Hello,

I posted a comment on your blog, but it got chopped off at the first
Shavian character.  Here's what I wrote (and hopefully it works
here...):

Inspired by your GNOME Shavian post I've learned to read Shavian.
I've been developing some scripts and building up a pronunciation
database in my accent.

I just thought I'd point out a few bugs in your system (I hope the
characters come through here...).
- You've got "Shavian" spelled wrong!  It should be "·𐑖𐑱𐑝𐑾𐑯"
(with 𐑾 "Ian" not 𐑳 "up").
- The naming dot "·", which indicates a word that would usually be
spelled with a capital, is missing.  I have them in my database for
proper names, and add them automatically to other capitalised words
which don't seem to be at the start of sentences (though this still
gets some wrong).
- Perhaps it's a problem with the database you're using, but it seems
that (in the Firefox extension screenshot at least) the places your
program puts 𐑳 "up" should really have 𐑩 "ado".
- Heteronyms are often incorrect (e.g. read 𐑮𐑰𐑛/𐑮𐑧𐑛, minute
𐑥𐑦𐑯𐑩𐑑/𐑥𐑲𐑯𐑿𐑑, close 𐑒𐑤𐑴𐑕/𐑒𐑤𐑴𐑟).  This is a tricky
one.   I've opted to leave them in Roman letters, but it should be
possible to use grammatical analysis to disambiguate some of them.

Michael