Log in

No account? Create an account
delirium happy

Just keep on trying till you run out of cake

Previous Entry Share Next Entry
Introduce yourselves
delirium happy
There are, according to my userinfo page, 101 people who have me on their friends list but who are not on my friends list. Some of you I know well, some of you I know tangentially, some of you I haven't a clue who you are. Some of you I probably ought to know, but have either forgotten who you are or was being dense or unobservant when you added me and never realised.

So, my invitation to you people is to introduce yourself in comments on this entry. Say hi. Tell me about yourself. Remind me who you are. Or whatever. Comment with just about anything you want, really.

(And so they don't feel left out, those who are on my friends list can play too, because I'm feeling generous.)

  • 1
I think it was karen2205 - it was the small world post about how there's some sort of relatively closed superset encompassing the various oxbridge, sf, goth, bi, geek etc. communities. I wonder what would be a good definition of such a superset would be actually - how would you get a computer given a graph of connections to detect one?

I don't know if the 100 million spiders website reached you, but if it did, one of my theories that it suddenly ground to a halt is that such a set exists. In which case there are just over a 1000 of us :-)

The only person on your friends list who I recognise who I've vaguely known in real life is tinyjo, but not sure if she'd remember me. I'm sure oxfordhacker and archie would though.

Ah, yes. I remember that one. It proved to be rather popular, and oft-linked, which is good. And to rather prove the point, from your friends list, I used to know megamole and j4 fairly well (though haven't really spoken to either in years), vaguely know imc (though don't believe I've ever met him in person) and very vaguely know lnr. Plus a few other names that I recognise, but where I've never known the person.

As for the detection of such a node, I guess that the obvious way to describe it woud be to have a group of people such that x% of the sum of all connections that these people had was to other people in the group. Though you'd then probably be able to take away and add a few people (possibly even adding people not at all associated with the group) and still stay over the limit. Maybe you'd have to try to search for any local maxima for x in some phas space that sounds likely to make my head hurt if I think about it, and then you could examine the different maxima, and apply some criteria for "core group" or "extended group" or whatever?

I don't know, really. I'm neither a mathematician nor a computer scientist, so anything I might suggest on the subject is nothing more than vaguely educated rambling.

I think I know all those four folk through slightly different ways too. (Though to be honest I can't remember exactly how or when I met j4 or lnr for the first time. They are people I think I've have found it hard not to meet somehow!)

I think one of the problems with a percentage based approach is possibly that you might get someone who is definitely part of that community, but also has a specific group of friend outside that community. Say if all the people I went to school with were on LJ, then that might dwarf the other people I know in numbers. But maybe the fact that none of my friends from school are on LJ says something?

I'm going to have to think about this. It's almost about relationships between different cliques (in the sense of groups of people who all know eachother) and what counts as cliques overlapping. But I have no idea how to define that!

  • 1