Whatever happened to FOAF?

FOAF stands for Friend of a Friend. This open data standard is part of the Semantic Web which was something Tim Berners-Lee (inventor of the web) thought could add value to the web by distributing useful data across web sites.

This is a fairly technical post, but it may have an audience here.

FOAF
Image from Wikimedia

The idea of FOAF was that you provided data about your friends and contacts in a data file on your web site. Software or 'agents' could extract that data to gather information about who knows each other. So it would be like your Facebook friends list, but you would control the data and there was no need for a central database. The data would be stored in one of a variety of data formats. This may have been one of the barriers to mass adoption as you really needed special tools to create and maintain such a file. There were also ways to embed the data within a web page. A related project was XFN that let you specify your relationship to someone when you linked to their site.

Some social sites did create FOAF files for their users. There was a site whose name I've forgotten that made it simple to traverse the data from the FOAF files available to see the connections, but I think I remember they shut it down as some people found this a bit creepy, even though the data was still out there. There are still tools out there that can render the data in a readable format.

So why did it not take off? Cory Doctorow (creator of Boing Boing) wrote an essay setting out some reasons why open metadata would not work. One is that people will lie and another is laziness. I think it's also partly due to the fact that most people will not run their own web site these days. In the early web you had to do that to get on there, but now everyone just joins various social sites that jealously guard their data. Also, many people want such data to be private. There are still plenty of services where the data is public, so you have to be careful about what you want to expose about yourself. I do have my own site, but I've not been maintaining it lately. It has a FOAF file, but that's out of date, so it seems Cory was right.

A few years back I was very enthusiastic about the possibilities of the semantic web, but you don't hear much about it these days. Of course people can, and do, mine the Steem blockchain data, but it doesn't include much detail beyond who follows who. You can't imply too much from that. I don't think the developers plan to allow us to specify how we know people. You can't even provide much personal data beyond a display name, location and web link. A lot of people choose to remain fairly anonymous.

So I see this as a lost opportunity. It might have been useful in certain fields. For example, scientists write papers that list the authors. If this data is in an open format then you can extract it to see the connections. If each site uses its own format (or just had text on a web page) then it's much harder to use the data. The open source world could also use it to see who has worked together. You can do this with sites like Github, but I'm not sure if the data can be easily extracted.

Anyway, I hope this has been of interest to some of you and I'm interested to hear your thoughts on whether there is a place for open data standards on the web. There may well be active projects out there that I don't know about.

Steem on!