L2/18-181

[Richard Wordingham via Unicode]

On Wed, 16 May 2018 13:46:22 -0700
Doug Ewell via Unicode <unicode at unicode.org> wrote:

> http://www.unicode.org/L2/L2018/18181-n4947-assamese.pdf
> 
> This is a fascinating proposal to disunify the Assamese script from
> Bengali on the following bases:

> 3. Keyboard design is more difficult because consonants like ক্ষ
> are encoded as conjunct forms instead of atomic characters.

Users of X do have a valid gripe here.  An X keyboard mapping can only
accept single codepoints; sequences require explicit support by the
application.  Advanced applications get round this by using an input
method, but they can be unreliable, particularly over networks.  (I
ended up creating an X keyboad mapping as back-up, but when I use it
I lose all my 'ligature' keys.) However, that seems to be an argument
for deprecating Bengali, rather than for disunifying Bengali and
Assamese.

I think simple Windows keyboards have a limit of 4 16-bit code units;
for an Indic SMP script, one couldn't map <x> to a single key, as it
would require 6 code units.

It would be handy to have characters whose only use was to input text;
adding characters that are subject to composition exclusions would not
change whether text is in NFC, in NFD, or neither. 

Of course, if the scripts were disunified, would we have to ban Assamese
domain names in the new 'Assamese script' as they would be ambiguous
with Bengali names.

Richard.