The Gauss database10 is an example of work aiming at the dissemination of knowledge about software patents. We have close connections with the developers and maintainers of that database and we have the intention to make a connection to this database and our work. We will use the ToolBus coordination architecture [8] to connect analysis components with the Gauss database.
Another public effort that is badly needed is to set up a searchable archive of prior art for software. Here one finds the systematic investigation into the non-triviality and prior art violations of existing patents. We intend to develop a patent monitor which enables the public to systematically submit their opinions about existing patents. In addition various forms of text-mining and cluster analysis can be employed to unlock the knowledge in the patent databases.
A crucial element in any patenting regime is the role of prior art. We propose to investigate the possibilities for
Several proposals have already been made for revising the various regimes. These should be studied and compared in detail.
There is no reason to believe that one unique software patenting regime can be designed, assuming that one exists at all. Thus many different regimes should be investigated. For each regime a set of questions has to be settled: what constitutes prior art, what is an infringement, how to define the particular 'patent speak' and its semantics, definition of the appropriate life cycles, and so on.
An important step might be to develop a collection op hypothetical software patents, i.e., rewrites and perhaps simplifications of the software development history in which known developments are ordered in such a way that some steps can convincingly be patented. The historical development of computer software might even be simulated in a game-like fashion in order to study the impact that some patents (had they existed) might have had.