FOAF

The Friend of a Friend (FOAF) project (wiki) is about creating a Web of machine-readable homepages describing people, the links between them and the things they create and do. (FOAF:FAQ)

Brian Kelly and Leigh Dodds wrote about the benefits of FOAF:

  1. the underlying data can be integrated with current and future application
  2. achieved without any agreement with, or notification to, other consumers of FOAF data
  3. other application developers […] can easily develop or extend applications to process it

Interesting issues:

Networking
One of the most important aspects seems to be the linking to other people you “know”. The definition of “know” is vague on purpose: “We take a broad view of ‘knows’, but do require some form of reciprocated interaction (ie. stalkers need not apply). Since social attitudes and conventions on this topic vary greatly between communities, counties and cultures, it is not appropriate for FOAF to be overly-specific here.” [1]
Security
There is a proposal for signing FOAF files using PGP and thereby hooking into the existing PublicKeyInfrastructure and the resulting WebOfTrust. [2]
IRC
There is a FOAF-bot that you can use to find information about people on the network. [3] Check #foaf on freenode.

Challenges

The main problem with FOAF, is that it’s hard for people to make FOAF pages.

This was noted three years ago, everybody more or less agrees on this, but there has been little action on that front.

One site that has recently made progress in this area is vIdentity.

(see also: 2005-11-16)

Links

See also OpenSocialNetworks.

Discussion

It’d be interesting to see an implementation of FOAF for Wiki, using the RdfForWiki framework.

Yeah!

But- what do you mean, when you say “FOAF for wiki?” Do you mean making FOAF pages in wiki? Or setting login/account data by FOAF? Or do you mean extending FOAF to describe wiki groups, or the wiki itself?

(Or all of the above?)

I recently saw the idea of a wiki “family tree.” What wiki inspired other wiki? We could encode the origins/inspirations for our wiki into the system.

So, for example, MeatballWiki points to WardsWiki, and CommunityWiki points to MeatballWiki.

It’s even simpler. WikiFolks

Lion: I was mostly thinking that it’d be nice to add the triple format described on RdfForWiki to Wiki HomePages.

Like, say, either EvanProdromou or EvanProdromouFoaf would have my RDF Foaf info. It might even be viewable as an RDF/XML file.

Whatever. The most genius step ever. To put “knows” on a wiki page. Fuck FOAF. Do it here. Completly open. Putting a foaf.rdf is too much for 99.99 % of mankind (I lost all my passwords for my stabile sites btw, so my foaf.rdf will remain the same for ever), changing a wiki homepage is 99.00 % maybe, which makes a decisive difference. The whole foaf concept on wikihomepages, open and watched carefully by the others.

I agree with FOAF for wiki, and RdfForWiki. I would use ESW:NotationThree or something for it.

We can use either RDF, or a special language.

If we use a special language, we have to write a program that interprets it. This is easy to do.

If we use the raw RDF, we don’t. This is also easy to do. RDF isn’t complicated. But it’ll cut out a lot of people who don’t know that, or aren’t comfortable with that.

Regardless, you can store your document anywhere. You can even do it on this wiki: Just post your RDF, and link to the raw version of the page, like so: http://www.emacswiki.org/cgi-bin/community/FOAF?raw=1

The only thing is, you want to put your FOAF entry somewhere that people aren’t going to be bothered by your updating the file. And you want it to be somewhere sort of permanent.

People aren’t going to want to keep updating their links to your FOAF file. Eventually, it’ll all be automated upon transfer, with event notification and all that. But until that day comes, we need to keep it in one place for a forseeable time.

We probably don’t want to store them here, because we don’t want RecentChanges to pop up whenever someone links or delinks someone or something, or adds some trifling detail to their FOAF file.

SO: Obviously, we need another wiki. I can set up “rdfstore.wiki.taoriver.net” if people like. We can just store the FOAF files there. Or, we can just use text warehousing entries. I already have http://warehouse.wiki.taoriver.net/ - if anyone wants to store their entry in a Storage Cell, or to just make a page with their name and “Foaf” at the end, it’s all good. To see it in the raw, you just put “?action=raw” at the end of the page URL.

