Re: [Mlterm-dev-en] problems with accents

[MINAMI H. <mi...@mi...>]

On Sat, 28 Jun 2003 16:10:53 +0900 (JST)
Tomohiro KUBOTA <ku...@de...> wrote:

> Sorry I don't understand the reason why XMODIFIERS is important in this
> case.  Thus, I cannot write any rational explanation.  Can you explain?
> (Japanese OK).

AFAIK,

- KeyPress events should be processed using either XmbLookupString()
   or XmbLookupString().

 - XLookupString() cannot handle a context of input
   (at least in a portable manner). 

 - XmbLookupString() can do compose but needs an input context to work.

- To create a XIC(X input context) structure, XIM should be opened.

When XMODIFIERS is not set, mlterm do not activate XIM for now.
In that case, only XlookupString() can be used and composing will not work.

> However, if there were any methods European people don't have to think
> about XMODIFIERS, it would be better.  I imagine most European people
> don't know at all about XMODIFIERS.  I think they should not need to
> know about it.

Though it is possible to always activate XIM by assuming "XMODIFIERS=none",
activating XIM will increase memory consumption and make mlterm slower. 
So I'm not sure whether it's a better default behavior. 

--
MINAMI Hirokazu <mi...@mi...>