Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!rpi!bu.edu!snorkelwacker!bloom-beacon!LARRY.MCRCIM.MCGILL.EDU!mouse
From: mouse@LARRY.MCRCIM.MCGILL.EDU
Newsgroups: comp.windows.x
Subject: Re:  input problem to X11 on a Sun....
Message-ID: <9008302245.AA20347@Larry.McRCIM.McGill.EDU>
Date: 30 Aug 90 22:45:57 GMT
Sender: daemon@athena.mit.edu (Mr Background)
Organization: The Internet
Lines: 42

>> A new strange problem has cropped up on *one* of the Sun3/50's that
>> use X11 here.  The user will be typing away and suddenly, all the
>> characters appear (in xterms) in uppercase.  Additionally the mouse
>> buttons are ignored everywhere on the root window.

Sounds as though the X server decided the capslock/shiftlock key went
down, or perhaps one of the shift keys went down but didn't come back
up.  Why it would decide this I can only speculate; perhaps your
keyboard hardware is flaky.  Try pushing your capslock/shiftlock key,
and perhaps the shift keys, a few times and see if that cures it (the
shift keys to generate valid down/up events to get the server in sync
with reality there; the lock key in case that's the problem).

>> Note also that *only* the characters [A-Z] appear in a/any window
>> (whichever has focus, yes, we can move the pointer though button
>> clicks are ignored...).  None of the other ASCII characters
>> (./><&$, etc...) appear in the windows.

Does this apply to just new keystrokes or to characters that already
existed?  If the latter, your server has a strange and wonderful new
bug I've never heard of before.

Hmmm, you can't type anything but letters?  Definitely sounds like a
flaky keyboard; when Lock is in effect non-letter characters normally
work just as normal.

It's a somewhat risky thing to try, but I've occasionally managed to
fix out-to-lunch keyboards without having to kill whatever is going on
by unplugging and replugging the keyboard itself.  This will usually
abort the machine (like L1-A), but a c (continue) command will, in my
experience, bring it back to life.  This is not a Factory-Approved
Solution; I even seem to recall some Sun documentation warning "don't
do this".  I don't know whether this was because most people will freak
when they get "Abort at 0x0f0087ca" / "> " or because there's a real
danger of damage to the keyboard; power-cycling equipment always
carries some danger, but I've never had anything worse than an abort
happen.

					der Mouse

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