Re: [tlug] Japanese page has been revised

["Stephen J. Turnbull" <stephen@example.com>]

>>>>> "Scott" == Scott Robbins <scottro@example.com> writes:

    >> Anyway, I don't think you said exactly what was makes anthy
    >> preferable over canna.

    Scott> Because Stephen said so.  :) Actually, it's subjective, and
    Scott> seems to be the preferred solution these days.

Right; that's precisely what I said.  :-)

Personally, I've never seen a purple cow ... ahem.  Personally, I've
never used Anthy, and haven't used Wnn or ATOK in 10 years.
XEmacs<->Canna satisfies almost all my needs, and the rare cases where
I'm desperate for nihongo in a browser I use Safari on the Mac or
Firefox<->kinput2<->Canna.  I mean, how much Japanese can you type in
a Google search box?  :-)

    >> Before I follow the instructions on your page and remove my
    >> working canna in favour of possibly improved anthy, I was
    >> wondering if you'd be willing to say a little something about
    >> what makes you like one over the other. On the web page you say
    >> anthy is more popular, but is that its only advantage?

Canna is a mature project; that means its APIs have fossilized.  There
has only been two major releases since I started using it in 1992.  It
also has its own built-in "front end processor" (the pre-editor that
edits the kana and sends commands to the dictionary engine).  You can
rebind the commands, but it's awkward for an expert and novices just
don't even try.  Most of the development recently has gone into
improving the dictionary (and that even forked).

Anthy is a backend dictionary engine only, although it's closely
related to the uim front end.  So is probably more pleasant to program
for.  That gives it an edge in developer enthusiasm.  I bet it also
integrates better with GTK than Canna does.  Woo woo!

The web-based Anthy documentation says nothing about the algorithms
used for learning and extracting morpho-syntactic information.  So I
would guess there's not much new there.

Finally, the two big selling points on the Wiki are that it's free
(who thinks that makes it a better conversion engine, besides RMS? and
anyway both Canna and FreeWnn are free) and secure.  "Secure"
apparently simply means that (1) it doesn't do networks so the
conversion protocol can't be sniffed to find out what you're typing,
and (2) if you make your private dictionary readable only by yourself,
nobody can read your private dictionary.

My feeling is it's a typical Japanese fork just to be idiosyncratic,
but it looks like it's going somewhere so it's popular.  From the
user's point of view, just take whatever you've got at hand.  :-)

N.B. all of the above is based on reading the Anthy Wiki; I guess I'll
have to try it one of these days, but I haven't done so yet.

-- 
School of Systems and Information Engineering http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
               Ask not how you can "do" free software business;
              ask what your business can "do for" free software.