some comments on the processor profiles draft

I agree with Norm that the new version is simultaneously more precise
and more confusing :-)

Some more detailed comments...

(1) [in "Background"] I just noticed the phrase, "building on top of the
[XML Information Set] or other similar XML data models"

The XML Information Set is a glossary; I realise many people informally
use "the set of items to which terms are given in the infoset spec" to
refer to something a bit like a data model, but, it isn't one.

Suggest recasting the sentence as
[[
XML applications are often defined by building on definitions of the
output of an XML processor, for example via the XDM [XQuery 1.0 and
XPath 2.0 Data Model (XDM)].
]]

(sorry not to have made this comment earlier, if I didn't!)


(2) the sentence (again in "Background")
[[
Some stems from the growth of the XML family of specifications: if the
input document includes uses of XInclude, for instance.
]]
used to be parallel to "Some of this stems from the XML specification
itself," but the new text has moved them apart, and now the second
sentence might be better cast as,
"Some of this uncertainty..."

(3) In 2 XML processor profiles -

[[
The latter is specified by reference...
]]
the previous sentence mentions more than two things (including indirect
objects) and for at least some readers might be clearer if it said,
[[
These requirements are specified...
]]

(4) I think I see what Norm meant by confusing -- in 2.1 item 3,
"classes A, B', P and X" makes absolutely no sense.  Later in the
document these letters are given some meaning, but the definitions
appear to be circular: B' is for items provided by 2.1, but 2.1 requires
faithful provision of the information items in class B'

In the list starting with "Document Information Item: A" (I thought at
first this was defining class "A", but ended up deciding the "A" was a
copy/paste error), does "A" by [base URI] mean that all profiles must
report a document's base URI?

I suggest (numbering my changes)
(4a) change A to All, B to Basic, B' to WF, and X to R, or Recommended,
and V to Validating
(4b) move the beginning of Classes of information to the start of the
terminology section, up to but not including the note, with the header
Classes of Information
(4c) Add a new header at the start of the 2nd part of Classes of
Information, Properties by Information Class, with a sentence at the
start to say, "This section defines which classes of information must be
reported to an XML application conforming to each of the profiles
defined in this document" (followed by th existing note and text).

5 If I read the document correctly, a Recommended processor would have
to read a DTD (if supplied) in order to normalize attribute values and
to mark ID, IDREF etc., but does not need to be capable of validating.

There's no class defined here that requires DTD validation. Is that
worth saying somewhere? Maybe not.

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/

-- 
Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/
Pictures from old books: http://fromoldbooks.org/
Ankh: irc.sorcery.net irc.gnome.org www.advogato.org

Received on Monday, 7 March 2011 18:20:07 UTC