When the realXtend project was born, we had a discussion about standardizing virtual world platforms. Much like world wide web has a standards to make sure that many kinds of servers and browsers can talk to each other.

One comment by Monday Beam at Twitter made me think about how people may see the efforts to standardize virtual worlds. How do they look like? What does it mean to get them standardized?

The standardization of virtual worlds does not mean that every world needs to use the same technology, or the same protocol, or even the same 3D approach. Long live the diversity of look and feel, theme and style.

Maybe the best way to explain is to define the effects of the standards.

Without standards:

• separate client software for each virtual world platforms, and most of the worlds use proprietary platforms
• virtual goods can be made for and used only on one specific world
• Avatar names are valid only within one world, you need to find your friends again if you go to the next world
• Moving between worlds means usually shutting down the connection and launching another client software
• If you want to visit a world you have not been before, you need to fill in web forms and to memorize new passwords

With standards (optimal situation):

• One client software for all standard virtual world platforms, capable of using plugins to enable world specific functionalities
• My virtual goods can be imported and exported around – if it makes sense – but technically can be done
• My Avatar and avatar name can be used on the worlds Ichoose
• Moving between worlds is as easy as clicking a link
• You can choose how much you reveal personal information to the world you intend to visit

Technology should not place limitations here. The ones who should be able to place limitations are both users and world service providers. Why? Lets take a look of some use cases:

Virtual Church

The virtual church owner may want to place dress codes, and even enforce them as a server setting. The user who enters the virtual church gets a selection to a) go away or b) accept automatic change of clothing during the visit. Some items will be disabled – like your favorite two-handed sword can not be shown or used.

Children using virtual worlds

Parents may set the age of the user to the virtual world browser, which then automatically limits and modifies items and avatars to make the world a safe place. An example could be that all the other avatars are shown as decently clothed human beings even if they are really naked or using scary costumes – which is quite common…

Ants role playing game

The illusion of the game world is broken if users appear as humans. Even the avatar names are Ant-like to fit to the mood of the game world. When a new user enters the world, her avatar name is twisted by the server to make it fit to the Ant world, and the user appears as an ant. The new ant-user may carry some facial or other features to make them look different from other ants.

Conclusion

One thing is certain, regardless of standardization, we will see more and more different and diverse virtual worlds. Lets just hope that things get a bit easier for the user. The real benefit of the standards is that it creates trust and confidence to the whole phenomenon and makes the virtual worlds mainstream adoption really possible.

Maxping wrote a while ago about the special requirements 3D web places to Internet and the end user hardware. This is a focused article how things look from a mobile device’s point of view.

When enterprises start to take virtual worlds into serious use, they will require mobile device connectivity to their virtual applications. Many business people are traveling a lot, and will continue to do so also in the future – at least until virtual world technologies enable as good fidelity as the real life.

During the travel, business people attend to meetings with their mobile phones. Virtual world meetings are not an exception to this rule and there needs to be an access to those from a mobile device.

Even today’s virtual world solutions allow a degree of connectivity for handhelds. In the near future, I expect to see a full range of solutions for mobile virtual worlds – from full blown virtual world clients to simple presence status messaging.

Sun Microsystem’s Wonderland has the possibility to visualize an attendad who is calling to the meeting. The caller is shown as a sphere, and her voice is spatialized to come from the object. Avatars can take the sphere and move it to a suitable location – even for a private talk if that is needed.

realXtend allows avatars to define their skype addresses. You can initiate a skype call by right clicking the avatar, and selecting “skype call” from the pie menu that appears. You can even set up your real world phone number as your skype address. That allows you to start receiving skype calls to a mobile phone from a realxtend world. The world can be your virtual shop, where your avatar is waiting for customers.

Genkii, a japanese virtual worlds company, recently launched an Instant Messaging client for Second Life and Opensim. See more about Sparkle IM for iPhone.

Technically there are some special properties of mobile devices that differ them from desktops.

