----------------------------------------
Friend of a friend of a frie...
May 31st, 2020
----------------------------------------

It  is not  important  how it  went,  but I  recently  had a  short
discussion  on  IRC  about  users having  several  accounts  spread
throughout the  tildeverse. Why you'd  need to is  currently beyond
me. But it got me thinking; with users spread all around the place,
with phlogs,  blogs and glogs  (a little gemini-joke)  in different
places,  and connections,  both  friends  and services  intertwined
through it all, wouldn't  it be cool if we had  a way of cataloging
it all?

We could have a world readable  file detailing our accounts and the
services we use,  maybe even in a machine-readable  format, so that
we could  build cool crawlers  that may traverse this  insanity and
visualize it  in ways never known  to man, and bring  us all closer
together, like the organic mess that we are.

Sounds familiar? I thought so. My brain virtually itched, before it
hit me: FOAF! You may even have thought about it all the time while
reading this,  you sly  devil, you.  What can  I say,  we're nerds,
we've heard of things.

If you haven't heard about it,  FOAF[0], or; "Friend of a Friend, a
machine-readable ontology describing persons, their activities, and
relations  to  other  people  and objects"  (lifted  more  or  less
entirely from wikipedia[1]).

There's also an article from 2004  at The Guardian[3] about it, but
it's mostly interesting from a historical perspective.

FOAF is simply  a file that stores pieces of  information about you
in some structured  manner. Typically it's an rdf,  or even json-ld
file, apparently.

I won't go into  the specs right now, but you  may for instance say
something along  the lines  of "My  name is sndr,  and I  use these
services, where  my username is  this. These are my  emails (though
they are hashed so you may  confirm that an adress is mine, without
me receiving any  spam). These are my friends, and  their sites are
at this place,  their FOAF-file is here,  so you may look  at it at
your own discretion."

Obviously the  FOAF-files are  *way* more structured,  with Agents,
SeeAlsos, and Persons and Groups, and whatnot.

FOAF have been around for quite  some time, but it never caught on.
And if I  were a speculatin' man, I'd guess  that have something to
do with my thoughs when I  discovered it years ago: "Cool! But what
is it for?"

While I  *now* see the appeal  with something like FOAF,  where you
have a real decentralized network of  people, the truth is it's not
meant for, well, people. You don't *need* software to parse it, but
reading basically  xml for hours  on end, probably gets  quite old,
quite fast. Not to mention writing it.

I'll  admit, I  have  neither  read the  specs  entirely nor  fully
understood all  the parts  I have  read. But I  *do* wonder  if you
could build a  sort of tree of relations with  the current features
of FOAF, or if you'd need an extension of some sort.

Say if tilde.team had a file  in the webroot, that basically say "I
am not a person, but here are a set of my users, and here are their
relations-file. I  am related to  tilde.club in some way,  and here
are it's relations-file." And I could have one such relations-file,
under *my* webroot, that say "I am  a person, this is my name. I am
a user of tilde.team. I have an account on this server too. This is
my nick in  this irc-server. These are my friends,  and their files
are here."

I'll have to  actually get to grips with the  specs. And then think
some more.

I don't know, maybe I'm just describing finger. A cool finger.

--

[0]http://FOAF-project.org
[1]https://en.wikipedia.org/wiki/FOAF_(software)
[3]https://www.theguardian.com/technology/2004/feb/19/newmedia.media