1 view
Skip to first unread message

Daniel Sheppard

unread,
Aug 19, 2007, 9:58:38 PM8/19/07
to social-networ...@googlegroups.com
In regards to this assumption
http://bradfitz.com/social-graph-problem/#assumptions-dontsync:

It's recognized that users don't always want to auto-sync their
social networks. People use different sites in different ways, and a
"friend" on one site has a very different meaning of a "friend" on
another. The goal is to just provide sites and users the raw data, and
they can use it to implement whatever policies they want.

I believe that's what people are doing out of necessity, not what they
want. People segregate their social networks between social networking
sites because those sites only provide a flat view of social
relationships.

For example, the friend list of a public figure on myspace would overall
be regarded as three main groups:
- 'subscribers to my blogs/bulletins'
- 'people whose blogs/bulletins I subscribe to'
- friends

That friends group would be be then same list of people that they would
be friends with on Facebook, but that group will probably consist of
other groups, such as:
- work colleagues
- school friends
- various other social groups (theatre, sports and other
activity related groups)
- close friends

Some groups will exist solely on facebook, some solely on myspace - not
just because of the differences in the sites, but the differences in how
the user wants to communicate differently to different groups of people.

As a user moves between social applications, it will not be the
individual friends that you'll want to associate with in the new
setting, but groups of friends. When notifying a social application of
your friends list, it should be possible to allow that application
access only to specifically tagged friends.

Daniel.

Julian Bond

unread,
Aug 20, 2007, 3:22:15 AM8/20/07
to social-networ...@googlegroups.com
Daniel Sheppard <dan...@pronto.com.au> Mon, 20 Aug 2007 11:58:38

>In regards to this assumption
>http://bradfitz.com/social-graph-problem/#assumptions-dontsync:
>
> It's recognized that users don't always want to auto-sync their
>social networks. People use different sites in different ways, and a
>"friend" on one site has a very different meaning of a "friend" on
>another. The goal is to just provide sites and users the raw data, and
>they can use it to implement whatever policies they want.
>
>I believe that's what people are doing out of necessity, not what they
>want. People segregate their social networks between social networking
>sites because those sites only provide a flat view of social
>relationships.
>
>For example, the friend list of a public figure on myspace would overall
>be regarded as three main groups:
> - 'subscribers to my blogs/bulletins'
> - 'people whose blogs/bulletins I subscribe to'
> - friends
>
>That friends group would be be then same list of people that they would
>be friends with on Facebook, but that group will probably consist of
>other groups, such as:
> - work colleagues
> - school friends
> - various other social groups (theatre, sports and other
>activity related groups)
> - close friends

I feel like a lone dissenting voice. But I think this problem is going
away as people understand what it means to live in public. I also think
it's related to age and insecurity or the lack of it.

My reaction to all this is that "you have no privacy, get over it". And
"If you don't want that information to be known about you, don't post it
on the net." But I also think that as a society we're going to have to
come to terms with people's histories. We all know that just about
everyone does things in their late teens and early 20s that shouldn't be
held against them when they're 35. So how does having a public record of
your teenage foolishness change anything?

Of course, that's unrealistic. In the short term, if this project
results in people thinking their privacy has been exposed without their
agreement it will cause problems. Last year's noise about Facebook
Minifeed RSS is a classic example.

--
Julian Bond E&MSN: julian_bond at voidstar.com M: +44 (0)77 5907 2173
Webmaster: http://www.ecademy.com/ T: +44 (0)192 0412 433
Personal WebLog: http://www.voidstar.com/ skype:julian.bond?chat
*** Just Say No To DRM ***

sky

unread,
Aug 20, 2007, 9:57:12 AM8/20/07
to Social Network Portability
The problem does not go away, because if this is successful, the
social network is not just a bidirectional
graph, but a communication medium. It's not just 'who are my friends'
but 'what am I telling my friends.'
Sure, what people say when they're 18 won't be judged harshly, but
what about what I say to my legal team? Or co-workers? Inherently,
privacy will need to be part of the model precisely because some
communication
with/via the medium will need to be private.

