On Thu, 9 Dec 1999, Henry Nelson wrote:
> > Henry, we say hanKaku not hanGaku.
>
> Right. Amazed you could read it (only because my sendmail has qkc as a
> frontend). BTW, were you able to read the final "na" of "katakana?"
I had to use nkf with a tricky way.
% nkf -e your_mail | nkf -S | less
I couldn't find the last "na", which seemed to be dropped.
# I'm sorry this is not concerned with Lynx.
> > Anyway you mean Shift_JIS is wrong because it has code for half width
> > katakana? I guess you think EUC-JP is also wrong by same reason.
>
> No. The ambiguous *WAY* Shift_JIS encodes it.
I see your thought. Thanks.
> Anyway, since you can
> program in c, make Lynx a better guesser. If you really think Shift_JIS
> is worth something, help Klaus fix the search code, or assure him it is
> working correctly for all code sets. Thank you.
I would do so if I had enough time and skills...
Anyway I'm going to test the search code this weekend.
--
Takeshi Hataguchi
E-mail: [EMAIL PROTECTED]
