31-7-2009

Short term roadmap

(the coming two months or so)

- Marketing! An open source project is heavy to do on your own. Blender, Python and programming forums should do nicely.

- Get the techn

- Document! Here is a stub on all the items that need to be documented. ome of the things are covered in the reference document and in the tutorials, but not everything. The reference document should be updated too.

Middle term roadmap

( the coming 6 months or so)

- The Blender part of the 3D libraries needs considerably more change to be clean. No more ugly spacehandler copying! However, better to wait for Blender 2.5.

- Spydermaster offers no means of editing system variables, or browsing them at install time, and its new/outdated lists are damn ugly. This needs to be improved

- I started once on a grid editor/game maker for Blender written in Spyder. Spyder, and probably Blender too, have considerably changed. It is now possible to do these things much cleaner and more OO style.

- A decision should be made how non-Spyder class function objects (and data objects) are handled through the *export directive, especially in relation with documentation

- There should be a Resource class that fetches things from repositories unless they are cached on your hard disk

Long term roadmap

(road map to world domination, the coming decade or so)

- Spyder needs a broader scope, and more cross-talk between its scientific, 3D and webserver sections. There should be a Tarantula back-end not only for Blender, but also for PyMol. The Blender GUI should be ported to PyMol as well. Even better would be a general GUI app that can communicate with many programs.

- My boss will fund me for another half-year or so, till my girlfriend graduates. It would be nice to leverage Spyder into a general visual flowchart programming tool, a la Taverna. This will require a major extension of the conversion engine to deal with all kind of flowcharts, a nice gui, and web integration too (automatic creation and detection of Spyder web services).

- It would be nice if Spyder web files would be an alternative for SOAP/XML/JSON in the web world... may be wishful thinking though

- Spyder should really mix well with efforts to compile Python into something faster (Pyrex, RPython, ShedSkin or whatever). Just define a switch to replace he existing core module scripts with something that spits out these languages. Spyder is a good interface between those languages and pure Python, too.

- Perhaps the conversion engine could be rewritten in Haskell.

- Spyder should be good for games. Really really good. It provides a way to integrate and extend objects in a game context, makes multiple and transparent back ends, blurring of data and code, etc. There should be a Spyder based network protocol application or something, into which objects can be plugged in.