Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rutgers!sri-spam!ames!ucbcad!ucbvax!UCBVAX.BERKELEY.EDU!holly%jason
From: holly%jason@UCBVAX.BERKELEY.EDU (Holly Heine)
Newsgroups: comp.windows.x
Subject: (none)
Message-ID: <8703161938.AA04634@jason.berkeley.edu>
Date: Mon, 16-Mar-87 14:38:10 EST
Article-I.D.: jason.8703161938.AA04634
Posted: Mon Mar 16 14:38:10 1987
Date-Received: Tue, 17-Mar-87 04:16:19 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Distribution: world
Organization: The ARPA Internet
Lines: 23


	I have been writing programs using the xwindow package available
	on the sun workstation. In writing a simple window I have uncovered 
	a very annoying problem that I was wondering if anyone knew anything
	about. The program prompts for user input and uses the XWindowEvent
	command with the mask set for Keypressed to handle the event. The
	problem is that every time a key is pressed it generates 2 events.
	The first is the intended key character and the second "fake" event
	is an "unprintable" character. This fake eventis then in the event
	queue so that the next XWindowEvent attempts to use it and 
	causes the program to work improperly. This problem was isolated
	while running the window problem under dbx. The fake event was
	of type keypressed. Two events (the intended and the fake) were
	generated every time a single key was pressed. However if two keys
	were pressed simultaneously, ie. shift-character, only one event 
	was generated. This simple program works correctly with no extra
	events if run on the uvax. The program will also work if extra
	XWindowEvent statements are put in to eat up the fake events, but
	this is in no means a solution. Has anyone else seen this problem with
	any of there xwindow programs? Can anyone suggest a solution or some
	one else to present this problem to?
	Thank you very much,
		Holly Heine   (holly@jason)