The ARCH

This is a guest post by Jaz Beverly, cross posted from her blog ‘Metaversity.’

photo by Jaz Beverly

On day three of Architecture Days on Orange Island, there was much discussion of balance between how much is preplanned, and how much of community architecture and infrastructure should emerge from residents. General agreement emerged that too much pre-planning kills community and can create very boring architecture. But too little can lead to a visual jumble and lack of community.

Michael Linden made the point, contested by some, supported by others, that there is a certain amount of infrastructure – roads, railways, monorail, green spaces – that can be extremely difficult if not impossible to retrofit once a community has moved in. Better to provide those up front, and, as the Linden Labs are currently working to do, perhaps provide a suggestion of a theme for an area, and then let the people who move there work out the rest.

JeanRicard Broek riposted that historically communities have come together in communal places oriented around trade, but in Second Life teleporting makes it possible for the shops to be anywhere. No need for communal spaces around economics. But in Caledon, as well as in other sims, real community has emerged from shared interests.

He argued that creating a themed space and expecting people to move in and form communities of interest afterwards is going about it backwards.

(I found this particularly interesting because I wondered how it might affect themed builds like my beloved hometown, SL London. It has a talented building and management team and an active core community of devoted residents. In the process of creating room to grow, a significant number of shops and apartments have recently been built that are now waiting for people to show up and rent them. The interests that bind the group are loose and general, such as an affection for RL London, the UK, and modern urban aesthetics or the desire to party and play with other international citizens. These broad themes facilitate community diversity, which is much appreciated by current residents and visitors. But will the general ties that bind us also make it more difficult to build and maintain a strong base of participants, since they can easily have these very general needs met many places in the metaverse?)

Broek asserted that in Second Life the most successful communities and builds have come from people who have come together around common interests – the people gathering FIRST – and then creating a place together where those interests could be shared and other people with similar interests could join them.

(It does seem likely that Caledon and her sister communities indeed benefit both from the people having coming first and from how specific their common interests are. Such specificity results in community events and connections that are only relevant and available in a very limited number of sims. Intentional community does tend to arise from shared values, ideals, and/or vision. But there is also tremendous amount of unintentional community in the world – coworkers, neighbors, family members, etc. who may have very little in common except geography, the same boring work environment, or some random DNA in common and yet they manage to come together in meaningful and supportive ways.

But I digress…)

Michael Lindon suggested that over the next 10 years, the way Second Life is structured will naturally evolve and bring with it changes in architectural and community patterns. What if instead of parcels being in square meters they were in 3 or four dimensions? Or if multiple audio streams were possible within a single parcel? What if tools emerge that allow communities to build things like road collaboratively? Simple things like this could radically change the space in which things are built and the ways people come together. [Such ideas, he was careful to note, merely serve in this context as hypothetical examples and in no way are meant as any kind of indication or announcement of specific upcoming features from Linden Labs.]

There was also some buzz in the panel and in among the attendees about how thus far most of the architecture in SL merely reflects the RL, but how that will change as time goes on. This thread was picked up by a later panel, in which there was some discussion of asking the pixel what it wants to be. The later panel and discussion also raised some interesting ideas about where architecture in SL could ultimately evolve including

(1) being more like a liquid than a solid artifact — flowing in response to how the space is used, making the right response in that moment of time

(2) more hybrid vigor from collectively generated work than from architecture produced by a single person’s vision

(3) more deleting and rebuilding! One panelist challenged communities not to be so static in their builds but to keep changing things. Why should buildings be any more static than avatar clothing or shapes?

The other session I really appreciated was very brief, but involved the presentation of a truly remarkable and intriguing achievement: a WikiTree to facilitate collaborative design. The presenter, Keystone Bouchard (Jon Brouchoud in real-life) co-founder Studio Wikitecture with Ryan Schultz (Theory Shaw in SL), worked with i3D to develop a tool that would help to evoke the much vaunted wisdom of crowds. They believe that decentralized groups of people can work together to create much more than the sum of their individual parts, and that this is in fact occurring and will continue to do so especially in the three dimensional web. He asserted that this will result in an explosion of advances in all sorts of areas that will rival the industrial revolution.

His current project, wikitecture, is about wondering how this could be facilitated in Second Life. They began by creating a meeting place for architects, and inviting them to build their ideas of how a particular type of space could work. The results were all individual and incoherent, with everyone just building something in their own style. So they learned collaboration is more than just letting everyone build in a common place.

In the next phase they asked everyone to allow mod and copy on their builds, so that they could use and build on each others’ ideas. They also included a Flickr group, so people could share photos and ideas about what they were working on, and developed an archiving kiosk which allowed rollbacks – which was closer to how a wiki would work. All of this resulted in better and more collaborative design.

In the current, 3.0 phase of the project they have built a new tool to support inworld collaboration – the WikiTree. The trunk contains tools for both construction and evaluation of forms. They ask everyone to use prims they rez from the tree when building models for group consideration, and the embedded scripts in those prims communicate back to the tree, and permissions are pre-coded to allow mod and copy.

Each design resubmitted to the WikiTree creates a leaf in the canopy that has a different design idea. Click a leaf and that idea is rezzed in the space. Each leaf is branch off of its parent leaf if it is a modification of an idea taken from the tree. Everyone working on the project gets 3 positive and 3 negative votes. The votes color the leaves to reflect the votes received regarding each particular idea. The green leaves are most popular, red least, with yellow and orange in between. The red leaves are automatically pruned by the tree.

In this way, the WikiTree’s canopy represents visually the evolution of the design concept, and which ideas are most successful. When you submit a new leaf, it is also automatically reproduced on a web site so that it can be accessed out of world.

Using this method, a group of architects collaboratively developed a plan for a medical clinic in Nepal in this year’s Open Architecture Challenge. The design developed collaboratively using the wiki tree won the Founders award, which proved some potential. They are now working on a collaborative design of a virtual classroom for university of Alabama. For more info, see the blog at studiowikitecture.wordpress.com

Future plans for the WikiTree include a greater degree of granularity so that it would be possible for the group to vote on a particular aspect of a building in addition to the design as a whole. The team isn’t yet ready to open source the code, as they are still developing protocols and methodology. Still crashing a bit.

One thing the model shows is that even a little change or improvement can be a big contribution. Better results are achieved when participants submit designs to the tree for group consideration early and often (rather than attempting to submit a finished, polished design after working solo for an extended period of time).

One final takeaway for me from this experience was the realization of how much more efficient panel discussions are SL than in RL. As questions arise, they are submitted via chat and picked up by the moderator whenever they best fit the flow of the discussion. This also gives panel members a chance to see what questions are arising and reflect before being asked for a response, and other panel members can chime in with additional feedback without interrupting the speaker. A model of elegant efficiency!