[rdfweb-dev] FoaF RDF structure question(s)...

David HM Spector spector at zeitgeist.com
Sun Jun 1 15:26:57 UTC 2003


I've working on some software which will make extensive use of FoaF  
type RDF and am running into some real conceptual  
roadblocks/cognitive-dissonance points on the structure of FoaF  
objects.  Here's my current couple of quandaries:

  - Given that FoaF seems clearly to want to be a superset of the  
capabilities inherent in vCards and LDAP directories why not start with  
the vCard and InterOrgPerson schema a starting points -- FoaF could  
then cleanly bridge the worlds of LDAP and contact management systems  
and bringing them into the semantic web without forcing multiple data  
sources to exist and/or a grand reconciliation at some point down the  

- Is there any reason that the FaoF description cannot be a tree?  The  
current foaf 0.1 seems incredibly top-heavy in that all possible  
attributes of a person are assigned to the person at the outermost  
scope if you will  and don't allow for people to be very dynamic.  For  
example, I may have several jobs each of which has several projects all  
funded by different  groups/organizations.  All of the major attributes  
(address, email, AIM ID, phone #, title, web page, etc) could be  
different depending upon which role we're looking at in any given  
situation.   Being able to decompose an FoaF markup into a person who  

'sorry if these are naïve questions, but I have found any answers in  
any of the usualy online spaces that address these issues...

   David Spector
David HM Spector	 	                            spector (at  
architecture/design/planning - security consultation - due diligence
Office:(631)261-5013                                            Cell:  
Website: www.zeitgeist.com          Blog: zeitblog.zeitgeist.com:8668

More information about the foaf-dev mailing list