I've got theoretical and applied interests in the idea of semantic operating systems and what these really are and how their value differs from that of existing forms of OS. I think that's the theoretical bit.

The second part is more about (and this is pretty bad of me as a CS lecturer!) how to build an OS of this kind (or of any kind for that matter). I'm not sure if using open source for existing OS is the best starting point, and essentially tweak that, or if it is better to start bottom up, using open source code as the basis (along with other sources) for thinking up from the very beginning, i.e. as a greenfield development.

Third part is simply the role of and appropriateness of specific languages in this. While Erlang (because I know it) strikes me as the place to begin (for me), I am thinking (perhaps erroneously) that a declarative language (as in, like ProLog) would be most appropriate for this kind of project, for someone producing a low fidelity model to start with (not really talking about actually using ProLog, just the notion of that potentially being more intuitively analogous to this work - for a cod CS lecturer like myself anyway).

Anyone who has any experience in this (or even simply with philosophy of OWL, RDF and the semantic web) would be a great help in this, I am sure. I'd welcome your suggestions, guidance and how, not only, I can get started, but also in helping me clarify what I'm actually talking about in terms of definitions and concepts. Not having done anything at this level or of this nature beyond early undergraduate level, I really am - I admit - a bit lost on this. Clearing up my thinking would take me light years from where I am, and hopefully to the 'start line'!

Similar questions and discussions