[SCIM] Re: scim question

[Ming Hua]

On Sun, Sep 26, 2004 at 12:44:05AM +0200, Osamu Aoki wrote:
> Hi,
> 
> I uploaded scim :-)
> 
> I now have question.

Good questions.  I'll try to answer them, but since I am not really
familiar with scim internals, I'll forward these questions to main
stream mailing list, so that the developers there can give better
answers.  (To scim mailing list -- this is about scim 0.9.7, I am
working to package 1.0.x.)

> 1) What happens if scim is started multiple times?

I think if you use the same user to start scim multiple times with the
same command line (i.e., same server, config and frontend modules),
nothing bad will happen.  scim is clever enough to know it's already
started and do nothing the second time.

> 2) If user foo start scim, can user bar use scim?

Not unless you do something special.  scim user a per user daemon by
default, and I think other users can't access the data.  However if you
set up the socket server, you can somehow use the same server.  This is
how it should work over network.  Only one machine (user) start scim
with real im method servers with socket frontend, and other users can
use the socket server with their own frontend (x11 in ordinary case) and
access all the imput methods provide by the socket.  However, that means
other users still need to run scim command, just with different options.

> 3) If we have 2 X sessions on one PC, what happens?

I don't know, and I never tried.  I doubt the XIM thing will work, but
the GTK im module probably can deal with that.

> 4) What happens user logout and log in again to run startup script
>    again?

I saw that if I use ~/.gnomerc to start scim, it will be killed when I
log out from GNOME.  So the next time I login, a new scim is started,
and everything is fine.  I remember scim 0.9.7 can't recover gracefully
when there is no X session, so you can't run scim before X is started.
I don't know anyway to keep scim running in the background when I log
out (nohup, perhaps?), and I doubt it will work.

> Osamu

Ming
2004.09.26

--------------------------------------------------------------------------
Physics is much too hard for physicists.                 --- David Hilbert
--------------------------------------------------------------------------