Re: bNodes as graph identifiers (ISSUE-131)

(hoping migrating of email does not cause problems)

Sandro:
>>
>>> BTW: Are graph literals in N3 tidy or not?
>>
>> I don't know what "tidy" means in this context.

Tidy literals means there is only one occurrence as I understand it.

:s :p "foo" .
:t :q "foo" .

so when is "foo" tidy (as in RDF) it has incoming links :p and :q

If it were untidy, there would be two instances of "foo" in the graph
(mathematical sense - 2 nodes in the node set).  It affects what code 
can do when walking the graph node by node.  It does not affect SPARQL. 
  We all tend to see them as the same but (as various discussions from 
WG RDF-2004 seem to show), sometimes it's not.

In N3(ish) with graph literals it is more pronounced:

:s :p { :a :p :b  }
:t :q { :a :p :b  }

is either

:s :p _:g1
_:g1 = { :a :p :b  }

:t :q _:g2
_:g2 = { :a :p :b  }

or

:s :p _:g1
_:g1 = { :a :p :b  }
:t :q _:g1

The N3 submission isn't clear to me in this area.

It is detectable.

ASK
{ :s :p ?X .
   :t :q ?X .
}

It may make a difference when quoting; I haven't had time to dig deep
enough but something around:

{
   <x> :claims _:a .
   <y> :claims _:b .
}

_:a { :WeatherToday :value :Sunny }

_:b { :WeatherToday :value :Sunny }

You can argue with modelling - but if :WeatherToday is a URL location 
whose contents changes over time, I'm maybe over simplying but if so, 
only in a way that people do already trip over.

Do <x> and <y> make the same claim or do that make separate claims
containing the same information?  Maybe later different times stamps appear.

Looks a bit like stating vs statement - unclear (to me) ATM.

	Andy

Received on Saturday, 1 June 2013 18:26:49 UTC