W3C home > Mailing lists > Public > w3c-rdfcore-wg@w3.org > January 2003

Re: Review of RDF Vocab LCC

From: Brian McBride <bwm@hplb.hpl.hp.com>
Date: Wed, 15 Jan 2003 10:28:50 +0000
Message-Id: <>
To: Dave Beckett <dave.beckett@bristol.ac.uk>, RDF Core <w3c-rdfcore-wg@w3.org>


Great review.  Thanks for all the detailed comments.

As noted below most of formatting/linking/typo errors now fixed.


At 13:19 10/01/2003 +0000, Dave Beckett wrote:

>I reviewed
>   http://www.w3.org/2001/sw/RDFCore/TR/WD-rdf-schema-20030117/
>   $Revision: 1.3 $
>   Several HTML things and links to fix.
>   Some clarifications but pretty much OK to go.
>  Odd indenting near 6.1 & 6.2


>P2, P3, ... - I think the style is to link/<cite> actual document
>titles as well as [CITE].

Will take after last call.

>P3, the sentence would read better if the full concepts title wasn't
>used i.e from
>   and Resource Description Framework (RDF): Concepts and Abstract Syntax
>   , concepts and abstract syntax

after last call.

>P8 mentions "RDF's vocabulary description language, RDF Schema," once
>too many times for me.  I think by now I've rememberede it.

after last call.

>P9 & P10 start with the same phrase again; seems rather redundant,
>although it does reinforce the non-schema title.

after last call.

>P10 rdfs:domain & rdfs:range are <tt> and namespace-prefixed, before
>this has been introduced.  I suggest since these just get regular
>links, since in P8 they were linked as normal words.


>P10 one bare Document vs eg:Document previously.


>P10 any link to the more recent webarch docs here useful?

consider after last call

>P11 DAML+OIL should have a ref,

after last call

>   Owl=>OWL


>P12 "a collection of RDF resources".  What kind of collection?
>It isn't an RDF collection.

after last call.

>P12 introduces namespace and rdfs.  This was already used in P10 and

No longer.

>Add Link to rdf namespace section of xml syntax wd.


>P13 the last sentence is a bit hard to read in what is concatenated
>depending on how you read "associated".  You should reword to
>make it clear it is a URI-Reference made from
>  (the URI-Reference associated with the prefix)
>    concatenated with
>  (the suffix)

After last call

>P14 Is it worth pointing out any benefits or costs that the WG or
>editors have already considered on changing the namespace URI so that
>it is clear some of these have been discussed.

I wasn't planning to.

>Section 2 Classes
>P2 in the middle of the example it says:
>   [[ It is possible for these classes to have exactly the same
>   instances, yet to have different properties.]]
>Took me a while to work out it meant the classes might have different
>properties, not that the (same instances) could have different

After last call.

>P3 Zermelo-Fraenkel - needs a reference to somewhere

I'm tempted to just delete the sentence.  After last call.

>P4 "a class called rdfs:Class" isn't that "a class with the URI-ref
>rdfs:Class" or would named be better?

after last call

>2.1 header isn't in right format; is a <p> not a <h2> etc.


>Add a link to concepts defns of resource?

Can't find one!

>I've just noticed that only some of the rdfs:... words are links.
>Here there are none, but 2.1 has one linked and later sections vary
>too.  This should be made consistent.

That's a case where the term rdfs:Class is used in the section defining 
rdfs:Class.  I had a policy of not linking sections to themselves 
throughout.  I claim it is consistent (but maybe wrong).

>'plain literal' and 'typed literals' need linking to concepts dfns
>too especially since it says that this doc doesn't define plain


>  No mention of language.

Not needed here.

>one bare rdfs:Datatype (not <tt>).

<tt> fixed.  others would be same section references.

>   Being both an instance and
>subclass of rdfs:Class seems a special case, is this worth expanding

Nope.  That would be for the primer.

>Section 3 Properties
>Link & <cite> the concepts title

After last call.

>P3 Bare rdfs:Class (not <tt>)


>P4 Bare rdfs:range

same section reference

>section title is not <h2>etc. form


>P4 Bare rdf:Property, rdfs:Class


>P5 Bare rdfs:domain

same section reference

>P7 bare rdfs:Class


>section title is not <h2>etc form


>P1 bare rdf:Property

There are a lot like that.  fixed.

>P5 bare rdfs:Class
>P1 bare rdf:Property


>P5 bare rdfs:subPropertyOf

same section reference

>P5 bare rdfs:label

same section reference

>Section 4
>P2 starts off with the old name "RDF schema"

Changed; naughty of me; its not really a typo.

>Section 5
>Better say if this section is informative (otherwise is normative)

I consider it normative.

>sentence 2 needs a ',' after "identical".


>bare rdf:Seq, rdf:bag, rdf:Alt.  could be links to next sections too.

Decided not to.

>P3 is rather cute but it could be better put with the statement
>first: "A property of an RDF container is not necessarily a property
>of all of the members of the container".

Too cute.  Meant to take it out.  Fix in last call.

>I'm not sure about the "human reader" phrasing.  Is this like an
>rdfs:comment, only for people to use?

address in last call

>P2 you could point to the example of an parseType collection in the
>syntax wd, or to the grammar rule that defines it.

pointed to example

>the "class of RDF Lists"?  Isn't this the class of RDF Collections?

We have a terminology problem here.  The class is called rdf:List.

>If you want to just define lists, I would expect to see some
>explanation in 5.2 that an RDF List is an RDF Collection, or that the
>only currently defined RDF Collection is an RDF List.

address in last call.

>P5 the link over rdf:List goes too far and covers "rdf:List. The"


>I would avoid sublist since that begs the question of how the two
>lists relate.  How about:
>    "used to indicate a list that contains the second and onwards (if
>    any) items of the current list"

address in last call

>you could even mention the lisp terms (car cdr), but I can never
>rememeber which ones are which.


>"states that L is an instance of rdf:List that has one element which
>should be indicated by its rdf:first"
>It doesn't quite say that, it just tells you if there are 2 or more
>items in L.  How about something like:
>"states that the L is a list with at most one item.  The first item
>may be present if L has an rdf:first property."

address in last call.

>[terminology throughout 5. - item or elements of list?  I'd go with

Thought I'd fixed that.  Found one that I missed and fixed.  Consider it a 
typo :)

>I'm not sure I can say anything useful about reification but I'll
>have a go.
>Didn't we decide these were statings?  Does that need mentioning here?

Address after last call.  Didn't want to get into that in this doc; left it 
to the primer.

>links to concepts for 'triple'


>  and 'predicate of an RDF statement'

that was already done.  haven't linked them all - just the first one

>[is statement==triple ?]

No.  I had hoped that was clear.  Address in last call.  We'll get lots of 
comments on this section anyway.

>Section 5.4 Utility Properties
>P2 bare rdf:value

same section reference

>P3 HTML:  r<code>df:value</code> should be  <code>rdf:value</code>


>Section 6
>normative lists?  Are these derived from the RDFS RDF/XML doc?

Now marked as informative.


>   In
>case of error...?
>Need links to the definining sections

Yes, maybe later.

>The comment at the bottom of 6.2 needs links to the precise defn of
>rdf:_n section, earlier and to rdf:nil.
>Appendix A
>Note that it is intended to be used at the RDF schema namespace URI
>eventually?  (whatever the URI is)
>I checked the
>and it parses OK as RDF/XML
Received on Wednesday, 15 January 2003 05:27:33 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:54:03 UTC