[immodule-qt] Current Qt 4 status

[Lars Knoll]

On Tuesday 23 November 2004 04:01, YamaKen wrote:
> At Wed, 17 Nov 2004 16:07:51 +0900,
>
> kaminmat at cc.rim.or.jp wrote:
> > Thank you for explanation detailed enough for judging what is necessary.
> > But, I have one question.
> >
> > YamaKen <yamaken at bp.iij4u.or.jp> wrote:
> > > > I have removed the setMicroFocusHint() method and instead rely on the
> > > > input method to query the information it needs using
> > > > inputMethodQuery() as this makes it a whole lot easier to implement
> > > > IM enabled widgets.
> > >
> > > It will cause a problem in some cases.
> > >
> > > For example, some input methods may want to show its auxiliary
> > > widget adjacent to cursor always regardless of existence of
> > > preedit. In this case, the notification from the widget is
> > > required to know cursor moving because no other trigger to
> > > activate the input context will occur.
> >
> > I think that it is enough to acquire required information whenever
> > auxiliary widget is displayed, in order to solve this problem.
> > Is this idea insufficient?
>
> Not enough.
>
> Suppose an input method that has a status widget always
> shown. To satisfy the user interface requirements of the input
> method, the status widget has to follow cursor even if preedit
> does not exist.
>
> To implement such behavior, There are two ways.
>
> 1. Invoke inputMethodQuery() at every opportunity (such as
>    filterEvent() and mouseHandler()) to detect change of cursor
>    position in the input widget
>
> 2. Invoke inputMethodQuery() when the input context has been
>    notified about something changed in the input widget
>
> Lars's API design implies 1, but I prefer 2.

You're right. We've gone through the problem here again, and it looks like we 
can not avoid having a notification from the input widget for this. There are 
two places this is needed: input method support and accessibility. So we will 
add some notification method again. 

Cheers,
Lars

>
> Note that the input context itself is not changed while the
> cursor position is changed in this case. I think that Lars's
> design is assuming any change of properties occurs by some
> change of input context.
>
> -------------------------------
> YamaKen  yamaken at bp.iij4u.or.jp