🚀 go-pugleaf

   In article <EBERSMAN.91Mar12195818@cfmartin.uu.net> ebersman@uunet.uu.net (Paul Ebersman) writes:
   >
   >I notice that a fair portion of the time, when I use the middle
   >button on the mouse to paste, it doesn't show up until I hit some
   >keystroke. It hasn't been happening consistently enough to establish a
   >real pattern, but it is very annoying.

   I have noticed exactly the same thing (and yes, it is annoying).
   A posted solution would be really helpful!!

   **> jacques

   Jacques Vaisey              |     email -> vaisey@cs.sfu.ca

I include a previous posting discussing the problem containing a
workaround that we use(d) in the sky-mouse package
(icsi-ftp.berkeley.edu pub/ai/sky-mouse). But it may not solve all the
problem in your system. Because you may have mouse functions that do not
call mouse::set-point. The real fix would be in xdisp.c I think.

Perhaps this is a start ... 

-- Heinz
--------------------------------------------------------------------------
Heinz W. Schmidt                                     hws@icsi.berkeley.edu
International Computer Science Institute, Berkeley   (415) 642-4274   x175
/\\/\\|;; 
\//\//|--  Make it as simple as possible, but not simpler


Return-Path: <hws@ICSI.Berkeley.EDU>
Date: Thu, 29 Nov 90 15:02:05 PST
From: hws@ICSI.Berkeley.EDU (Heinz Schmidt)
To: epoch@cs.uiuc.edu
Subject: Hanging redisplay


Redisplay seems to hang from time to time as reported also by others
on this list.  Below is a workaround that fixes the bug by explicitly
calling redisplay on set-point but I think a real fix would go with
redisplay instead. The extra redisplay should be avoided since
set-point will be called by many mouse commands at the beginning of
command execution.

The bug can be reproduced as follows:

   Have 2 screens beside each other. Select two different buffers in
the different screens. Use the mouse to set point in screen A to the
end of a line (click behind the line) and to set point in screen B
someplace.  Move the mouse over screen A to select it, but don't click
to set-point. Type a character. Move the mouse back over screen B and
set the point a couple of times by clicking (without dragging).  The
cursor does not move although the (internal) point follows correctly
as one can see when typing any character in screen B.
   Drag calls redisplay explicitly. But subsequent clicks continue to
leave the cursor behind. Also any even-numbered iteration of
type-to-other-screen-click-here releases redisplay but the next
iteration blocks it again.

;;; modifies epoch-lisp/mouse.el    
(defun mouse::set-point (arg)
  "Select Epoch window mouse is on, and move point to mouse position."
  (select-screen (nth 3 arg))
  (select-window (nth 2 arg))
  (goto-char (car arg)) 
  (redisplay-screen)		;workaround 'hanging redisplay bug'
  ;; traces bug when previous line commented, point moves, cursor stays
  ;;(message "%s %s %s %s" arg (point) (query-mouse) (query-cursor))
)