At SIGGRAPH 2008 a meeting on Multi
User Virtual Environments (MUVEs) and social networking took place.
This is a report on the presentations and panel.
Location: Los Angeles Convention Center
Room: 507
Date: Wednesday 13 August, 2008
Meeting Time: 12:30-2:30pm
link: http://www.siggraph.org/s2008/attendees/birds/
Chair: Chris Thorne (dragonmagi
<at> gmail <dot> com)
Many thanks to the wonderful
group of people who came to the inaugural MUVE Moot and were prepared
to spend over two hours of their time giving individual presentations
and for joining the panel at the end.
This document is mostly a transcription
of the presentations and provides links to other information,
including, hopefully the video. Note: the video was taken in 1080i HD
and represents a bit of a challenge to process on my equipment and
put online, so it will be awhile before any of it appears.
The meeting was organised to provide a
venue for virtual world creators and users to present on products and
also explore the concept of merging social network and virtual world
technologies. Speakers were given up to 10 minutes each and there was
a panel at the end which went 15 minutes into teardown time (about 35
minutes)
Speakers
1. Chris Thorne, PhD, The University of Western Australia. Meeting Chair and MD of http://vrshed.com.
2. Michael Wilson, CEO of There.com, http://www.there.com/
3. Mick Brady, Professor Emeritus, Russell Sage College, http://www.mikimojo.com.
4. Doug Twilleager from the Wonderland group, Sun Microsystems, https://lg3d-wonderland.dev.java.net/
5. Rafhael Cedeno, CTO & Co-Founder, The Multiverse Network, Inc., http://www.multiverse.net
6. Don Brutzman, Associate Professor, Naval Postgraduate School, Monterey, CA, http://web.nps.navy.mil/~brutzman/
7. Tony Parisi, Chief Platform Officer of Vivaty, http://www.vivaty.com/
8. Peter Schickel, founder of Bitmanagement Software GmbH, http://www.bitmanagement.com/
9. Greg Spencer,
Senior Software
Engineer, Google, Inc., Google Lively, http://www.lively.com.
Chris described his motivation for the
meeting came out of the expectation in 2007 that virtual worlds and
social networking technologies may merge to form what he terms super
social networks.
Chris went on to present 3D environment
and virtual world projects he recently lead. One project involved the
rapid (in 3 months) development of a virtual world called The Virtual
Universe Project (http://www.csse.uwa.edu.au/virtual/). This project
began as a virtual mirror of The University of Western Australia
campus but is intended to expand to include other places. A number of
other students at the University of WA and various people from across
the world contributed. Last year Chris lead another project involving
a university team that put the campus into Google Earth as part of a
Google 3D campus competition, which the team won. Videos were shown
of the projects and a prototype visualisation of the Australian
Square Kilometer Array radio telescope project (ASKAP).
Michael began with describing a number of firsts that distinguished There.com.
There is the first social virtual world, launched in 2003. One interesting thing was some people actually paid to be in the beta program.
First virtual world to make business partner technologies: some original designers were data warehouse providers which gave There the immediate capacity to handle huge transactions from the virtual world. It is able to throw out thousands of dimensions of information "out the back end" activities in the virtual world, which is great for advertisers.
There was one of the first to have a
licence to allow participants to watch MTV, virtual Laguna
Beach, in the virtual world (from 2005). Michael sees this as a
tipping point in online virtual worlds because it did not require very
specific hardware/software platform just so people could use it for
network TV viewing.
There is unique because it
provides a
PG13 brand protected environment. The reason for that is they thought
more people would be offended by seeing non PG13 content than not
seeing it, and that seems to have held true so far. They also brand
protect screen items so if you see a product of a certain brand it is
the genuine product not an imitation. Michael listed many high
profile commercial partners.
There have over 3 million users with
average age of 22. 58% are female, 70% are between 13-26. An average
customer spends 20hours per month in There.com which Michael noted is
still a small number compared to games like World of Warcraft. Data
from Comscore.