• First, and maybe the most important is battery life (With heavy CPU, network and 3D chip load it is easy to drain small battery fast)
• CPU is selected on the basis how good it is to conserve energy and may not perform well
• 3D graphics support is still coming, even it is already present on the newest models
• Network speed is low when compared to desktops, but may be already on megabit range
• Network use is expensive in terms of battery life, radio transmission power consumption is huge when compared to anything else
• Displays are small but good in quality

In the future we can expect to see virtual displays that allow people to project the display to a big screen on a wall or plug in to wearable head mounted displays. It should be also easy to make the mobile’s own display to show in real stereo, as it is small and watched from hand – which makes it perfect application for passive stereo image technologies.

With the external display present, the user input may be easy to arrange so that the device’s 3D accelerometers are used as the steering device.

How should a virtual world client on a mobile platform work?

To conserve energy – and maximize battery life, virtual world client should minimize network traffic and cpu intensive work. As the memory is cheap and well available, many things can be precalculated and cached into memory. Calculations should be moved to the server as much as possible. The networking solution should be client-server and not peer-to-peer. Peer to peer tends to cause more network load even it has many other benefits. See earlier Maxping article how this is contradictory to how things should be arranged with desktops.

Fistest.org is a community project to think and research through all the architectural and hardware needs of networked virtual worlds. This post is written as a part of that work. Follow me on Twitter to stay updated on virtual world technologies.

The core assumptions on which the internet infrastructure is built are fundamentally different from what 3D web requires ( Maxping definition of 3D Web). Thinkbalm suggests in their recent study, that year 2009 will be the watershed year in which we will see how virtual world applications begin to really emerge. Maxping wrote about how big companies are gathering around Opensim and related open source projects to better understand and even take part in the development of the new 3D web. 

Check out  the new study from ThinkBalm for further proof of the business value of virtual worlds.

Although the software platforms for virtual worlds are still not ready, we can, in addition to those needs we know of, foresee many of the requirements the 3D web will place for the end-to-end system. 

The current internet is an answer to old needs

During 1996 and 1997 Internet experienced very dramatic exponential growth in traffic and bandwidth consumption. That comparatively short period was an exception to other years and the Internet traffic has been growing approximately only 100% each year after that. The dogma “traffic doubles every three months” (originating in the dotcom bubble era) led to an investment leading to massive overcapacity to the core network.

What virtual worlds really miss network-wise is a low lag connection to the virtual world server and to the peer users. High lag causes big problems when it comes to showing a timely world response to the user actions. The lag should be kept below 50 ms for good results, but applying dead-reckoning and prediction algorithms the lag can grow up to 200 ms before users notice it – this depends greatly on the application and varies per use case, however.

The end user access to Internet is often via asymmetric services (ADSL), where bandwidth from Internet to the user is high – currently for example 8Mbit/second – while the bandwidth from the user to Internet is low – often less than 1Mbit/second. This asymmetry was built to serve web browsing, video and other media streaming – all services that stress the bandwidth from Internet to the user thousands time more than the upstream bandwidth.

The asymmetry becomes a problem with virtual worlds however, when users need to collaborate with each other. The shortest communication route is directly from peer-to-peer, stressing the uplink bandwidth much more than the traditional applications do. This also draws the attention to current firewalls and NAT systems which still need an efficient way to bypass them in a secure and reliable way.

The design principle for an enterprise database driven client-server application is different from virtual worlds. For thes eenterprises it is important to keep as much as possible of the application business logic, calculation and functionality at the server’s end. The clients are reduced to bare machines, preferably using only a browser user interface that communiates with the server. The application designers of the virtual future need to turn this upside down; while the business logic may still be at the server’s end, most of the computation happens at the client’s end.

Why? Because the user experience matters. Response time needs to be kept at absolute minimum.   

