ISO/IEC JTC 1/SC 34N0630

ISO/IEC logo


Information Technology --
Document Description and Processing Languages

TITLE: Disposition of comments on SC34 N0592: Document Schema Definition Languages (DSDL) - Part 1: Overview
SOURCE: Mr. Martin Bryan
PROJECT: CD 19757-1: Document Schema Definition Languages (DSDL) Part 1 - Overview
PROJECT EDITOR: Mr. Martin Bryan
STATUS: Agreed disposition of comments
ACTION: Editors to create next CD
DATE: 2005-05-23
DISTRIBUTION: SC34 and Liaisons
REFER TO: N0567 - 2004-11-16 - Document Schema Definition Language (DSDL) - Part 1: Overview
N0567b - 2004-11-16 - Ballot due 2005-02-16 - Document Schema Definition Language (DSDL) - Part 1: Overview
N0592 - 2005-02-18 - Summary of Voting - Document Schema Definition Language (DSDL) - Part 1: Overview

Dr. James David Mason
(ISO/IEC JTC 1/SC 34 Chairman)
Y-12 National Security Complex
Bldg. 9113, M.S. 8208
Oak Ridge, TN 37831-8208 U.S.A.
Telephone: +1 865 574-6973
Facsimile: +1 865 574-1896
Network: [email protected]

Mr. G. Ken Holman
(ISO/IEC JTC 1/SC 34 Secretariat - Standards Council of Canada)
Crane Softwrights Ltd.
Box 266,
Telephone: +1 613 489-0999
Facsimile: +1 613 489-0995
Network: [email protected]



1. Since this part is not normative, it should not be a standard but rather a TR (type 3).

To be checked with ITTF. Should overviews of multipart standard be TRs or part of the main standard?

2. Introduce a new section for schema inclusion and recommend the use of XInclude (or external parsed entities) for syntactical inclusion of schemas.

Accepted in Principle: XInclude is already covered in Section 6 (see Figure 1). A more detailed of the role of XInclude within DSDL should be added to this existing section.

3. State that other parts may be added later when new requirements are recognized.


3.4 Schema

Replace "XML-encoded set" with "set".       


3.5 The role of document schemas

Restate that an SGML document is first converted to an information set by validation against SGML DTDs. This assumption is stated only in the scope section.


6 Validation Management

1. Delete this section and move its content to Section 4 or 8.10.

Accepted in Principle: Rename section Validation Techniques and add information on use of XInclude requested in Comment 2 above.

2. Figures 1 and 2 are misleading, since RNG validation does not create any XML documents.

Change figures to show that RELAX NG, unlike W3C schema, does not require an Infoset to be generated. Indicate clearly where Infoset is generated and reused.

8.2 Part 2

1. Since this part is RELAX NG, the first paragraph should not mention other regular-grammar-based schema languages.

Rejected. The first paragraph simply defines what is meant by "regular-grammar-based schema languages"

2. Remove the last paragraph, since new parts may be added anyway.


NB: Change the third paragraph to include the word "schema" before "language"

8.3 Part 3

1. Since this part is Schematron, the first paragraph should not mention other rule-based schema languages.

Rejected: the first paragraph simply defines what is meant by "rule-base schema languages".

2. Remove the last paragraph, since new parts may be added anyway.


8.4 Part 4

1. Drop the fourth paragraph (about NRL). NRL is not the only ancestor of NVDL and NRL is not identical to NRL. Moreover, unlike RELAX NG and Schematron, NRL is not widely used. The history of NVDL should rather be explained in Part 4.


2. Remove the editorial notes, since NVDL already provides a mechanims ("trigger") that specifies element names for creating validation candidates.

Accepted in Principle: Make editorial comments generally applicable.

>Remove the last paragraph, since NVDL schemas do specify which schema and which schema language is used for validation candidates.


8.5 Part 5

1. Replace "regular expression grammars" with "regular expressions", since regular expressions, regular grammars, and regular right-hand-part grammars are different from each other.


2. Mention the equivalence relationship. (e.g, are "1" and "01" equivalent?)


8.6 Part 6

1. Remove the third paragraph (about extracting fragments by XPath), since such extraction of fragments has never been in the scope of this part.

2. Mention better integration with Part 2 (e.g., static typechecking) as an advantage of this part.

3. This part should not have a mechanism for embedding fragments from other documents.

4. Do not mention XPath in Part 1, since different parts have good reasons to create different variations of XPath. In particular, path expressions in Part 4 are NOT XPath.

5. Since extraction of fragments by XPath should be done by an XSLT stylesheet, DSDL should not duplicate this functionality.

All rejected: Until we understand fully how our requirements match to the XPath 2.0 specification this section cannot safely be rewritten. It will be rewritten when we know better what the requirements are. The current text will be withdrawn and replaced by notes, including the above points.

8.8 Part 8

Drop entity renaming unless such renaming is implementable without changing the behaviour of conformant XML processors.

Accepted in Principle: Entity renaming will be an optional feature.


(comments absent from form; NB requested to forward comments directly to working group)


As these have not been received they cannot be disposed of.

United States

As noted in ISO/IEC CD 19757-1, Document Schema Definition Language (DSDL) - Part 1: Overview, at page 5, paths should be treated in part 10 of DSDL.