Maybe we should rename the warehouse the “Meat Locker,” if we start stuffing bodies into it. ;D

(BandwidthTheft:) http://www.escaflowneonline.com/ge999/05-02t.jpg (Galaxy Express 999)

Or maybe we should call it Pluto. :)


Separate FOAF for separate domains

Prompted by the social network part of the vision in ConnectingWikiEssay?, I was wondering if it wouldn’t be better for each individual to have multiple FOAF files, one for each “facet” of their online (& offline?) lives. For instance, I could have a separate FOAF to “wiki technology” and for “neuroscience stuff”. Lion might have a FOAF for “wiki technology” and a separate one for “visual language”.

The advantage is that these FOAFs could serve as a machine-readable and exchangable description of whose “check boxes” you have checked in various domains (in the ConnectingWikiEssay? vision). Much like blogrolls, this would allow third-parties to find “other people who are interested in area X, who are considered interesting by people whom I think are interesting”. This is the specialization of the relation “2nd-degree friend” to a certain area of interest.

In essence, this imposes an implicit type on each link in each domain specific FOAF-file; a link in Lion’s visual language FOAF means not just “I am socially related to this person”, but rather something more specific; “I am interested in this person’s visual language work”).

This could allow automated browsers to identity “communities” of interest more easily than with the current model.

I had another SocialNetwork idea today. So, say everyone switches to FOAF like I think they should; is there any way for social network sites to make money?

First, they could do ads, like Google.

But, second, they could allow you to opt to pay to send a message to someone in your social network. The payment would prioritize the message on their “messages received” list. This is worth paying for because many times people use social networks either to get in touch with a friend-of-a-friend or friend-of-friend-of-friend as a business contact, or to simply broadcast apartment vacancies, etc to friend-of-friend-of-friends.

So, I was thinking, should you allow the receiving end to distinguish between paid and unpaid messages? If seems like they’d want to distinguish, but if they could, maybe everyone would just screen out incoming paid messages, making the service worthless. Let’s make it worth the recipients’ while.

So, the idea is, split the proceeds of messages with the message recipients. Allow each user to set their own prices, which may vary depending on their connectedness with the sender. For example, you can say, “My friends can talk to me for free. Friends-of-friends must pay a penny to bother me. Friends-of-friends-of-friends must pay a nickel”. Of course, anyone can send you an unprioritized message for free.

Then, say someone else wants to send one of those “I have a used car to sell” broadcasts to friends-of-friends-of-friends. All they do is tell the system, “I’m paying $20 to priority broadcast this as far as you can”. Then the software makes a list of that person’s friends, friends-of-friends, and friends-of-friends-of-friends, and sorts that list from cheapest to most expensive. Then it works its way up the list until the $20 has been spent.

This is similar to the Google AdWords bidding system, except in reverse and applied to email-like messages.

Details

The recipients then see the message at the top of their priority queue, without any mark to distinguish it from something sent from a close friend. But they are paid for looking at it.

Perhaps people could be allowed to charge a higher rate for actually promising to read a given message. This could be verified by letting the sender enter a “quiz question” based on the content of the message, and if you promise to read a message, then there is a 10% chance that you’ll be selected to answer the quiz question. If you don’t answer it correctly, you lose the profits from the last 20 messages that you promised to read (of course, the corporation would punish senders who entered a quiz question that can’t be answered based on the message contents).

If you want you could even allow someone with no social connection to you (i.e. a commercial entity) to send you advertisments, in exchange for a fee that you set.

This system makes everyone happy. Senders have a way to pay however much money they choose to get more people to read their used car ads, or respond to their business proposals. Receivers have a way to be compensated for the annoyance of getting those used car ads from their friends’ friends’ friends; if they don’t want to be bothered much, they just set their prices high. And let’s say the social net company takes 10% of all transactions.


See also FoafProjectWiki (rc).

CategoryWebTechnology

Define external redirect: ConnectingWikiEssay

EditNearLinks: PublicKeyInfrastructure HomePage WebOfTrust MeatballWiki WardsWiki CategoryWebTechnology

Languages: