On Tue, 4 May 1999, Jonathan Stone wrote:

> 1) screenblank during idle leaves cursor sprite visible on sfb.

  I haven't seen this on the sfb on a 5000/25 using the default screen
blanker.  Another sfb system is running xscreensaver, so I can't tell on
that system.

> 2)  On an lk-201, caps-lock seems to only sense one direction
>     of the capslock key. To toggle capslock *off*, I have to
>     bounce it twice.     e.g., if I type
>     "hello world<CL>hello world<CL>hello world<CL>hello world<CL>"
> 
>     I get:
>  
>      hello worldHELLO WORLDHELLO WORLDHELLO WORLDhello world
> 
> Looking at xev output, it seems like the driver is only getting one
> message per capslock up/down - if I bounce <capslock> and watch in
> xev, i get an KeyPress for an  up/donw, then a KeyRelease for the next
> up/down, then a KeyPress for the next up/down, usw.
> That seems to be FUBARing the state machine.
> 
> I dont recall ever seeing this with the old Xserver. I'd guess the
> casplock key needs to be in "report up/down" state, and a reset string
> in Xpmax is putting it in some other mode?

  A quick work-around is to run the X server without keyboard extensions
(-xb option).  The problem is that XKB extensions tries to hand the shift
lock function, and Xpmax unconditionally does it as well.  I'm trying to
test a fix for this, but it didn't seem to be working on the 5000/260 I'm
running at work - but I'm not certain if I was actually running the server
with the fix.  I've rebuilt it and will be testing it later today when I
get to work.

Michael