W3C home > Mailing lists > Public > www-xml-schema-comments@w3.org > January to March 2004

Re: XML schema primer wording

From: David Fallside <fallside@us.ibm.com>
Date: Wed, 11 Feb 2004 14:17:37 -0800
To: Pat Hayes <phayes@ihmc.us>, www-xml-schema-comments@w3.org
Message-ID: <OFD2BD7E7F.F7FF5ECB-ON88256E37.007A5A14-88256E37.007A766F@us.ibm.com>

Thanks for your comments. I am sending them to the comments archive (per
Status section in Schema docs).

David C. Fallside, IBM
Data Management Standards
Tel 530.477.7169 (TL 544.9665)

             Pat Hayes                                                     
             02/11/2004 01:13          David Fallside/Santa                
             PM                        Teresa/IBM@IBMUS                    
                                       XML schema primer wording           


I wonder if I might make a suggestion about a possible improvement to the
wording of the XMLS Primer, if a future edition is contemplated.
Basically, it needs to spend a little more time setting the context, before
going into the purchase-order example.  The text begins
" The purpose of a schema is to define a class of XML documents"  but does
not really elaborate further.  By itself, this is a very ambiguous
statement. "Define" in what sense? What kind of information is a schema
supposed to provide about the documents in the class? (or, put another way:
given a schema, what are the documents in the class it defines supposed to
have in common?) Is it primarily syntactic; and if so, should we think of a
schema as something like an EBNF grammar? Or is it something weaker than
that? Does it impose constraints on the syntactic form of the documents?
(What kind of constraints? And WHY?) Or does it have to do with the
intended meaning of the documents in the class? (Etc.)

I think it might be helpful if the primer began by mentioning an example of
a typical intended use of a schema, if only briefly. When reading the
paragraph (section 2, beginning):

" Before going on to examine the purchase order schema, we digress briefly
to mention the association between the instance document and the purchase
order schema. As you can see by inspecting the instance document, the
purchase order schema is not mentioned. An instance is not actually
required to reference a schema, and although many will, we have chosen to
keep this first section simple, and to assume that any processor of the
instance document can obtain the purchase order schema without any
information from the instance document."

I was left puzzled, wondering why a processor would wish to access the
schema of a document; and after reading the entire primer I am still left
wondering this. In all the cases of document transfer with which I am
familiar (code, web pages, formal reasoning), the text of the document
itself all that a processor requires: that is usually how documents are
designed, after all, to enable them to be processed in some appropriate
sense.  What is a schema FOR?  What could a processor do with a document
and a schema that it could not do with the document alone? Except, of
course, check that the document conforms to the schema: but what is the
purpose of checking such conformity? If the only purpose of schemas is that
documents can be checked against them, then the entire schema-checking
effort seems to have no particular utility other than to give employment to
idle CPUs.

I am sure that the answers to my questions are so obvious to you that they
do not seem to be worth saying; but in a primer, even a few sentences to
make the 'unspoken' background assumptions a little more spoken can be of
immense value to a naive reader.

Many thanks for your attention, in any case.

Pat Hayes

IHMC       (850)434 8903 or (650)494 3973   home
40 South Alcaniz St.       (850)202 4416   office
Pensacola                 (850)202 4440   fax
FL 32501                     (850)291 0667    cell
phayes@ihmc.us       http://www.ihmc.us/users/phayes

(image/gif attachment: graycol.gif)

(image/gif attachment: pic28141.gif)

(image/gif attachment: ecblank.gif)

Received on Wednesday, 11 February 2004 17:20:12 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 23:09:01 UTC