Friday, August 23
09:30 - The Semantic Web: Opportunities and Problems
At the moment, an international interdisciplinary initiative, lead by the W3C, develops the second generation of the web, the Semantic Web. Its major advance is to provide uniform description languages for the static semantics of documents (ontology languages). With their help, concepts such as better web search, uniform document constraint processing, and composition of web services should be realized, and this talk illustrates these scenarios.
However, the Semantic Web is no "silver bullet" by which all problems of mankind can be solved; it incorporates many technical weaknesses and limits. Additionally, it suffers from many stakeholder problems, such as conflicting interests of parts of the society. And it produces many new dangers for the privacy of the individual human being. Basic human rights, such as the right of "information self-determination" should be introduced on a global scale, to ensure that the Semantic Web is beneficial to us all.
10:00 - Tea/coffee break
Have some tea, coffee, cookies and a chat in the cafeteria. All parts of course strictly optional. :)
10:50 - The sTeam project: Information Structuring
Document based work, even with support for synchronous work still needs possibilities to express the inter document structure of the field one is working in a persistent way. The world wide web consortium and its Semantic Web initiative provides innovative concepts to meet these demands. Annotations (covered by the Annotea standard) are out of document comments linked to a document. Topic Maps is an XML based technology to express relations between documents in a consistent way. Together, they provide the technological basis for new applications to support teams when structuring the information needed for their task.
11:30 - Active documents on the web
Web documents are intrinsically woven of data and software, they are "active documents". This is why software engineering should play a much larger role for web engineering in the future. This talk overviews several approaches to component based software engineering in the last 30 years, and presents a technology that unifies several of them for the use in active documents. "Invasive composition" works uniformly for software and data, permits defining component models for both worlds, and could easily be supported by Pike.
12:00 - Break
Lunch hour; we suggest you visit one of the restaurants near the campus with the rest of us.
13:15 - The sTeam project: Developing Application Servers in Pike
The people at the University of Paderborn are currently developing a client / server system running behind a Roxen or Caudium webserver. The core server is a largely generic application server, providing an inheriting ACL based security system, object persistency and event handling capabilities. Webservice specific modules like XSLT capabilities are added to demonstrate how application specific demands can be met by plugging into the generic system. Last but not least, an upgrade mechanism allows for rapid development and administration, without needing to restart the server and thus without long downtimes.
13:45 - Web Ontologies and Pike
During the past few months, Martin Nilsson and Leif Stensson have been looking at DAML, OIL and RDF for the purposes of making Pike the first scripting language for the Semantic Web. This has already resulted in some extensions to Pike in the form of a data type for relations and some XML namespaces support. In this brainstorming session, Leif will introduce some of the design and practical issues of the use and purpose of ontologies, in the context of Pike.
14:15 - Pike and Java
There is no Pike applet in the web browser. Building synchronous web applications like chat tools or shared whiteboards is possible only if you have a platform that is avaliable from every browser. The only answer to this dream is Suns java-applet technology or -- if you can live with the installation procedure -- java-webstart. Consequently, the client part of such a synchronous web based system has to be done in java. When interfacing to a Pike server, Java built in technology like RMI or the Shared Data Toolkit can not be used. An easy to use Java API has been developed to allow easy access to the open sTeam platform. This API is an abstraction layer to the same protocol that is used to interface between Roxen Webserver and the open sTeam platform.
15:00 - Tea/coffee break
Time for another round of tea, coffee, cookies and general friendly chatter with the rest of the attendants in the cafeteria.
15:30 - The Caudium project: background and future
Martin Bähr will hold a presentation and brainstorming session about the Caudium project, a Roxen 1.3 fork in active development.
16:30 - Additional presentation, if any
Room for unscheduled events. This slot intentionally left blank.