And...don't forget basic stuff like avoiding the spammers :-)

/sky

On Aug 20, 3:22 am, Julian Bond <julian_b...@voidstar.com> wrote:
> Daniel Sheppard <dani...@pronto.com.au> Mon, 20 Aug 2007 11:58:38

Ben Bangert

unread,
Aug 20, 2007, 6:41:45 PM8/20/07
to Social Network Portability
On Aug 19, 6:58 pm, "Daniel Sheppard" <dani...@pronto.com.au> wrote:

> As a user moves between social applications, it will not be the
> individual friends that you'll want to associate with in the new
> setting, but groups of friends. When notifying a social application of
> your friends list, it should be possible to allow that application
> access only to specifically tagged friends.

I agree completely. I suggested in a separate thread about having
separate FOAF's that are used by different SN's as a possible way to
re-use an existing standard in a way that will let the SN build an
applicable graph for its particular social context.

This does cause an issue of how the user is to expose specific FOAF's
to specific SN's in a way that maintains the privacy of each list
though.

- Ben

Zach Scott

unread,
Aug 21, 2007, 3:13:06 AM8/21/07
to Social Network Portability
On Aug 19, 7:58 pm, "Daniel Sheppard" <dani...@pronto.com.au> wrote:
> Some groups will exist solely on facebook, some solely on myspace - not
> just because of the differences in the sites, but the differences in how
> the user wants to communicate differently to different groups of people.
>
> As a user moves between social applications, it will not be the
> individual friends that you'll want to associate with in the new
> setting, but groups of friends. When notifying a social application of
> your friends list, it should be possible to allow that application
> access only to specifically tagged friends.

I haven't seen it mentioned yet, but I think personas are also a very
important part of partitioning friends. In the real world (some)
people
dress differently when they go to work than they do when they're out
having fun with friends. Likewise, I think it is important for a user
to be able to have different personas for different social contexts
which
may or may not be encompassed by your concept of groups. There
could easily be multiple social contexts within a single SN.

-zs

Julian Bond

unread,
Aug 21, 2007, 3:28:44 AM8/21/07
to social-networ...@googlegroups.com
Zach Scott <zach....@gmail.com> Tue, 21 Aug 2007 07:13:06

>I haven't seen it mentioned yet, but I think personas are also a very
>important part of partitioning friends. In the real world (some)
>people
>dress differently when they go to work than they do when they're out
>having fun with friends. Likewise, I think it is important for a user
>to be able to have different personas for different social contexts
>which
>may or may not be encompassed by your concept of groups. There
>could easily be multiple social contexts within a single SN.

I know lots of people see this need. But here's the dissenting voice.
I'm uninterested in categorising my friends into different groups or
running different personas. My work life is entangled with my private
life to such an extent that I can't really see the difference. If I have
two personas at all it's published on the net and not published on the
net. All I have is a single address book.

Maybe this just means that I have no life and need to get one! ;)

Mat Bowles

unread,
Aug 21, 2007, 6:27:45 AM8/21/07
to social-networ...@googlegroups.com
On 21/08/07, Julian Bond <julia...@voidstar.com> wrote:

> I know lots of people see this need. But here's the dissenting voice.
> I'm uninterested in categorising my friends into different groups or
> running different personas. My work life is entangled with my private
> life to such an extent that I can't really see the difference. If I have
> two personas at all it's published on the net and not published on the
> net. All I have is a single address book.

I'm sort of similar; I'm me, under my name, everywhere if I can be.
But the data I give out and allow people access to depends on who they
are and how I know them.

Professional contacts and clients don't have access to the stuff about
my personal life that I put on my journal for example; I know I can
lock them out using privacy controls &c. I have a friend working on a
Wordpress plugin that'll emulate the functionality LiveJournal gives
in that area, so that I can self-host everything, and that's looking
very promising. The appeal to me of this sort of project is that I'll
be able to define who can read my content after they've logged in at
my site before they've ever been there, currently I can only assign
rights to existing users not to notional OpenID logins.

> Maybe this just means that I have no life and need to get one! ;)

This is possibly true, I definitely don't regret putting my projects
on hiatus for the last 6 months-the fiancee I've gained is a most
excellent beta-tester...

--
Mat Bowles
Tel: +44(0)7855 307993
Email: matb...@gmail.com
Web journal: http://mat-bowles.taktix.org/

Zach Scott

unread,
Aug 21, 2007, 10:32:47 AM8/21/07
to Social Network Portability
On Aug 21, 1:28 am, Julian Bond <julian_b...@voidstar.com> wrote:
> I know lots of people see this need. But here's the dissenting voice.
> I'm uninterested in categorising my friends into different groups or
> running different personas. My work life is entangled with my private
> life to such an extent that I can't really see the difference. If I have
> two personas at all it's published on the net and not published on the
> net. All I have is a single address book.

