----------------------- InBCT 2.1 Plan for 2004 ----------------------- In 2004, we plan to go further on the most important accomplishments of the year 2003, which included - FenPDF is in actual real-life use by our research group for managing the academic literature related to our research. - We have developed the LEGO + optomechanical mouse custom controllers paradigm for easy and cheap construction of custom computer controllers - We have developed functional programming -related methods to speed up view generation: if a certain part of a view only used RDF nodes that have not changed between frames, that part can be reused automatically since the functional programming paradigm gives guarantees about the result not depending on anything else. - We have developed methods for enhancing the readability of freely rotatable/scalable/deformable text rendered using graphics accelerators - The first Storm blocks have gone through a P2P network between machines - The completion of Navidoc, a UML - Javadoc documentation unification tool for HTML. - We have developed the Fenfire Loom, a focus+context RDF browser - The first published article about unique background textures - unique background textures are in use in FenPDF and have shown themselves to be *extremely* useful there, demonstrating that we were asking the right questions very far away from the mainstream. In 2004, we plan to (more information about any of these topics can be requested by emailing Tuomas Lukka) - Make our work accessible for the financiers by - documenting our architectures and code in detail, to make it easier for people from outside our project to "get into" the code (Navidoc is essential for this) - creating internal technical reports about subjects that the financiers express specific interest in - publishing scientific articles about several parts of our work that have matured enough, for example: - FenPDF, the literature comprehension tool, the first real user-level product of this project - Alph, the xanalogical referential fluid media implementation which includes important innovations related to realistic implementations of RFM - Libvob, the UI development system we use - The Functional programming view approach - Irregular viewport edges - Start testing shrinking of the FenPDF user interface to mobile-sized screens (specifically requested by Nokia in beginning of 2003, causing us to create the framework for custom controllers) - Further develop FenPDF-like interfaces to be applicable to e.g. photographs, sounds, SMS messages (another request from Nokia). - Develop Loom to combine with FenPDF and other applitudes to get the full hyperstructure functionality - Develop UI mapping techniques: now that we have a *real* FenPDF bidirectionally linked structure, there are several ideas in user interfaces that we can test, such as creating a multi-focus view which shows the routes between the foci in the graph. This is one of the core areas of new innovations that our approach enables, since we're doing things differently from the mainstream: on a Web/Filesystem -based system, such views would not make nearly as much sense. - Research the uniqueness of the unique background textures further. The first article used a rather ad hoc distribution for the textures, we have ideas about how to make the uniqueness more well-founded by using user experiments. - Develop libvob vobscene recursion, enabling even more interesting functional programming -related techniques in views. This technique is useful for retaining fast interactions even with complex views without too much resource consumption. - Storm-based transparent collaboration: two people meeting in a coffee shop and working together on the same document, both with their own computers, transparently synchronizing through a *local* network, with no servers involved, later synching the results to the whole working group. Requires: - Further development of Storm - RDF vocabulary-based modular change merging - UI techniques for understanding changes to documents in FenPDF