* Port boinkmarks to ECL or create some other way to test performance of
ECL/git and produce nice records of that.
* Profile and optimize ECL' C library. Key places include for instance the
interpreter, CLOS dispatch routines, uses of ecl_process_env(), etc.
* Work on the FFI creating optimizers for compiled code. This basically
amounts to creating compiler macros that, in case of large optimization
settings, expand FFI forms into optimized C-INLINE forms. This would
dramatically improve access to foreign data when the types are specified.
* Work on a better API for synchronization of multithreaded programs. The
Boehm-Weiser garbage collector already provides a nice library for atomic
operations which could be used to implement compare-and-swap or other cheak
* Finish the MOP.
* Optimize our implementation of CLOS. This includes things like rethinking
the way we optimize access to object slots, writing specialized dispatch
functions for methods that depend only one argument, etc.
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)