composition: include (with matching target NS,
with unspecified target NS),
                   import, redefine.

-- Always
-- no
-- no
-- no

--------------------
Part2 (Datatypes)
none
none

Which simple types does your impl. support?    some
string, float, double, decimal, integer, long, int, short, byte, binary

Supported facets:
*  length, minLength, maxLength, precision, scale

Did not encounter any impl. difficulties.
NO

Do you support derivation of simple types by
- restriction  -- yes


========================
Priority feedback request
========================
Part1 (Structures)

xsi:null.....

xsi:null is a useful feature.  The XSU generates XMl from the DB so the
ability to explicitly specify nullness is very desirable.

impl problem?
NO

Setting schema-level defaults

Local element declarations  ....
-- Extremely useful and simplifies mapping considerably.

Order of content model and attribute declarations...
-- It is a god idea to require fixed order.
-- Attributes followed by content model.

-------------------------
Part 2 (Datatypes)

Non-URI characters and URI references...
-- A single type allowed to contain non-ASCII chars could be preferable
-- Maybe we can use type derivation to have a "pure" URI_REF


Any other comments...
-- facets should be derivable
-- support for simple type templates -- ability to create a new simple
type based on existing type but the new type might require the user to
provide certain facets  (e.g. from string create my_string where my
string requires that the length facet be set).

Any suggestions for improving this form?....
-- Add questions on equivalence classes.
-- Add questions on derivation of facets.
Received on Thursday, 11 January 2001 18:03:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Sunday, 6 December 2009 18:12:49 GMT