Automating the group creation process is something I've been thinking about lately. In the GNE prototype, we have the ability for people to make explicit 1-to-1 relationships. The other players with whom one has made a relationship are collected in the contact list. (The 2,244 players in the game's first round of prototype testing created 40,784 relationships.)

To “make someone's acquaintance”, they need to agree to become acquainted, but once you are acquainted you can tag them as a friend unilaterally (we also now have enemies, and there will be a slightly richer set of 1-to-1 relationships in the big game, but acquaintances and friends are enough for most purposes).

Here's a diagram of a small chunk of relationship data where acquaintances are indicated by thin lines and friendships by thick lines (diagram is slightly idealized; all the friendship relations were made symmetric):



If we define a group as a maximal set of actors where every actor is a friend of every other actor you could write an algorithm to detect all the groups (that would be, um, approximately NP-badass, I think). For example, here are the four groups which existed in the first diagram:


Note that it is quite messy and there is lots of overlapping. That's because this stuff is very messy and there is lots of overlapping. There are four distinct groups, one member of which is common to all:



But three of the four groups have at least two members in common with at least one other group — the whole thing is very close to being one group. And you could alter the algorithm so that configurations like this count as one group; but purely social groups (in “real life”) tend to be very subtle and no matter how the automatic detection worked, it would only be a suggestion: very often groups of friends operate on a level that is influenced but not entirely determined by the individual 1-to-1 relationships of their members (a group of three people may have reasons to exclude some fourth actor that none of the three group members has as an individual).

In fact, most groups are formed based on criteria other than the 1-to-1 relationships of the members: e.g., roommates, co-workers, sports teams or reading groups on the small scale, or religions, political parties, professional associations, etc., on the large scale. The groups are explicitly formed or founded and individuals are either explicitly invited or can join through some explicit action on their part, using whatever process the group has set up.

If individuals can make their interests, preferences, desires, goals, etc., explicit in a way that is consistent and machine-usable, then we can make machines that come up with things like “you and these 13 (or 37 or 1142) other people look like a group to me” (though given the messy and overlapping bits the suggestions maybe a lot more muddled that than). Even then there are all kinds of idiosyncratic, non-formalizable reasons that a particular individual may not wish to be a member of the suggested group. (Machine-generated suggestions like “hey, this person who you don't know shares a lot of interests and opinions with you — here's a way to get in touch with them” will probably be more useful.)

Still thinking about it, but I'm not too optimistic about automating the formation of groups.
` ` ` ` ` ` ` ` ` ` ` `

I was reading my mail when this spam message came in, at just the same time as my phone made the “I just got a text message” noise. The text message was also a copy of this spam:

Subject: A LOGIC ANSWER TO A LOGIC QUESTION

World Events are affecting the way we live.

It's time to fight back!

Indeed! O sublime banality, how can I not love you!? Fight back against world events! (The spam went on to ruin it by trying to sell me dubious heating oil futures.)
` ` ` ` ` ` ` ` ` ` ` `

I got my review copy of The Elements of User Experience yesterday and it is been my temporary downtime reading since then. Very well put together — congrats Jesse!
` ` ` ` ` ` ` ` ` ` ` `

www.sony.com had a redesign at some point in the not too distant past and interestingly now offers the following eight top level choices:

My Sony | Shop | Electronics | Playstation | Online Games | Movies, TV & DVD | Music | More of Sony

First, two out of the eight are gaming. Second, for a company with $55B in annual revenue that is one of the biggest producers of consumer electronics and largest publishers of movies, TV and music, it's remarkable that online games makes it as one of their fundamental categories, especially given that they only have one massively multiplayer title out right now. I keep saying, this stuff is gonna be huge(r).
` ` ` ` ` ` ` ` ` ` ` `

You Arent Gonna Need It (“Always implement things when you actually need them, never when you just foresee that you need them”) is perhaps my favorite extreme programming principle.
` ` ` ` ` ` ` ` ` ` ` `

And so the social software summit ends, which is too bad because I certainly could have used (at least) one more day. There were many great minds in attendance and the set-up worked pretty well, but I think there were probably too many people in attendance to make it as productive as it could have been (the most concrete work got done in the last few hours, once we split into four smaller groups).

