The goal of the software architectures of the WebTalk series is to allow different forms of cooperation between and among various visitors which are accessing the same website. While conventional web surfing is exploited in a solitary fashion for every single visitor, the ever increasing complexity of the information presented in a web site may benefit from special means which support the user in finding, understanding, and consuming information. We think that this result can be obtained by devising a cooperative mechanism, that is allowing each user to detect the presence of other users in the same web site, with the possibility to interact with them. It is not hard to envision a scenario in which users playing special roles, such as tour guides, virtual clerks or virtual consultants, assist website visitors in retrieving and understanding information from a site with a high level of complexity, due to its structure or to the very nature of the information provided. This holds not only for communication between 'privileged' users and end-users, but also between end-users themselves, thus enhancing socialization and distance learning. Adding to web navigation the support for collaboration can be done in different ways: 2D Interfaces Systems, which make use of traditional bi-dimensional GUI elements, and 3D systems, which exploit 3D representations in order to offer new functionalities. The WebTalk Team strives to experiment cooperation between users, supporting it using both approaches - 2D and 3D - , and by developing in-house middleware components, as well by studying and integrating third party toolkits and products. While the initial steps of the team have been in mailnly developing lightweight middleware to support 3D systems, the activities are broadening to comprise, alongside the development of a completely new middleware infrastructure, the design of cooperation patterns methodologies and usability tests of current cooperative-2D components. All the activities lead to one and only final achievement - discovering the different ways to cooperate in consuming web resources.

WebTalk-I Platform   WebTalk-II Platform   Going Wireless
The first solution deployed by the team was the WebTalk-I platform. It consists of a client-server architecture, entirely written in the Java language, to support the creation of Internetworked Virtual Environments. These virtual environments can represent any 3D ambient and allow web users to navigate in it. Each user is represented by a human-shaped figurine named avatar: users can thus see other surfers in the same space, and share with them space, actions and activities, while exchanging opinions about what they are seeing and learning by means of a chat application. The goal of WebTalk-I is providing such functionalities using only standard technologies: for this reason, the 3D representations are implemented using the VRML (Virtual Reality Modeling Language) standard, while the server and client middleware is entirely written in Java. The client system can be easily loaded from inside a browser window, after the proper plugin (CosmoPlayer 2.1, has been downloaded. The advantage of this approach is that VRML geometries are common and easy to build. Once the geometries are ready, they can be quickly plugged in, in compliance to the WebTalk-I requirements. The server can run together with any web server, and on any Operating System and Hardware Platform combination. The ability to jump from the virtual space to the regular web space made of HTML pages makes this system an innovative shell to conventional 2D information. The cooperative virtual worlds become thus an additional layer, which supports retrieval and understanding of complex informations, thanks to the possibility of close interaction with other users within the virtual environment. An example of such an application is a virtual museum. A real case study has been deployed by the team, in a partnership with the Italian National Science Museum "Leonardo da Vinci". A virtual exhibition of Leonardo's machines has been setup to allow visitors to learn the way the machines worked and their real scale. The system is up and running at           The team is currently working at the development of the WebTalk-II platform. The new system aims to gain in complexity and scalability, by allowing authors to design general-purpose access structures of their websites, collect the 2D resources (text, images) and the 3D elements (geometries, sounds, spatial links) in databases. It will then be possible to generate automatically any virtual environment needed, modeling it on the desired access structures. The system is designed to support loads of hundreds of concurrent users, and it is capable to load geometries in all the currently available formats, such as VRML and 3DMax. Instead of requiring a plug-in, it uses a Java3D graphical engine which will self-install from a user-friendly installer, avoiding the end user the hassle to go through special plug-in configurations. At the moment it is not certain which will be the distribution format of the platform. For users with a Java Virtual Machine already installed properly, the downloading size of the system will not probably exceed 1 MB. As part of the project, a theoretical foundation will be given to the issue of Cooperation Metaphors. This research will try to determine and formalize which are the base rule for cooperation between users in a cyberspace environment, and how these base rules can be composed in order to form behavioral patterns and cooperative interactions.          The WebTalk Team is working not only in 3D, and not only on conventional desktop technologies, but it is performing a full-scope researching on cooperative consumption of distributed information. A plan is envisioned to integrate wireless devices in our architecture: the idea is mainly to devise, both theoretically and by proposing architectural platforms and prototype implementations, a way to integrate hand-held PDA's and next generation (G2.5 and G3) cellphones in the cooperative system. This will allow mobile users to access web information in a cooperative fashion, as it is already provided for in desktop environments. The challenge will of course have several aspects, one for all how to represent effectively cooperation within the limited possibilities of today's hand-held visors.