VRML (Modelling Virtual Reality Modelling Language) provides a flexible, multi-platform environment to model urban form and move towards the construction of virtual cities. VRML files are specifically designed for distribution on the WWW, allowing a block model of a large urban environment to occupy less than 100K. However, to achieve realism in the construction of virtual urban form, VRML is highly texture dependent. The higher the number of textures used within the VRML model, the higher the degree of realism and therefore the overall impression gained by the avizen. However, due to current bandwidth constraints is it not practicable to constuct heavily texture dependent models.

At present VRML 2.0 is replacing VRML 1.0 as the standard for 3D Modelling of the built environment. VRML 2.0 allows the developer to increase the level of achievable realism by allowing behaviors to be assigned to objects. Behaviors allow a scene to include features such as movement and directional sound, significantly increasing the level of realism. However, as with VRML 1.0, VRML 2.0 is texture dependent and features such as directional sound and movement increase the overall file size and therefore download time.

An example of a 'typical' virtual city on the web is pictured right, Virtual Tokyo by Plannet 9 Studios. The use of textures is extensive to enable a medium level of realism. While this presents an actuate representation of Tokyo, it does not represent a Virtual City. Virtual Tokyo is characterrstic of many Virtual cities on the web in that is is merely a three dimensional representation of urban form, it does not have any social or interactive elements.

Social and basic levels of user interaction have been built into the next level of VRML browser by Blaxxun Interactive. Blaxxun Interactive using their Passport Browser (pictured left) were the first company on the World Wide Web to offer VRML based multi user interaction. Blaxxun Passport integrates a JAVA based version of Internet Relay Chat (IRC) with a VRML compliant browser. Avatars are dynamically placed in the virtual world allowing any already developed Virtual world, such as the Plannet 9 example of Tokyo, to be adapted to include avizens. The avatars are dynamically placed into a VRML world by a server, which is hosted on the authors system. One of the easiest to use multi user browser/server is Sony's Community Place. Community Place is similar to Blaxxuns Passport, in that it allows the use of avatars in VRML worlds, but it also allows the use of real-time speech, adding another level of realism to multi user worlds. Community Place is also easy to set up, allowing easy development of complex multi user environments. However, there is currently no one standard of VRML multi-user environments. Worlds developed using Blaxxun Passport will not be viewable in Sony's Community Place and visa versa. Discussions are taking place about developing a standardised system, however until one is developed, VRML based multi user worlds will be dependent on users running specific software packages.