Skip is copyrighted.
How about a non-counting (mostly), "take-it-apart" method?
So I would look up "oku" (English??) under person-sound-heart.
I would prefer a method that involved as little counting as possible.
$B!!!z$8$e$&$$$C$A$c$s!z(B
$B!!;d$O$m$3$($s$i$+$Y$5!#(B
Riddle of the week:
What song is 35971040100?
That is not a catalog number.
Hint: the chorus is 3597104042
--- Original Message ---
$B:9=P?M(B: Tom Emerson <tree@basistech.com>;
$B08@h(B: James Kass <jameskass@worldnet.att.net>;
Cc: rscook@socrates.berkeley.edu;unicode@unicode.org;
$BF|;~(B: 01/07/09 11:03
$B7oL>(B: Re: Erratum in Unicode book
>James Kass writes:
>[snip]
>> An ideal index for the casual or non-CJK user might be quite
>> different in approach. Perhaps the first component drawn in
>> any character would be a good basis for indexing rather than
>> the significant radical. But, as you've pointed out, not all
>> components are immediately recognizable as pertaining to a
>> particular radical, especially to the casual user.
>
>If the goal is to generate an index for the non-CJK user then perhaps
>something similar to Jack Halpern's SKIP index used in his kanji
>dictionaries would be worth considering. For those not familiar with
>it, the basic idea is that characters are classified based on their
>primary structural pattern and the stroke counts in each component. It
>is easier to learn to count strokes than it is to recognize radicals,
>IMHO.
>
> -tree
>
>--
>Tom Emerson Basis Technology Corp.
>Sr. Sinostringologist http://www.basistech.com
> "Beware the lollipop of mediocrity: lick it once and you suck forever"
>
>
This archive was generated by hypermail 2.1.2 : Mon Jul 09 2001 - 08:20:06 EDT