Orhography (and phonology) dilemmas

Right now, the orthography of tAruven is somewhere between
being phonemic and phonetic, the distinctions are (probably)
finer than they need be. For instance, I haven't decided yet
if there are two phonemic r's possible in the onset, or just
allophones... breathy(aspirated) trilled r <'r> like in
<'riaN> "(telepathic) voice/communication" and 'ordinary'
trilled r <r> like in <rI> "give". I guess I need more words
starting with (')r to decide... They sure weren't -meant- to
be two different phonemes :)

Furthermore, I'm thinking of changing how I do vowels,
glides and geminates yet again... right now, 'legal',
(possibly) syllabic gemination is marked by uppercase: <aRO>
"dog", and other gemination (due to derivation, and all
geminate stops) by doubling: <cigga> "foreigner, foe". I
used to disallow geminate stops alltogether, but they kept
showing up, and since I'm discovering this lang, not
horsewhipping it to comply to rules set in stone, I've had
to allow it :)

Point is, I don't like the uppercase-for-gemination idea as
much as I used to... but if I were to double them, the
geminate vowels would conflict with the so-called 'double
vowels', like the passive-marker -aal in <suMaal> "to be
judged, sentenced".=20

It won't interfere as much with the glides (tAruven-style
polyphthongs) as it used to, as I now explicitly mark them:
  1. <fi=E0x> (f-i-agrave-x) "to board",=20
  2. <ciar> "black",
  3. <i=E1luin> (i-aacute-l-u-i-n) "to whisper, hum"
marking 1) coming from standard-length, full i towards short
a, 2) both i and a full and standard-length and 3) short i
going towards standard-length, full a respectively.

The double vowels are of two qualities as well, Vv
(stressed-unstressed) vs. vV (unstressed-stressed). I could
of course mark them up as I do the glides, <suMaal> would
then be <suMa=E1l>. Using doubling instead of uppercase for
gemination could then produce things like <kir,aa=E1l>
(k-i-r-,-a + a-aacute-l) "was cut" and <a=ECraaa>
(a-igrave-r-a-a + a) "the center/middle of the rabbit". =20
Yuck :) Not exactly readable...

Pronounciation-info at
http://www.nuts.edu/~taliesin/taruven/sounds.html
Vocab at http://rhiann.nvg.ntnu.no/cgi-bin/search.py


- said a frustrated
taliesin,=20
eagerly awaiting the miracle of unicode and the death of
7-bit charsets