Re: {Disarmed} Re: blank nodes (once again)

On Mar 25, 2011, at 5:48 AM, Michael Brunnbauer wrote:

> 
> re
> 
> On Thu, Mar 24, 2011 at 09:40:21PM -0500, Pat Hayes wrote:
>> Nor to me, but it does not make sense as stated. What is a "bnode for [a] person"? Bnodes by their very nature do not identify anything. 
>> Can you give some actual example RDF which you feel will be mistreated if its bnodes are skolemized?
> 
> livejournal uses bnodes with foaf:weblog as IFP. It looks like this:
> 
> <foaf:Person>
> <foaf:nick>foo</foaf:nick>
> <foaf:weblog rdf:resource="..."/>
> <foaf:knows>
>  <foaf:Person>
>   <foaf:nick>friend1</foaf:nick>
>   <foaf:weblog rdf:resource="..."/>
>  </foaf:Person>
> </foaf:knows>
> <foaf:knows>
>  <foaf:Person>
>   <foaf:nick>friend2</foaf:nick>
>   <foaf:weblog rdf:resource="..."/>
>  </foaf:Person>
> </foaf:knows>
> .
> .
> .
> </foaf:Person>
> 
> These documents are generated on the fly out of some database. If we allow
> the bnode URIs to be generated on the fly too, the FOAF documents of the
> friends of the person described here will each contain another bnode URI for
> this person. What seems even more horrible to me: The URIs will change every
> time the document is requested.

OK, so why do you feel it is more horrible to have URIs doing this than bnodes doing it? The bnodes are generated on the fly and are new each time. Why would it be worse to have these being URIs? Is it just a matter of there being more characters to generate? (I agree that is a consideration, but I think a minor one.) Or are you concerned that there would be lots of redundant 'names' for the people involved? Because if so, I don't think that one should think of these URIs in that way. They will not be resolvable, and nobody should expect to be able to find owl:sameAs assertions linking them to any other 'real' URIs. They are just syntactic 'placeholders' for what is currrently a blank node, a kind of syntactic plug for a syntactic hole. 

> 
> If we do not allow bnode URIs to be generated on the fly and request that
> once a bnode URI has been generated, it must be reused

No requests, but if its easier to re-use it, I bet it will get re-used :-)

> , it will be much
> easier for programmers of software like livejournal to switch to real URIs
> instead of bnode URIs.

Here is one possible actual utility for bnode URIs. If people use a generating technique which encodes the date/time of creation into the URI name (one obvious way to ensure uniqueness), then the URI can be used to track history and provenance of triples which use it.

> We would make using bnodes more difficult than using
> real URIs. Maybe people will stick to bnodes and nevertheless generate the
> bnode URIs on the fly ?

The way I visualize it, bnodes will be simply illegal in RDF text. So RDF-with-bnodes might be an 'internal' data structure inside some apps, but it will have to be skolemized before being published. Or it might be purely a GUI device to make RDF look prettier.

> If we make it too complicated we scare people away
> or provocate errors. Producing RDF should be easier than consuming it
> (although it should not be made too easy IMHO but that is another discussion).

Sounds like an interesting discussion, that other one :-)

Pat

> 
> Regards,
> 
> Michael Brunnbauer
> 
> -- 
> ++  Michael Brunnbauer
> ++  netEstate GmbH
> ++  Geisenhausener Straße 11a
> ++  81379 München
> ++  Tel +49 89 32 19 77 80
> ++  Fax +49 89 32 19 77 89 
> ++  E-Mail brunni@netestate.de
> ++  http://www.netestate.de/
> ++
> ++  Sitz: München, HRB Nr.142452 (Handelsregister B München)
> ++  USt-IdNr. DE221033342
> ++  Geschäftsführer: Michael Brunnbauer, Franz Brunnbauer
> ++  Prokurist: Dipl. Kfm. (Univ.) Markus Hendel
> 

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973   
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes

Received on Friday, 25 March 2011 14:14:23 UTC