User experience drives the 3D Web

The basis of all succesfull products is an excellent user experience. Even as it is an undeniably important factor for traditional applications, it becomes even more crucial for virtual worlds. The basic assumption of a virtual world is that it is immersive, i.e. it creates a feeling of being there. The immersiveness can not achieved succesfully if the simulation quality and the manner in which the user experiences it, is not good enough. People’s brains are good in catching differences between virtual world and the real world, in a way that breaks the feeling of immersion. Examples of such things are lag in user interface response, poor or unsteady frame rate, rendering artifacts, breaking basic physical rules in an odd way – meaning that it is ok to fly as long as it looks like smooth and intentional. 

The end user equipment is often optimized for traditional office applications – word processing, spreadsheet and the like. The 3D web needs high end gaming computer performance, especially efficient rendering performance is critical. Often a neglected aspect is sound quality, it must be spatial 3D and of high quality – often headphones offer the best bang per buck – and the speakers can not be just the cheapest tin cans. CPU and memory requirements are easily met with a modern system, but still “more is better” for virtual worlds use as with almost any other modern software application. 

The smooth 3D web user experience needs natural-fit input/output devices, like Ball-it control ball, high quality web camera, and Emotiv headset. The rise of 3D monitors is also something that adds a great deal into immersiveness of a virtual world application. See how the vision for the user interface looks like from the Maxping article “The world builder“.

For mobile and small devices the story changes. What is the point of enabling a virtual world for a mobile phone? Possibility to join virtual meetings from a mobile device will be a requirement after the first successful pilot of a such virtual meetings at a corporation. The user interface needs to be different to a desktop, and the mobile use has different rules than what stated so far: For example the computation must be shifted to the server as much as possible, and peer-to-peer networking is not feasible in many cases.  

Summary – what needs to be done?

• End user needs more bandwidth to upstream direction to facilitate growing need for P2P communication.
• The computer equipment must be optimized for 3D graphics and sound to provide smooth user experience.
• The input and output devices must be re-thought to support more natural interaction.
• Application computations must happen at the client end as much as possible. 
• Internet core networks must be built to provide low-lag connections. 

This article just scratches the surface of  the ideas that need to be thought through to make this mayor changeto the whole infrastructure of the internet that is needed. Many new and old businesses will find ways to make things easier for the users.

This change is going to be a huge opportunity for everyone, once again.

First prototype of the 3D web is already run at thousands of Opensim servers all around the globe. The 3D Web bears similarities to 2D Web; Users can follow links to teleport from a 3D world to another one. 3D Viewers are used to browse the 3D content on the servers.

The 3D web software consists of Opensimulator server, comparable to Apache, and a 3D browser, comparable to Firefox. Today there are already many 3D browsers to choose from, the most advanced one called realXtend viewer.

The Opensimulator movement started early 2007 when Darren Guard published his C# reverse engineered Second Life compatible server. The Opensimulator (Opensim) project was born (read some history here).

Later many companies, most notably IBM and Intel (check Intel’s ScienceSim effort and their latest work with Opensim) have joined the Opensim project. Microsoft started to support Opensim indirectly via its C# user community. Even though there are big companies involved, the project backbone is built on hundreds of volunteer contributors.

Nokia joined the band by backing up the realXtend project. The realXtend project is mostly focusing on the 3D Viewer and user experience; It developed an enhanced 3D Viewer for Opensim based on Second Life viewer. Now the realXtend project is building a from-scratch 3D Viewer for the Opensim platform and Modrex module to support viewer enhancements at the server.

3D worlds have already, even in the current prototype form, spawned many business cases: IBM has demonstrated data-centre managing application and Virtual meetings with integrated Sametime product. Architects are using 3D web to show house plans to customers. Green Phosphor is visualizing data for medicine industry. Immersive Education initiative, collaboration of hundreds of universities, is evaluating and using virtual worlds for education.

Open source seems like the only way to implement the new 3D Web. No company or government wants to tie their applications to a single commercial closed source software provider. The big companies know this and they can not afford to ignore open source movement.

By the end of 2009 3D Web is developed into a stable and usable form that most probably allows mass adoption.