Hi Christophe, I had seen your announce on the caml-list. This is a really nice initiative, and I would like to help. The reason why I haven't moved yet is because my time is already very fragmented this year, and I wouldn't want to engage in something I can't do correctly. I guess I can delay that reflection until somebody applies for a project on liquidsoap?
I'll try to fill-in your wiki pages soon. I have a couple of project ideas, which I'll mention quickly for savonet users and developpers: 1. A first idea is to compile "pure" liquidsoap script functions for fast execution. This is a simple addition to the interpreter that detects "pure" functions (roughly, those that can be compiled down easily to ML or even C) and compiles them and loads them transparently to get fast execution. This is necessary make operators such as map() or map2() usable. 2. A more radical (but perhaps simpler) version of (1) is to directly compile liq scripts to ML, trying to get as close as "normal" OCaml code for fast execution. When this is not possible the result of the compilation just calls the embedded interpreter. 3. A graphical interface for composing scripts. 4. Various language extensions: (polymorphic) variants, records, objects. 5. A constraint-based type inference for our labelled calculus. 6. This is probably too prospective, but it could be interesting to investigate "hot reloading" heuristics. More later, -- David ------------------------------------------------------------------------------ The ultimate all-in-one performance toolkit: Intel(R) Parallel Studio XE: Pinpoint memory and threading errors before they happen. Find and fix more than 250 security defects in the development cycle. Locate bottlenecks in serial and parallel code that limit performance. http://p.sf.net/sfu/intel-dev2devfeb _______________________________________________ Savonet-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/savonet-users
