Re: Editorial thread for BGP matching

Reading through, some miscellaneous 
comments/questions/suggestions(?).
------
2.1.4
"Triple Patterns are grouped together with {}(braces)."
Possibly mention here that these groupings determine scope 
of bnode identifiers(?) <<Do they, in fact? That is, 
should we read
{{_:a :p :q .}
{_:a :r :b .}}
as having two bnodes in it, or one? Im presuming two, as 
otherwise what are the {} boundaries for? >>

2.1.5 Examples of Query Syntax

The reader is tempted to ask, what about mixing the 
variable prefixes in a single query, such as

PREFIX  dc: <http://purl.org/dc/elements/1.1/> PREFIX  : 
<http://example.org/book/> SELECT  $title WHERE   { :book1 
 dc:title  ?title }

Is that legal? If so, suggest fix one of the examples to 
show this.

BTW, some of the example answer tables have the variable 
name with ? included, others not: suggest they should be 
consistent for clarity.

2.1.7
The bindings shown are to quoted strings which aren't 
typical. Might be better to show a binding to a URI(?)

2.2

"This definition of RDF Term collects together several 
basic notions from the RDF data model."
//
"......the RDF data model, but updated to refer to IRIs 
rather than URIs."
(Could refer to 
http://www.w3.org/TR/rdf-concepts/#section-Graph-URIref 
"Note: this section anticipates an RFC on 
Internationalized Resource Identifiers. Implementations 
may issue warnings concerning the use of RDF URI 
References that do not conform with [IRI draft] or its 
successors.".)

"Note that all IRIs are absolute; they may or may not 
include a fragment identifier [RFC3987, section 3.1]. Also 
note that IRIs include URIs [RFC3986] and URLs."
? Should perhaps say here that they include URI references 
(?)

Definition:
"A query variable is a member of the set V where V is 
infinite and disjoint from RDF-T."
Why infinite? We only need enough for a single query.

  "The following triple pattern has a subject variable 
(the variable book), a predicate of dc:title and an object 
variable (the variable title)."
predicate of//predicate

Triple pattern: Why not allow bnodes in property position 
as well, with the same disclaimers about not matching any 
current RDF graph? There isn't any good semantic reason to 
forbid that case either. (If this would require a WG 
decision, forget it :-)

---------

I started commenting on section 2.4 and 2.5 but its easier 
to redraft them. (The current definition of 'pattern 
solution" is wrong and can't be fixed until the scoping 
graph is mentioned, which requires re-ordering some 
material).
For my best attempt so far, see 
http://www.ihmc.us/users/phayes/Section2.4revized.html  (I 
cut out some of your html and edited it. <<comments like 
this>> )

Im rather confused, I confess, by the various distinctions 
(pattern solution, query solution,...) in section 2.4, so 
that section is still a muddle. I think some of it ls left 
over from earlier versions. For example, Pattern Solution 
refers to "RDF terms occurring in G" which is wrong. In 
fact, I don't think there is any way to distinguish 
between variable substitution and pattern solution at this 
point, since the latter can't be defined properly until 
the scoping set has been mentioned.

-------

Sorry, I only got back to this tonight, so it likely needs 
more work but I'll send this now to get it to you before 
the telecon.

Pat

Received on Tuesday, 24 January 2006 05:23:59 UTC