Path: utzoo!attcan!uunet!decwrl!wuarchive!psuvax1!psuvm!art100
From: ART100@psuvm.psu.edu (Andy Tefft)
Newsgroups: comp.sys.apple2
Subject: Re: 8-bit screen blanker
Message-ID: <90281.104905ART100@psuvm.psu.edu>
Date: 8 Oct 90 14:49:05 GMT
References: <5335.26e5359c@vax5.cit.cornell.edu>
 <90280.134725ART100@psuvm.psu.edu> <2101@unsvax.NEVADA.EDU>
Organization: Penn State University
Lines: 29

(Scott suggests interrupts from a clock card...)

Well I'm doing this on a //c so that's kind of out, although
I was planning on making it work with 'vblclock' anyhow (which
I know doesn't always run). Maybe I'll see how useful the basic.system
version is first.

The main idea of this is to grab the kb input vector first and
save the time each time a key is pressed. Then at every interrupt
it checks the time and if the time since last keypress is greater
than some interval n, the current text screen is saved (and
switched to in case a graphics program is running) and then blanked.
I'll put some fancy little doojiggy in there so you can tell that
you can press a key to continue, maybe a bouncing clock. Then you
will hit a key to resume what you were doing. My first impression
is that I can't tell whether the person was looking at hi-res,
lo-res, 40-col text, etc. so I'll let them cycle through the modes
and select the one they want.

I guess the program should keep running "in the background"
while the screen is blank, till it stops when it wants a keypress.
The keypresses needed to restore the screen will not be passed
to the program (as it is with other screen blankers i've seen).

This will of course be limited to programs which work with vblclock
and use ksw for input.

The problem then is how do you tell when the last user activity
was, if the program doesn't use ksw for input?