The Emperor <the_emperor at m...> the_emperor at m...
Sun Dec 15 19:05:07 UTC 2002


> > Yes sorry I wasn't being very clear there. Obviously its not
> > desirable for the user to define their relationship with the
> > community. I am generating a list of the members FOAF files:
> Other than to "allow" a user to be a confirming source for that 

Quite ;)
> There's several angles to consider.
> A group wants to express it's members.
> A member wants to express participation in a group.
> Does the group want to "allow" the users to join?
> Do the users want to "allow" a group to state their membership?
> These *will* be extremely polarizing issues if they're not taken 
> consideration. So while the examples extended from the 'core' foaf 
examples are
> useful for 'controlled' groups they're don't quite address these 
more subtle
> concerns. Not that they can't, the example don't illustrate 
allowances for
> such. And perhaps more importantly, how to resolve ambiguities.
Yes - my concerns are with more strictly defined groups but if it is 
going to work it needs to be more flexible that that. As mentioned in 
the other post (quoted below) that the group vouching for you 
authenticity is really only important in the more defined groups 
(like claiming to be a professor of Physics at Oxford University or a 
qualified doctor) and these are the easiest to authenticate. When you 
get vaguer web designers, extreme sports fans there possibly isn't 
any need for the group to specifically authenticate your claims 
(although how one would express is this tricky - perhaps that isn't 
important and we could leave the judgement of claims up to the 
intelligence of the individual - but that is dangerous ground too).

As far as FOAFing my own community - they allow us to display a 
certain amount of information in their profiles and this is just an 
xtension of that data. As it is tied into RSS files (forum postings 
and tutorials at the moment but comments, news, book reviews, etc.) 
the connections start falling apart if too many people opt out - and 
it isn't really required.

Anyway I suppose its a matter of kicking around examples from the 
grey area between well defined and more vague communities - plenty to 
chew over.

> > All good points (and you can't beat making a point with a quote 
> > Groucho Marx ;) ). I can see that a top down approach would work 
> > discrete entities like a web site, company, university, etc. but 
> > the community gets larger and more vague like web designers,
> > Buddhists, etc. it might be more difficult to do - although it 
> > also not be necessary or desirable.
> This is sort of like the informality (or ambiguity) of 
foaf:knows. "Knows" is a
> term laden with a whole range of subtleties. Do I "know of" 
someone? Do I
> "know them"? In the biblical sense or are they just some poor 
shlubs I've met
> on the street? Qualifying such this might help get folks to see 
foaf as being
> extensible.
> For example:
> <foaf:knows type="http://wordnet_or_whatever/coworkers"/>
> <foaf:knows type="http://wordnet_or_whatever/ex-wife"/>
> Two /tremendously/ different forms of awareness.

Yes I was looking at the trust and relationship modules:



but I was concerned that they weren't flexible enough to cover 
everyone's needs. I much prefer the approach you advocate.
> > I'm sure there is a good solution to this (I'm just not sure what 
> > is). I'll have to give this more thought.
> Input and debate about this stuff will, hopefully, lead to some 
more robust
> examples of possible ways to express extensibility.
That was my intention I'm still new to this and am seeking input from 
ideas from the experts. I assume there are a large number of possible 
solutions and I am looking for the most efficient way of going about 

As well as FOAFing the Gurus Network I am in discussion with other 
people about this and I have influence over 3 or 4 communities (and I 
know people in plenty of others) and will be encouraging others to 
get FOAFing communities but I want to be sure about the best approach 
before pushing things. I mean I could lash together a schema, roll it 
out to a few thousand people and see if it takes off from there but 
its not worth doing if it isn't done well.

> > OK good :) I'll keep fiddling and see what I come up with.
> Yep, it's good to hear from folks that a) have data and b) want to 
express it in
> XML and RDF. Especially when the people and the data are not 
just "geek
> related".
Well I'm afraid we are still a bit geeky (well I am - I won't tar my 
associates with a similar brush) - I suspect early adopters will tend 
to be (no offence intended to anyone) ;)


More information about the foaf-dev mailing list