Re: storing mouse-movement events in recent-keys and keyboard macros

[Kim F. Storm]

"Eli Zaretskii" <address@hidden> writes:

> 
> Please be sure to test these changes with mouse-avoidance-mode in its
> various submodes.  The special treatment of help-echo events is there
> to prevent unpleasant effects when you set the mouse-avoidance-mode
> to `banish', for example.
>

Originally, I think the treatment of help-echo is there because I
reported it (during pretest) as a problem having C-h l flooded with
<help-echo> events.

Anyway, I don't see how my changes - which only influences what is
returned by (recent-keys) - can have an effect on mouse-avoidance-mode
as it doesn't look at (recent-keys).  Actually very little code does
that...

More strange, I cannot get mouse-avoidance-mode == banish to work even
on a vintage 21.1 if I enable tooltip-gud-tips-p.  For some reason -
even when the mouse has been moved out of sight - every key I press
(when the active window is a .c file with active mouse-tracking)
is followed by a mouse-movement event!  So I cannot even type C-h l
to see what's going on without clicking on a non-.c buffer.

This is very similar to the problem I reported a few days ago,
saying that mouse movement after C-s  clears the mini buffer
(without exiting isearch).

I see this on two diffent GNU/Linux systems (running Redhat 6.0 and 6.2),
both under X (Xfree 3.x) and KDE desktop (1.x).

-- 
Kim F. Storm <address@hidden> http://www.cua.dk