Michael next talked about the idea of
whether virtual worlds were games. There is bringing game-like
aspects into the virtual world, such as the sort of environment,
challenge, aspirational and rewards systems one finds in
games.
Although the graphics are not as hi
fidelity as some might expect, There runs on even 2003 vintage PCs
and over 56k dialup. It does, however allow incorporation of
streaming video.
One can use external geometries such as
content generated in Collada, 3DS Max, gmax, or Maya.
Michael demonstrated some of the game
like aspects such as Cosmogirl magazine content, an aspirational
example. The Cosmogirl Village is an online interactive magazine.
There is physics correct (if
some physics happens every witness sees the same behaviour) and
includes voice over IP and embedded web browser. Michael also showed
an example of the challenge aspect: the There Summer Games - a series
of user sponsored games.
Mick described himself as primarily an artist who had thought he had retired until, about a year ago, he was invited to help build a virtual campus for the Santa Barbara City College in Second Life, and, for some reason, things suddenly got busy! He was invited to participate primarily because of his track record as a writer, visual artist and innovative thinker. Mick's presentation, "Art is What Happens When You're Rezzing Other Prims", is about the nature of art in Second Life. He makes the interesting observation that the term 'art' itself is not quite big enough to encapsulate what is happening in Second Life, and therefore suggests using the term 'creative', which refers not only to the creation of quality content, but even to the the tools used to create that content. In a world where even the ground one walks on is a work of art, the very notion of art becomes an elusive thing. This idea, along with Mick's mention of artworks created through collaborative efforts, was later picked up on during the discussion part of the meeting.
The first example shown, was from "Kiss the Sky", an exhibit of a new form of art called Hyperformalism, a show which was displayed across a large section of the sky in Second Life. Many of the works were fully interactive and included motion and sound. The next example was the Garden of NPIRL Delights (Not Possible in Real Life), was the largest exhibition in Second Life, created by over 100 artists working in collaboration. The working premise for this show, as might be expected, was to create things which could not be created in real life. Much of the creativity on display in this exhibit came from unexpected quarters and many disparate walks of life, such as PhDs in English, engineers and medical surgeons.
One work notable for its subversive inventiveness was a chair that morphs your avatar into weirdly contorted forms resembling nothing if not a Picasso sculpture when you sit in it: in other words, you suddenly realize that you've become the artwork!
Mick described many other fascinating
art works, so I recommend you watch the video for more details, as it
is too much to describe here; the subject being primarily of a visual
nature, it is far better to see the art itself. More on this subject
can be
found at http://www.mikimojo.com.
One of the motivations behind
Wonderland came about when Sun surveyed its employees and found that,
at any given time, 50% were not in the office. Many of those that
worked remotely had some issues with working from home. Project
Wonderland is an open source virtual world collaboration platform.
Supports hi-fidelity spatialised audio (if a sound is from behind you
to the left it will seem to come from behind to the left). Although
Doug was showing a 3D client, the platform supports different types
of clients, even a connection via a phone line.
For the server side Sun uses a game
technology for hosting high performance online games, called Project
Darkstar.
Sun's own example world project is
called mpk20, can also be downloaded for free. It supports
application sharing in the virtual world. Users can take control of a
shared application and others will see the changes they make, then
someone else can take control and continue the collaboration.
Sun is currently working on increased
scalability in Darkstar. With Wonderland they are moving to jmonkey
engine - a gametech. Also moving to include Collada import
because of its increasingly widespread use in the industry. Sun are
rebuilding their avatar system because an important aspect of virtual
world interaction is the inclusion of facial and avatar expressions
are to communicate emotional quality when talking to someone face to
face. Sun are also developing a game engine wrapped around the client
system to exploit new multi-thread and multi core architecture:
called mtgames, which will allow its performance to scale nicely with
the number of cores.
Examples. Cone of silence. MiRTLE:
eLearning system by University of Essex. App share , vidcon, Music in
Wonderland - not confined to normal spaces. building
applications in world.
