W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > October to December 2000

Editorial comments on XML Schema Part 2: Datatypes

From: Bruce D. Sinclair <b.sinclair@liant.com>
Date: Thu, 30 Nov 2000 17:54:06 -0600
To: <www-xml-schema-comments@w3.org>
Message-Id: <00Nov30.175412cst.115201@jumpgate.liant.com>
Abstract, second sentence.  Seems to be missing an "in" or "as" in the phrase
"to be used specifications".

2.4.2.5 Enumeration.  Has an extraneous "that" in last line of the phrase "of
the datatype that from which it is derived."

3.2.1.1 Constraining facets (of string).  Since the default for the whiteSpace
facet is "preserve", a parenthetical note of the form '(default: "preserve")'
would be helpful following the whiteSpace facet listing.

3.2.2.2 Constraining facets (of boolean).  Since the whiteSpace facet is fixed
as "collapse", a parenthetical note of the form '(fixed: "collapse")' would be
helpful following the whiteSpace facet listing.  Similarly for all datatypes
other than string or derived from string.

3.3.11.3 Constraining facets (of integer).  Since the scale facet is fixed at
zero, the list of facets would be clearer if fixed facets were noted in the
list.  Thus, a parenthetical such as '(fixed: "0")' following the scale facet
would be helpful.  Similarly for all the datatypes derived (ultimately) from
integer.

3.3.14.3 Constraining facets (of long).  It would be helpful to note the
defaults for maxInclusive and minInclusive as parenthetical notes following
these facets.  Similarly for the datatypes int, short, byte, unsignedLong,
unsignedInt, unsignedShort, and unsignedByte.

3.3.24.3 Constraining facets (of timeInstant).  While it is true that a
timeInstant has the duration and period facets, these are both fixed at "P0Y".
The list of facets would be clearer if fixed facets were noted in the list.
Thus, a parenthetical such as '(fixed: "P0Y")' would be helpful following the
duration and period facets.

3.3.25 time.  Last line says 'equal to "P1D"', but the normative schema in
appendix A says "PT24H".  While probably equivalent as used in XML Schema, the
mismatch is an unnecessary editorial discrepancy.

3.3.25.3 Constraining facets (of time).  The duration and period facets are
fixed at "P0Y" and "PT24H", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

3.3.26.3 Constraining facets (of timePeriod).  The period facet is fixed at
"P0Y", so the list would be clearer with a parenthetical note following this
facet in the list.  Also, a parenthetical note that the duration facet must be
set before use would be helpful.

3.3.27 date.  Last line says "P1D", but the normative schema in appendix A says
"PT24H".  While probably equivalent as used in XML Schema, the confusion is an
unnecessary editorial discrepancy.  Also, the preceding phrase "independent of
how many hours this day has" is not meaningful if "PT24H" is the specified
duration.

3.3.27.3 Constraining facets (of date).  The duration and period facets are
fixed at "PT24H" and "P0Y", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

3.3.28.3 Constraining facets (of month).  The duration and period facets are
fixed at "P1M" and "P0Y", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

3.3.29.3 Constraining facets (of year).  The duration and period facets are
fixed at "P1Y" and "P0Y", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

3.3.30.3 Constraining facets (of century).  The duration and period facets are
fixed at "P100Y" and "P0Y", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

3.3.31 recurringDate.  Penultimate line says 'equal to "P1D"', but the normative
schema in appendix A says "P24H" (sic).  While probably equivalent as used in
XML Schema, the confusion is an unnecessary editorial discrepancy.  Also, the
normative schema fails to include the time separator, which would seem to be
required (the rules in 3.2.6.1 are somewhat inconclusive on this issue).

3.3.31.3 Constraining facets (of recurringDate).  The duration and period facets
are fixed at "PT24H" and "P1Y", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

3.3.32 recurringDay.  Penultimate line says 'equal to "P1D"', but the normative
schema in appendix A says "P24H" (sic).  While probably equivalent as used in
XML Schema, the confusion is an unnecessary editorial discrepancy.  Also, the
normative schema fails to include the time separator, which would seem to be
required (the rules in 3.2.6.1 are somewhat inconclusive on this issue).

3.3.32.3 Constraining facets (of recurringDay).  The duration and period facets
are fixed at "PT24H" and "P1M", respectively, so the list would be clearer with
parenthetical notes following these facets in the list.

Appendix A Schema for Datatype Definitions (normative).  The whiteSpace element
in the simpleType definition for "boolean" should include the attribute
'fixed="true"' to agree with the 2.4.2.6 requirement that the value cannot be
changed from "collapse" for any atomic datatype other than string.  Similarly
for "float", "double", "decimal", etc.

Appendix A Schema for Datatype Definitions (normative).  The value attribute on
the duration element in the simpleType definitions for "recurringDate" and
"recurringDay" use what appears to be an invalid form of "P24H" that should be
"PT24H".  The rules in 3.2.6.1 seem to imply that the time separator "T" is
required if any time items are present, regardless of whether all date items are
absent (but the rules could be made clearer on this issue).

Appendix C.2 Constraining Facets.  The whiteSpace facet is omitted.

Appendix C.2 Constraining Facets.  The lists in this appendix would be more
helpful if they included default or fixed values for the facet when used with
the listed datatype.  For example, on maxInclusive, '(default: "0")' for all the
unsigned datatypes, on scale, '(fixed: "0")' on all the integer datatypes.
Received on Thursday, 30 November 2000 18:56:05 GMT

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