Unfortunately for me, we never got to what I consider the meatiest bits of the design of social software: methods for formalizing identities, 1-to-1 relationships and groups. I was hoping that we could get into how to make machine-usable descriptions of groups, since, eh, otherwise we're not going to be making a whole lot of social software (other than simple platforms for communication which people can use as media for social activities — not that there is anything wrong with that).

I really want to know what the possibilities are like when we move beyond this:

<group>
	<member />
	<member />
	<member />
	...
</group>

Ray Ozzie deliciously suggested automatic schema creation for groups but that's something that I'm not smart enough to envision (yet).

Groups can be “surmised” via analyses of 1-to-1 relationships and ad hoc groups constructed out of views I create on the relationships and facts-about-individuals inside my super address book (cf. this peterme post) but these don't replace explicit groups. More on this soon.

(Highlight of the meeting: on the first day, Scott Heiferman saying “The essential thing about social software ... no wait, I'm wrong.” How refreshing! Also, Brad's pics.)
` ` ` ` ` ` ` ` ` ` ` `

Inching towards the dream: we now have working relay bots (long way from public; sorry) that pass messages back and forth between AIM and MSN Messenger! Who cares? The idea is that the bots can act as a proxy allowing people in the game to IM with other players when they are not in the game (without ever having to divulge their out-of-game identity, if they don't want to).

Groups can register with the bot, address their IMs to the appropriate bot account and have the message distributed to other group members who are listening in on whatever platform they choose (game, AIM, MSN, Yahoo, etc.) Not sure if I'm explaining that well enough, but it is super freaking cool and extremely exciting! There are so many great pieces of systems out there, and being able to yoke them together makes possible all kinds of amazing things.
` ` ` ` ` ` ` ` ` ` ` `

My body, presumably in a protest move to decry the last few month's lack of sleep, ceased functions before dinner last night leaving me face down on the hotel bed until about 4:00AM, after which I was not able to get back to sleep (bad move, body). So, here I am in the 24-hour cafe across from the hotel, making good use of NYC's ubiquitous wifi (two open networks available in the hotel, two different ones available here, across the two lane street — last night I dorkily check mail while the cab idled at a red light).
` ` ` ` ` ` ` ` ` ` ` `

And, here we are in New York City. Surprisingly, upon reflection, it's been two whole years since I've been here (!) and I notice that there are European numbers of people talking on mobiles.

I meant to post this night photo of the construction going on next door for a few weeks now and this photo of the pumpkins I made for somebody on Halloween:(Thanks to Eric and Ben for helping with the pumpkins and also providing the background vocals during the pumpkin presentation ritual.)
` ` ` ` ` ` ` ` ` ` ` `

Back from Amsterdam and home for all of 40 hours. As usual, it is a little bit humbling to be back in North America and see, for example, how very wide and ill-used our streets are. I didn't have a chance to take pictures, but I am now officially fascinated by the 2-4 grade street designs in Amsterdam. I saw one that went:

pedestrians / bicycles / (lowered) cars / (raised) trams / bicycles / (lowered) cars / pedestrians

All this in the width of what here is a normal three-lane street, and all constantly used to near the maximum number of trips of each type.
` ` ` ` ` ` ` ` ` ` ` `

I wrote:

The web is closer to a providing something we could call a social environment, but even that seems a little dodgy — I'd be more inclined to say that the web is a medium for lots of smaller social environments, but one that is still generally parasitic on social environments (and cultures) that exist elsewhere (Limp Bizkit webrings don't spring up in the absence of Limp Bizkits).

Matt wrote:

The www is like a city?

Touché.
` ` ` ` ` ` ` ` ` ` ` `

Dylan Foley writes in with a better bookmarklet for retrieving definitions from Merriam-Webster's m-w.com using Mac/IE5 (which you probably shouldn't do anymore) and includes another one for searching xrefer as a bonus (scroll down a few posts past the m-w link).