Rafhael described the Multiverse
platform then went on to talk about what has been leaned about what
people are trying to do with technology and how it is interacting
with social networking.
Multiverse is an end-to-end multiuser
server and client virtual world platform built to handle a large
scale user base. There is also a network of virtual worlds you can
navigate. Some of the differences from other virtual worlds are that
Multiverse is supporting a single browser client into multi virtual
worlds mapping and also looking for people to develop video games as
well as social worlds on the platform, all with multi-media:
flash/video/socnet integration with social networks such as facebook.
Where to in the future? Avatar
portability for one.
Multiverse wants to get away from people reinventing the wheel, which makes no sense. Therefore reusable technology is needed with support for social nets, machinima, etc.
Multiverse is exploring what does it
mean to have APIs that expose all these 3D, virtual world and social
net features. e.g. Similar to how SL has enabled a lot of artwork to
be built on top of its platform.
Next Rafhael showed examples. Hi
fidelity virtual worlds with scalability across platforms of
different capacity (such as high end gaming systems to lower end
ones more optimised for 2D). They are trying to break limits of
gameplay. Going for hi fidelity with extensive simultaneous
multi-video playback.
Rafhael returned to the meeting theme
of social network and virtual worlds. He expressed interest in the
"matrix aspect": A virtual world managed on a machine
with no visual display [the server network] which keeps track of
where people and objects are, what they are doing, what people are
saying. This is not just applicable to a 3D spatial model but
includes the huge digital world of facebook, flicker, myspace etc.
We have the beginnings of people having their app inside of facebook
or who bring facebook and flicker content into their virtual world or
go the other way. "Its about virtual worlds reaching out to the
bigger Web". It's happening with IM and voice calls, beebox
etc.
Multiverse also has a Flash client to
connect to same server platform - a much lighter weight, but
someone using this light client can be interacting with someone using
a hi fidelity 3D client.
Can do the kind of interoperable things people want if we have 3D browser with Web technology embedded. It's what people want / are meant to use and what Multiverse aims to support.
Don described the standards based
developmental work of the Web3D Consortium (Web3DC) and NPS in
particular. Their efforts are to support some of the great work in
virtual worlds being portrayed in the meeting.
Don described X3D Edit : originally
designed as a tool make it easier for people to create and edit 3D
models in an ISO standard form that is easily ported to other
applications. This tool has been extended to more of a collaborative
integrated development environment.
X3D edit is created in Netbeans -
an open source platform from Sun Microsystems which is capable of
integrating a great many things, not just for use as a programmers
integrated development environment. You can plug in other programs
e.g. for voice/chat text/ 3D viewer collaboration. The text chat is
implemented in XMPP - directly supported by netbeans.
Don described how 3 people collaborated
between US, UK and Australia, editing documents and working on X3D
specification.
He described how NPS have added the
IEEE simulation protocol, DIS. The DIS standard is serialised to XML
for transmission across a network, including networking between X3D
Browsers. For the future, the Web3DC are looking at supporting the
binary XML W3C encoding for more efficient XML based networking. They
also hope to build more tools ... exposing open technologies which
people can integrate into a variety of applications.
Don then talked about including avatars through the Hanim specification. Mapping to/from common formats. He noted a student thesis compared SL, Olive, Wonderland, X3D - assessing them for quality and other factors and recommended it as a worthwhile read.
Vivaty is part of the Web3D Consortium,
helping to develop international standard format for interactive
online web 3D. Tony described the Vivaty suite of tools which
includes an X3D Web 3D viewer, Vivaty Player and a 3D modelling tool,
Vivaty Studio.
Tony said the avatars are built upon
the humanoid animation specification that Don mentioned, but with
extensions for dynamic behaviours. He also described the SWaMP
protocol that Vivaty has published for multiuser virtual world
communications. The virtual worlds protocol is in development with
the IETF at this time.
