Primer review -- partial - thumbs up

I had wanted to have a reasonably quick read of all of it - but the sheer
length has defeated me.

This is of course a very useful document, but it would be better if it were
a lot shorter.

Minimally I think there should be some clear way to print out 10 or 15 pages
that is enough to get started with RDF.

I think however much pruning the editors do, I believe the primer will
benefit from being a multi-page document. One of the pages should be Quick
Start.

(This point above is not critical to this release).

===

Some more detailed comments:

(I have not looked at anyone else reviews yet, so there will be duplicates).

Abstract: much too long.
An abstract is a summary of the entire document, not a repeat of the
introduction.
Suggest:
1st sentence.
"RDF can be used to represent information both about things that can
be retrieved from the Web, and about things that can be identified on
the Web but not retrieved.
RDF provides a common framework for expressing information so it can be
excahnged between applications without loss of meaning."
Last two sentences.

Introduction:

...The complete specification of RDF [[suggest: and RDF Schema]] consists of
...

suggest-delete [[(from the W3C's point of view)]]

2.1 suggest move into appendix, in multi-document

... urn: (Uniform Resource Names, [[suggest-delete: intended to be]]
persistent ...

... Addressing Schemes [ADDRESS-SCHEMES] [[suggest-delete: and it is a good
idea to ]] ...

(note use of 2nd person is hard to translate consider phrasing)

[[strongly-suggest-delete: While nearly every other part ... the Web
together.]]

[[suggest-delete: Since the URI is such a ... anything, it should not be
surprising that]]

... has no physical [[suggest-replace:virtual]] existence on the ...

... how RDF statements can be [[suggest-delete: physically]] represented and
exchanged.

2.2 suggest move into appendix, in multi-document

... So additional mechanisms [[strongly-suggest-replace: must be // are ]]
provided ...

Last para 2.2, keep in main body of doc.

2.4 a way after figure 5
... renaming their node identifiers [[suggest-delete: are considered to]]
represent ...

2.5 after Figure 9
[[strongly-suggest-delete-a-large-chunk:
RDF datatype concepts borrow a ...
RDF doesn't have]]

also [[strongly-suggest-add:
Many RDF applications are expected to support XML Schema Datatypes.
]]

2.6
[[suggest-delete: from the W3C's point of view]]

Note: the name of the concepts doc has changed to RDF Concepts and Abstract
Syntax

... the abstract [[suggest-delete: graph]] syntax of RDF ...



3.1 after figure 11, para about QNames

... Because the URIref is being used as an attribute value, we
[[strongly-suggest-change: cannot// do not ]] abbreviate ... element and
attribute names [[strongly-suggest-delete: (this is due to the need to
conform to XML syntax)]]. Instead, we [[strongly-suggest-delete: must]]
write it out ...

after figure 12

... For the most part, we will continue to use [[strongly-suggest-change:
XML style (untyped) character // plain ]] literals ...

The basic serialization is particularly [[strongly-suggest-change:
recommended // appropriate ]] for applications ...

3.2 rdf:ID="10245" is an illegal id try rdf:ID="id10245"
(rdf:ID matches NC_NAME which starts with a letter or _)


thats me hosed.

hope this helps.

===

We (the WG) have a serious problem with too much material. I suspect that
Brian will be the only one of us sufficiently dedicated to read every page
this time round.

We do the community a disservice if we do not select the best material, and
either downgrade or ditch the rest.

One possible downgrade path would be to produce a book (RDF: the definitive
guide - a load of WG names) after having produced a shorter set of specs -
we might even make a buck or two.

Jeremy

Received on Thursday, 7 November 2002 14:30:56 UTC