` ` ` ` ` ` ` ` ` ` ` `

Matt Webb is smart as a whip and his weblog (¿recently? named The Strongly Connected Component) has consistently been the place where I have found the links that have most accurately hit the intersection of all the thought-paths that have me interested at the moment I was reading. Even when I don't have time to follow and read the links*, I dig his uninhibited enthusiasm for the day's fancy.

But I also think he was wrong when he wrote (perhaps summarizing a position not his own)

Now we've seen that social environments can be created technologically (what's worked: email, www. What's not been as big: MOOs), what properties do successful technologies have that we can harness and steer to not just communities but directed groups

First, citing email as a platform for a social environment seems pretty after-the-fact: email is a tool that can be used in social interactions (in the loose sense that any communication device can be employed in a social interaction) but the environment comes from somewhere else entirely.

The web is closer to a providing something we could call a social environment, but even that seems a little dodgy — I'd be more inclined to say that the web is a medium for lots of smaller social environments, but one that is still generally parasitic on social environments (and cultures) that exist elsewhere (Limp Bizkit webrings don't spring up in the absence of Limp Bizkits).

Second, saying that MOOs haven't been all that popular is misleading: no applications that rely on command line interfaces have been all that popular. Even things like irc and usenet never got that big until there were simple windowing interfaces: for most people (myself included), scoping conversations from interfaces and from other conversations is too hard if only a serial, linear presentation is available.

(I used pine, nn, and talk but still can't really do regular file management tasks in CLIs and switched to windowed interfaces for all of those applications (except for talk, obviously) as soon as they became available.)

The first graphical MUDs* have hundreds of thousands of players shelling out US$150+ a year for a subscription (plus the box upfront). Lineage has 4 million players in Asia (almost all South Korean).

The first thing that really looks like a graphical MOO is coming out in a few weeks and is pretty likely to be the biggest game of all time by revenue — others are the works. So, we'll see how big MOOs get, now that interfaces which support the possibility of mass popularity are available for them. Give it a few years.

* Two recent wish-I-had-time-to-read-them-right-now links: I Have No Words & I Must Design (#SCC) and Message Types in Goal-Oriented Discourse (#SCC)

**When I say “MUDs” I mean games (usually RPGs, and usually involving orcs) where the players reside in a world which is created in a different context and using different tools than those by which the game itself is played. MOOs are less goal directed and have a greater emphasis on the things created by the players, in the game. I don't know how closely my usage corresponds with the consensus usage.
` ` ` ` ` ` ` ` ` ` ` `

(I've tried a few times over the last 12-18 months to get this site up and going again, but I've always been too swamped. Now I will try again, with a little bit more spirit. Because "at the end of day, it's the team that got the puck into the net the most times that's, uh, gonna win the game.")
` ` ` ` ` ` ` ` ` ` ` `

My car only gets broken into at the most inconvenient times. Like at 4:00AM on the morning I leave for Amsterdam for a week (previous incidents include Friday night on a long weekend and Christmas eve; the car spends far too many days with broken windows). But at least I have a 10 hour flight ahead of me ... I haven't got that much sleep in a really long time.
` ` ` ` ` ` ` ` ` ` ` `

(Non-food.) Derek sends me this great paper entitled The Lessons of Lucasfilm's Habitat “arguably one of the first attempts to create a very large scale commercial multi-user virtual environment” with a client that ran on a Commodore 64. Some things are still true today:

The essential lesson that we have abstracted from our experiences with Habitat is that a cyberspace is defined more by the interactions among the actors within it than by the technology with which it is implemented. While we find much of the work presently being done on elaborate interface technologies -- DataGloves, head-mounted displays, special-purpose rendering engines, and so on -- both exciting and promising, the almost mystical euphoria that currently seems to surround all this hardware is, in our opinion, both excessive and somewhat misplaced. We can't help having a nagging sense that it's all a bit of a distraction from the really pressing issues. At the core of our vision is the idea that cyberspace is necessarily a multiple-participant environment. It seems to us that the things that are important to the inhabitants of such an environment are the capabilities available to them, the characteristics of the other people they encounter there, and the ways these various participants can affect one another. Beyond a foundation set of communications capabilities, the technology used to present this environment to its participants, while sexy and interesting, is a peripheral concern.

