NOTES

======================================
LIST OF IDEAS AND TODO ITEMS FOR BABEL

  * VERY IMPORTANT: given that we steal a lot of code from OpenMCL we
    probably to license those files under the LLGPL (or is it the
    whole library?)  Figure this out ASAP.

  * Many more tests needed.
      - test invalid sequences.
      - use more interesting test files.
      - test with big strings that when converted would need a array
        with more than array-index-limit elements.

  * Add support for ennforcing endiannes for multi-byte encodings like
    UTF-16 and UTF-32.

  * Implement more encodings. (ISO-8859-*, windows, CJK encodings, etc)

  * null-termination, bom-markers.  Add to string-to-octets.

  * Unicode normalization routines?

  * Allow for non-fixnum indexes?

  * Documentation.

  * Address the various XXX and TODO items in the source.

  * Should we start adding a lot of encodings, a load-on-demand scheme
    might come in handy. It would also help with debugging.

  * Make the test suite reusable, very much like mappings are reusable.

  * Make the whole thing easier to debug.