RFE: warn about zero predicates with anytyhing other than []s

<vkolovski> {  :a :b :c. }
<vkolovski> log:includes { :a :b :c. { :a1 :a2 :a3} }
<DanC> TimBL: indeed, that's true. the nested { :a1 :a2 :a3} is the
subject of a statement with zero properties
<DanC> q+ to suggest that only []s should be allowed as the subject of a
statement with 0 predicates.
<timbl> {  :a :b :c. } log:includes { :a :b :c.
{ :a1 :a2 :a3}  :b1 :c1 }
<DanC> TBL: this longer statement is not true.
<timbl> {  :a :b :c. } log:includes { :a :b :c.    "a1 b1 c1"  }
<vkolovski> TimBL: when you have a zero property subject, you're not
asserting it's true
<DanC> you're just saying it exists.
<timbl> (1 2 3 ( 4 5 "dftgyhujk")).
<timbl> [   mother  anne ].
<timbl> [   mother  anne ]    father fred.
<timbl> We should have a warning if  it is not a square bracket thing
and it has np data.
<DanC> "warning: existence of literal is rarely what you meant"
<timbl> RESOLTION ^

Dan Connolly, research scientist, MIT CSAIL
Decentralized Information Group http://groups.csail.mit.edu/dig/
office: tel:+1-617-395-0241
mobile: mailto:connolly+pager@w3.org

Received on Wednesday, 28 September 2005 19:10:05 UTC