Wheels in my Head

[Brandon Wirick <brandon@yrick.com>]

I really am working on the program to visualize 7SD characters in a
Lojban context; don't anyone think I've put that on the back-burner. I
have, however, come up with yet another fun and mostly useless idea. I
would like to devise a ray to represent every conceivable Lojban
syllable in exactly two bytes apiece (I'm counting a consonant
followed by a {denpa bu} as a syllable). The way I've broken it up in
my mind is into four nybbles (four-bit numbers for those non-computer
people).

Nybble #1 would include the voiced and unvoiced consonant pairs (not
{ly}, {my}, {ny}, or {ry}, and I don't know about {xy}) and a null
sound.

Nybble #0 would precede Nybble #1 with any consonant or a null sound
(maybe a {slaka bu}, but Nybble #1 would set the voice, so Nybble #0
in both {tca} and {dja} would be the same (a sort of {ty}-{dy} pair)
but Nybble #1 would differ.

Nybble #2 could include the consonants {ly}, {my}, {ny}, or {ry}; the
vowels {ibu} or {ubu}; any vowel followed by {y'y}; a {denpa bu}; or a
null sound

Nybble #3 could include any vowel, lowercase or capitalized, or one of
the diphthongs {ai}, {au}, {ei}, or {oi}.

Some Nybble combinations would not be allowed, like those that result
in double letters, but the point is to cover the set of Lojban
syllables. This system is not perfect, so I'm open to suggestions, but
it probably won't serve any practical purpose, so feel free to ignore
this if you don't care.

mu'omi'e.uirik.