Tony explained Vivaty's approach was to
build a web architecture on web standards and plugins because they
believe that is the best way to deliver Web 3D content that
integrates with social networks, other web technology and the massive
quantity of information accessible on the Web today. Their aim is to
ensure the infrastructure will scale to support the hundreds of
millions that currently use social networks and the even greater
number that use the internet. What Vivaty is not aiming to do is
reinvent the wheel. These are some of the key reasons Vivaty chose to
build on open standards - they provide a solid basis for building the
future 3D Web and give developers the most freedom of choice.
Tony described the many content
creation options available for their platform, supported by Vivaty
Studio. Vivaty is currently running a content creation competition as
part of their outreach to developers.
In the future Vivaty will put in place
a virtual economy to support virtual commerce.
Next Tony demonstrated Vivaty, entering
it via facebook. He gave the first live demonstration it worked on
the very patchy wireless connection we had at the time. He showed a
virtual room with interactive content: facebook, flicker, youtube, and
was joined by Keith Victor's avatar during the demonstration. An
attempt to stream video however was a bit beyond the poor connection
capacity.
Vivaty takes the approach that not
everyone is a modeller so they provide room templates with pre
modelled items and avatars. Several other presenters at the meeting
take a similar approach and agreed that this was an important
approach for making Web 3D accessible to the masses. There are also in
world editing options in Vivaty to modify the appearance of objects
for those who wish to be more creative.
Snowcrash was mentioned twice in the
proceedings: once by Mick Brady and Peter opened his talk by asking
who had read Snowcrash. Peter was originally part of the Blaxxun
company whose business model "was Snowcrash". Time has
changed, 10 years on, now is the time where a lot of people are
trying to enter this avatar and virtual world space again and
bitmanagement aims to help with providing the technologies to enable
this without people having to reinvent the wheel and using a platform
that as been out there for "10 years or more with tons and tons
of bug fixing".
Apart from being CEO and founder of
Bitmanagement, Peter is also vice president of the Web3D Consortium
which started develop the VRML and now the X3D ISO standard.
Bitmanagement software also supports collada. Called BS Contact the
3D player supports vrml, x3d, collada, kmz. It has realtime shaders
and physics, stereo rendering.
There is now a linux version, windows
and a mac version in the works. BS Contact can operate standalone or
as a web browser plugin. Bitmanagment has a whole suite of tools, but
the most important for this meeting were BS Contact and BS
Collaborate- the multiuser server. Bitmanagement have a wide
international customer base and Peter showed a map of customers all
over the world.
Peter then showed examples such as a
soccer stadium with interactive avatar/viewpoint avatars where you
can take the viewpoint of players and participate in games.
Visualisation of 3D cities. Important
to visualise realistically. He showed San Francisco and Prague. He
showed an impressive hi fidelity model of a 200 square kilometer city
area generated automatically in just 4 days. Peter explained we are
now into a new era where it is very cheap to generate a city model
automatically and we will likely be exploring and interacting within
these city environments in the not too distant future.
Peter then showed an example of mobile
"life logging" in a virtual San Francisco, using a
product called Raygun. This a multiuser city simulation system for
mobiles developed by Planet 9 Studios in collaboration with
Bitmanagement. Peter's comment was that he thinks "reality"
is right now more important than "virtual".
Peter showed a number of other applications enabled with Bitmanagement's software, a notable one was the Bettlesmann 3D earth encyclopedia. Please see the video for more details.
Lastly Peter showed BS Contact running on the ee PC mobile platform with smooth realtime performance.
Lively was started as a Google 20%
project about two years ago and it has now been turned into a
project.
In Lively you create a room and can
send url to friends. Unlike SL, Lively is an unlimited number of
interconnected "rooms". Connection is just via hyperlinks
like on a web page. Rooms created by users can be embedded in any
web page, facebook integration for inclusion with facebook
applications.