And then there are these two lines in a MMOG blurb on Tom's Hardware:

Just look at the popularity of those mundane message boards on the Web. What if that social interaction could also incorporate gaming activities?

To both, I say “exactly”.
` ` ` ` ` ` ` ` ` ` ` `

NYT Travel section sez:

Vancouver is to Asian food what New York is to European: A place where a cuisine is often as good, and sometimes better, than in its country of origin. So splendid is the Chinese cooking in Vancouver that it rivals that of Hong Kong; so original and satisfying are its Japanese, Indian and Thai restaurants that frequent travelers to those countries who are familiar with Vancouver see little difference in the quality of the best establishments.

Vij's, my all-round favorite restaurant is first on the list and the rest is pretty standard. Lumière (13 course, 3 hour menu signature menu — happy birthday, honey) gets a mention, but Ouest doesn't (it should). And I've never had great Thai here, but I haven't been to any of the ‘best of’ places yet. Places to go! (But whither Guu? or Gyoza King?)

Maybe soon we'll even get our own damn Chowhound's board (you tell ’em, VanEats!)
` ` ` ` ` ` ` ` ` ` ` `

After my third visit, Guu with Garlic (1698 Robson, near Caper's) is now on my great restaurants list. Any Japanese place that doesn't serve sushi, tempura or terriyaki is automatically interesting, but GuuwG is takes it all the way to weird. Sample dish, as described in their badly photocopied hand-lettered menu:

Freshly cut Japanese radish mixed with scallop and spicy cod roe in a mayonnaise dressing spread over three types of lettuce dressed with a soy based dressing topped with crushed potato chips.

They know how to dress. (Also, I had rice with “wild moontain plants”.)
` ` ` ` ` ` ` ` ` ` ` `

Expecting the worst, I went to see John Zorn's Electric Masada project at the Vogue on Friday night. Though I have heard some great Zorn, and though I am a fan of some of the other members in the lineup, it was a group supremely capable of some serious atonal, irrhythmic fooferah (the sounds that make you hope that this is not literally the “avant garde”).

The first 30 seconds or so exceeded those bad expectations: skwrak, pfrt, eeeeeeeeeee, wuuhh-bing, etc, but the band slowly gathered steam and it was quite interesting to watch Zorn create/edit/delete the arrangement in real-time using some gestural command language. After the first few tunes the melodies came out and the surprising phrasing and unexpected closures shone. He is a great composer.

Medeski (MMW) did his usual mond-blowing extremely physical ultra-high speed organology —I love watching him play more than just about anyone. He is essentially a drummer who happens to play the organ: his hands fly up and down and always manage to land with the fingers splayed just right to hit the chords in a way that seems perfectly analogous to the way hand drummers' minute alterations in angle, pressure, release and damping determine the tone (and the percussionist, Cyro Baptista, oya!!! — best new find).

There were a few lousy solos and noisy bits, but the second set smoked and I was quite happy to be wrong. (Seattleites, tomorrow is for you. Note: google says Seattleites over Seattlites 4,170 to 1,590)
` ` ` ` ` ` ` ` ` ` ` `

33 Customer (p)reviews of The Sims Online at EB Games. Ouch. I'd still bet anyone $50 that it will set a record for most revenue from a single title, but maybe, just maybe, a thin-client game of similar concept with a lot more flexibility in the fundamental model and more hooks into people's pre-existing online personas will prove to be just as popular ...
` ` ` ` ` ` ` ` ` ` ` `

I got my copy of Christina's Information Architecture: Blueprints for the Web (pays to be on the right mailing lists!) And she also recently launched (with others) the AIfIA. Congrats Christina!
` ` ` ` ` ` ` ` ` ` ` `