I view the net a bit differently I guess. Just because I'm
interacting on
the net does not mean I classify that interaction as published. At
least not in the same sense you appear to. Many do and will continue
to have private discussions on the net.

The other aspect is fitting into a context and has nothing to do
with privacy. I would not show up to a wedding dressed in ripped
jeans, nor would I go to work in an office in rock-climbing gear.
It's
not socially acceptable. Likewise the social acceptability of any
single avatar or profile is not universal.

-zs

Owen Taylor

unread,
Aug 21, 2007, 10:57:46 AM8/21/07
to Social Network Portability
On Aug 21, 6:27 am, "Mat Bowles" <matbow...@gmail.com> wrote:

> On 21/08/07, Julian Bond <julian_b...@voidstar.com> wrote:
>
> > I know lots of people see this need. But here's the dissenting voice.
> > I'm uninterested in categorising my friends into different groups or
> > running different personas. My work life is entangled with my private
> > life to such an extent that I can't really see the difference. If I have
> > two personas at all it's published on the net and not published on the
> > net. All I have is a single address book.
>
> I'm sort of similar; I'm me, under my name, everywhere if I can be.
> But the data I give out and allow people access to depends on who they
> are and how I know them.

Some people may want mandatory privacy, but what a lot of people are
looking for is simply to avoid confusing mixing. You buddies from
college aren't interested in your professional contacts. Your
professional contacts aren't interested in your pictures of your dog.
And so forth. From this perspective, a persona isn't a covert alias,
it's just a "filter" onto a user's personal data. Right now, people
often try to use accounts on different sites for this
purpose .."LinkedIn is my business contacts", "Facebook is my college
buddies", but that tends to break down (your coworkers find you on
Facebook and add you there). Having good social network import will
definitely increase that problem.

The distinct challenge here is making something meaningful to the
user. When a user imports their network into a new site, you can't ask
them to write a complex select statement to specify which users in the
network to import. I'd go as far as to argue that if you are
supporting splitting contacts into multiple groups/personas/tags, then
that's only really sensible to do on a single site - the user needs a
"home site" for their social network. Merging networks together only
manageable when grouping contacts by the network they came from.

- Owen Taylor <ota...@redhat.com>

Janet Hawtin

unread,
Aug 21, 2007, 11:02:10 AM8/21/07
to social-networ...@googlegroups.com
On 8/22/07, Owen Taylor <owta...@gmail.com> wrote:

> The distinct challenge here is making something meaningful to the
> user. When a user imports their network into a new site, you can't ask
> them to write a complex select statement to specify which users in the
> network to import. I'd go as far as to argue that if you are
> supporting splitting contacts into multiple groups/personas/tags, then
> that's only really sensible to do on a single site - the user needs a
> "home site" for their social network. Merging networks together only
> manageable when grouping contacts by the network they came from.

or grouping by theme/tag regardless of originating site.
i think being able to carry a theme or tag family across a range of
contexts would be a nice way to facet traffic/connections for
different purposes

j

kevin curry

unread,
Aug 21, 2007, 11:34:06 AM8/21/07
to social-networ...@googlegroups.com
Right now, people
often try to use accounts on different sites for this
purpose .."LinkedIn is my business contacts", "Facebook is my college
buddies", but that tends to break down (your coworkers find you on
Facebook and add you there). Having good social network import will
definitely increase that problem.

The distinct challenge here is making something meaningful to the
user. When a user imports their network into a new site, you can't ask
them to write a complex select statement to specify which users in the
network to import.

There's some indication that people may be able to figure it out on their own.  I did this little viz of my del.icio.us network using a completely generic layout algorithm ( force-directed graph) and I was able to clearly see communities of interest.  The reason this happened is because I'm including the edges of the "back graph" - links among my friends who also know each other.  But, people in the two subsets do not know each other well.  So there is nothing holding those two groups together and I am able to see that and I know exactly what it means.  Now, 3rd parties don't know what that means, but I consider that to be good for me.  3rd parties will find other kinds value in here.  (Hopefully...that's Owen Taylor's concern...will they?)  I can also think of circumstance of when this will break down, but I was encouraged by this unexpected outcome.
Reply all
Reply to author
Forward
0 new messages