Lively is a Microsoft API with Active X
plugin, so it runs inside the browser. Currently only supported on
Windows. There is no user created content right now, everything is
created using catalog. Lively will have user generated content,
probably go similar route to Vivaty: importing Collada content and
content from other authoring tools. Lively do not plan to have a
system like in SL.
Google's approach is focused on social market. They want people chat and interact, watch youtube videos. Lively supports Google gadgets, which are based on javascript, html and Flash.
Right now lively has a large catalog of
100,000 items created by its own team. The project is targeting ages
13 and up.
For the future: support for user
generated content, improved support for external formats: Sketchup,
Collada, 3D warehouse... plus support for other platforms than
Windows.
Examples were shown of unusual things
people can do even without user generated content, e.g. an amazing
staircase that spirals into the distance made of hundreds
(thousands?) of "rugs".
Perhaps in keeping with the name, Greg gave the second live interactive demonstration at the meeting. Greg even went one step more challenging than Tony: he went live online under a virtual machine on a Mac. He started with a lively "room" with youtube video in the background, with avatars, animations etc. Some of the in-world modification features were described, illustrating how users could customise the environment and avatars using the 3D warehouse and customised using the in world editor.
The
Panel, from left: Rafhael Cedeno, Michael Wilson, Peter Schickel,
Chris Thorne (Chair), Greg Spencer, Tony Parisi, Mick Brady, Doug
Twilleager.
First question was about user adoption
at the point of installing the software. He noted that in their
surveys, for his company, with a signed java applet, they lose about
90% of potential customers "at the point saying ok". He
said "that's why we went to Flash". He asked to give
insight into plugins / executables and how they use them and provide
some idea of how many people they lose at the point of installation,
and what platforms to choose.
Tony Parisi answered: he can't give
figures as yet but the biggest fallout was due to lack of support for
Firefox. It was just a matter of timing and the complexity of some
javascript. Firefox support will be coming soon. Tony suggested a
reason for the fallout was because the many people who "liked
new cool stuff use Firefox" and so that is one of the platforms
they will be supporting. Tony thought people who point at the
requirement for a plugin as being the main user adoption impediment
and the main cause of fallout are worried about the wrong thing. He
observed that, compared to the recent past, downloading and
installing plugins/programs is common and not difficult. He recalled
the time when more people had the Cosmo VRML player than flash. Every
day now we are prompted for an update of the itunes/quicktime or
other product. People don't blink an eye [at the need for such
downloads]. So he does not think people should be worried about the
downloading aspect at all.
Rafhael added they were seeing similar
things and the biggest problem they found was not fallout at install
time but after people start using the world and they see many drop
off , perhaps depending on how compelling they found the world. Also
very interesting to see what other technologies are doing, Flash for
example is growing more 3D-like capabilities and shockwave has a 3D
engine built on top of it.
Peter added that it was very important to support Firefox: 80% of their customers used Firefox.
... lots too faint to hear
Michael said they have 30% of falloff
is because of people not having permission to install the software.
Especially parental controls for the young are in place. This is
important because virtual worlds are mostly used by the younger
generation.
It was at this point that a few heads
turned to the back because a guy near the back disclosed up front
that he works for Linden Labs. He noted presenters eschewed online in
world content creation for offline created content and he asked the
panel to comment on that and "what is the stench on in world
content creation" ... basically all the emerging technologies
seem to be "shying away from in-world content creation".
Greg answered that Lively were shooting
for users who want a simplistic environment, to have a nice world
where they can chat and interact without having to learn difficult
creation skills or spend too much time/effort on creating a good
looking environment. Google has a good modelling tool (Sketchup) and
artists who can create sophisticated content but did not want to
complicate things for people who just wanted to have a chat. They
thought it would be too much for the general users.
Tony agreed, adding that from a
consumers perspective this was true but from a producers perspective
there is another reason supporting the creation of content offline.
