On Thu, 24 Oct 2002, Bharathi S wrote:

> Today at 5:01pm, Jacky, Lau mailed to Me too :
>
> > Right, does your languague is realy some complex one??
>
>  Yes, Highly complex in both INPUT(KB) and OUTPUT(Display) side :(

  Input may not be as complex as you think although output
rendering certainly is. If  mechanical typewriters worked for
Indic scripts, it seems to me that Xkb approach should work as
well. <digression> Korean Hangul being as alphabetic as Indic scripts,
the same can be said about it. Koreans had used 3-set keyboard mechanical
typewriters for a few decades when the dictatorship in 1970's introduced
2-set KBD against advices of a number of experts including the inventor
of 3-set KBD  (he even had to leave the country for a while.) and made
things complex. While 3-set KBD can easily be translated into Xkb,
2-set KBD requires 'intelligence' of electronic typewriters and input
methods later.  Unfortunately, I'm one of those many people 'spoiled'
by the 'intelligence' of input methods. (for Hanja input,
we need input methods anyway)</digression>

 You appear to think that keyboard sequences have to
be directly mapped to  'glyph indices' of some custom-encoded fonts.
I could well be downright wrong and I'll stand corrected in
that case.

However, I found that at least for Hindi either Xmodmap or Xkb
works. See the following:

http://rohini.ncst.ernet.in/indix/doc/HOWTO/Devanagari-HOWTO-2.html

Wouldn't that also work for Tamil and other languages/scripts in
India? Actually, some people posted Xkb layouts for a few Indic
scripts to this list.

BTW, if you still think you need a complex input method, you may also want
to explore 'input module' of Gtk2. It's not as universal a solution as XIM
because it's limited to Gtk2 applications, but _might_ be easier to work on
than XIM.


>   Our Languages ( 15 Languages ) is based on single encoding(ISCII) and
> different encodings also(TSCII etc ). But during display depending upon
> the current language the encoded codes are rendered.

  IMHO, you should base your work on Unicode.  AFAIK, the encoding scheme
of Indic scripts in Unicode is almost identical to that of ISCII but with
distinct/disjoint code spaces allocated for each of 15 scripts/languages.
The use of ISCII, I believe, has to be limited to (internally) taking
advantage of non-Unicode-encoded  fonts for Indic scripts.

  Jungshik

_______________________________________________
I18n mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/i18n

Reply via email to