Unicode-based FreeBSD

Alexander Churanov alexanderchuranov at gmail.com
Wed Nov 12 07:30:38 PST 2008


Vadim,

Yes, I'm sure that transition from 256 to 512 characters is possible.
However, my suggested implementation plan for this feature is:

1) Create working version of syscons having the ability of transcoding UTF-8
to 8-bit on output to screen and encoding keystrokes to UTF-8 on input. This
will utilize maximum of current screenmaps, fonts, etc. The UTF-8 mode is
optional and can be turned on and off.

2) Having UTF-8 processing done, modify the renderer to accept full range
(21-bit) unicode code points on input instead of 8-bit characters.

3) Modify the renderer to draw greatest possible range of code points.
Probably 512 or more. Please, take into account that, as I know, there are
more than one rendering module. Some utilize text-mode capabilities of
graphic cards, other use graphic modes and deal with pixels. I don't work on
it right now and explained my current view of rendering. So then, on x86 it
is possible to draw even complex scripts like devanagari. The drawback is
the cost of implementation.

Alexander Churanov


More information about the freebsd-current mailing list