Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!snorkelwacker.mit.edu!bloom-beacon!dont-send-mail-to-path-lines
From: mouse@larry.mcrcim.mcgill.EDU
Newsgroups: comp.windows.x
Subject: Re:  XWarpPointer and ICCCM....
Message-ID: <9101210514.AA09897@Larry.McRCIM.McGill.EDU>
Date: 21 Jan 91 05:14:07 GMT
Sender: daemon@athena.mit.edu (Mr Background)
Organization: The Internet
Lines: 36

>> Section 6.2 of the ICCCM says "In general,  clients should not warp
>> the pointer".  More specifically,  Section 4.2.7 says "Clients
>> should not warp the pointer in an attempt to transfer the focus;
>> they should set the focus and leave the pointer alone".

> Then how do we accomplish this type of action?  I find it extremely
> constrictive to have to continuously move the pointer to the next
> text field in the "entry-form" application that I've built.

> How do people do it otherwise?

For the most part, I suspect they don't.  I've seen three programs that
use text-input widgets but mishandle this.  One of them is a commercial
desktop publishing program, one is a privately developed program
someone I know is working on; I won't discuss those.  The third one is
xpostit.  (Caveat: I don't know whether this is true of the current
xpostit.  It was definitely true some months ago when I tried xpostit.)
The problem is that when a note (implemented as a distinct top-level
window) is given focus by the window manager, keystrokes are
nevertheless ignored unless the pointer is inside the window.  The
other two programs exhibited symptoms with different details, but all
of them had the same basic problem: they were depending on the X server
to do keyboard event distribution to their various subparts, but they
weren't doing this right by setting the focus; instead, they were
simply depending on the pointer position to arbitrate at the
sub-top-level-window level.

Just a plea to try to get app authors to deal with keyboard input
properly.  (Meaning: test under many different focus paradigms; don't
assume that everything's fine just because it works for the developer
and a couple of others (who happen to use similar paradigms).)

					der Mouse

			old: mcgill-vision!mouse
			new: mouse@larry.mcrcim.mcgill.edu