Thousands know how to use Sketchup, Maya, 3DS Max and they do not
necessarily want to learn to use something different, and it [the
Second Life in-world creation tools] is very different from what they
know. Similarly with scripting. Vivaty's strategy is centered very
much around the type of content creation approach used by the
majority you would find on the show floor at SIGGRAPH, and that means
supporting the products of those enterprises. Vivaty does not have
anything against in-world content and they have some in world
editing/modification tools but their main approach is to support the
common content pipeline.
Peter agreed, Bitmanagement takes a
similar approach in supporting X3D, Collada, VRML and have developed
a plugin for the open source Blender modelling tool to help with
getting content into Web 3D format.
Someone else in the audience commented
that Second Life had enabled a whole class of people to create
content that have been put off by the interface of Blender etc ...,
people who were not able to do this kind of stuff before.
The Linden guy then pointed out the one
thing that the presenters [except for Mick Brady] had not addressed
was communal content creation. That is the one thing that SL has
enabled and it is "a little worrying to see that missing",
we are talking about communal spaces and ...that is the one huge
fault with offline content creation ... people like to go visit each
other and work together.
He agreed SL in-world content creation
using in world tools is a PIA, but someone can always make a better
version of in-world content creation tools. In SL, very few people
create things on their own. With offline content tools you miss that
communal creation process. People create together, what is called
server side creation: you make something and everyone else sees it
and can add/modify (if you allow).
Another in the audience said:
Just as artists are creating art in SL
that goes beyond just wall painting and flicker stuff on the wall,
educators are also investing heavily in SL. Not just because they
can educate with an online tool but SL is actually changing education
... so that educators can provide the facilities for students to
become involved in what they are learning ...it is a growing model
that students take what they learn and make something of it. In SL
students and educators are given those tools to take course material
and what they have learned and make something with it. The question
to the panel was: what are the panel members looking at in terms of
in world creation tools for education and things like facebook?
Second question from the same person:
"My daughter is 4 and not
represented here is the [builder?] bear, barbie and [webkin?]....,
and my daughter got into barbie and asked why can I not colour on
that dress?" So that is a user generated driver. She does not
want a "pick from this" type of system - which
seems to be the model offered by most of the presenters. What sort of
research are you doing for the [young] people of the future who will
be trying user generated content?
Tony pointed out the one thing he could
not demo because of network lag was any of the objects can have
properties changed ... like change colours or load texture from a
url. Not like a full painter program but allowing simple changes like
colouring/texturing nonetheless.
Michael: We have similar tools for
painting etc. Our problem is complicated by the fact that we have to
review everything for texture theft and inapropriate content, such a
pornography and [There] still don't have the tools to handle that
properly. Michael then discussed the issue of content theft: it is
great to enable everyone to create textures and other things but what
happens when you start getting texture theft. When everyone can
create everything and no one can create a market for everything you
start getting things like texture and model theft. That's the next
iteration that everyone has to start thinking about.
Mick added comments about building
educational content for Santa Barbara ... they started the Serious
Game Design Institute ...tried to create new ways for educating
people getting into game technologies. What they found was that SL
provided ... a creative working environment where people had to start
thinking about working together as a team and collaborating. "The
creative of the future is going to be a multidimensional thinker"
and this is one of the things that is going to change education:
education will have to incorporate wider collaborative environments
catering for broad based analytical and creative thinking which
becomes possible in SL. New ways of thinking and doing things. You
can create mind boggling things that have no correspondence in RL and
this is also mind opening and what will contribute to new ways of
educating.
Next, Michael was asked to comment on
the resources needed to review content for inappropriateness/theft
before it could be made public. He replied that There has actually
turned it into a science: they have spent a significant amount of
time building tools to assist in the process and this has helped a
lot but it still requires human in the loop and it is impossible to
keep up with the volume of content generated. He noted facebook and
youtube have to review content as well.
The next query was about what are some
of the challenges the panel sees for virtual environment creation,
solving a lot of different problems with virtual environments and
transporting avatar between them.
*** It was at this point the camera ran out of battery :( ***