... at:

http://srfi.schemers.org/srfi-74/

Besides typo fixes, the only change I've made is to make (endianness
native) be one of (endianness little) and (endianness big), instead of
a separate object.  (Thanks for the hint, Sebastian!)

Generally, the discussion has raised many interesting issues---thanks
to those who contributed.

Several people have expressed their wish to have blobs be
bit-addressed rather than octet-addressed, and have bit length instead
of octet length.

This is probably a good thing to have, and I can well imagine having
bit addresses.  It is just not within the scope of the SRFI.

Changing blobs to have bit length, however, would change blobs to be
bit strings instead.  That's probably also a fine abstraction to have,
but, in someone else's words, "this SRFI is not that SRFI."  With this
SRFI, I only want to explore what's within the scope that's spelled
out by the title and the abstract.  It may very well be that a bit
string SRFI would subsume this SRFI, but I'm not convinced of that at
this point, and I'm not ready to do it myself.

--
Cheers =8-} Mike
Friede, Völkerverständigung und überhaupt blabla