Re: tlug: Kinput2 and Wnn6

[Frank Bennett <bennett@example.com>]

"Stephen J. Turnbull" <turnbull@example.com> writes:

>     Frank> ... But why would Emacs and Kinput2
>     Frank> behave differently (in both instances Emacs and Applixware
>     Frank> respectively were running solo, by themselves)?
> 
> The interfaces are different, so they'll behave differently.

Sorry, I should have stated the point more explicitly: why should
the _server_ behave differently when it is accessed via different
clients?  On the server side there should _not_ be a difference.
I've solved the user interface problem on the client side ---
that was an earlier thread.

> One possibility is that kinput2 is a server, and therefore running as
> a specific user, especially if it's in xinitrc or similar.

It is an X client that waits around for connections from other X
clients connected to the same display.  It is owned by the user
ID in effect when the .xinitrc .xsession script decorates the
user's window with initial processes (including the lurking
Kinput2).

> I don't
> know how kinput2 passes private dictionary information to Wnn.

Neither do I.  The private dictionaries, though, are owned by
Wnn6, not the user.  They are created and maintained
automagically in the Wnn6 filespace.

> But if
> it just passes its effective ID, then wnn thinks that kinput2 is user
> kinput, while Emacs is almost certainly running as your personal user.
> This would result in the dictionaries getting decoupled.  Take a look
> at ps aux and see what userid kinput2 is using.

Nope.  As indicated above, Kinput2 is owned by the user who is
running the display.  But a failure of this sort seems the most
likely cause of the problem --- Wnn6 ends up adding the Kinput2
data in a "dunno" stack, or into the root dictionary.

Here's the result of test:

[bennett@example.com bennett]$ ps aux | grep kinput2
bennett  14572  0.0  4.2  3068  1300  ?  S   11:24   0:00 kinput2 -wnn -jserver

>     Frank> The server certainly _claims_ to be willing to speak with
>     Frank> anyone, without conditions:
> 
> I don't trust such messages to mean what I think they mean.

That's why I put emphasis on _claims_.

>     Frank> Is this due to some limitation in Kinput2 --- maybe the
>     Frank> patch for Wnn6 is missing something?
> 
> Could be.  Gotta love proprietary softo.

That's very helpful, thanks :)

Cheers,
-- 
-x80
Frank G Bennett, Jr         @@
Faculty of Law, Nagoya Univ () email: bennett@example.com
Tel: +81[(0)52]789-2239     () WWW:   http://rumple.soas.ac.uk/~bennett/
---------------------------------------------------------------
Next Meeting: 10 October, 12:30 Tokyo Station Yaesu central gate
Featuring the IMASY Eng. Team on "IPv6 - The Next Generation IP"
Next Nomikai: 20 November, 19:30  Tengu TokyoEkiMae 03-3275-3691
---------------------------------------------------------------
Sponsor: PHT, makers of TurboLinux http://www.pht.co.jp