Subject: RE: Starting X from /etc/rc.local
To: Frederick Bruckman <fb@enteract.com>
From: John A. Maier <johnam@kemper.org>
List: netbsd-users
Date: 01/13/2000 17:21:00
Oh, it gets better.  I kept saying xdm when I'm using kdm!!!

Argggg!

If you enter you password and hit enter, the "ghost" starts hitting   
enter.  If I enter the password and click on "Go!" everything works.

I re-tested on wscons and everything works okay.

The stock xdm work okay too.

Sorry for all of the trouble.

John A. Maier
Kemper Military School and College
Director of MIS
Voice: 660-882-5623
Fax: 660-882-3778
Email: johnam@kemper.org
Web: www.kemper.org


> -----Original Message-----
> From: Frederick Bruckman [mailto:fb@enteract.com]
> Sent: Thursday, January 13, 2000 3:35 PM
> To: John A. Maier
> Cc: netbsd-users; netbsd-help
> Subject: Re: Starting X from /etc/rc.local
>
>
>
> On Thu, 13 Jan 2000, John A. Maier wrote:
>
> > I tried starting X from rc.local with the command:
> > /usr/X11R6/bin/XF86_S3 -query johnam &
> >
> > For the most part it works, but after I login, odd keyboard
> stuff starts
> > happening.
> >
> > I open an xterm and it starts scrolling as if the Enter key
> is stuck.  I
> > hit some keys and it stops.  But then other characters will
> > spontainiously start printing.
>
> I'm doing the same thing since 1.4.1 on a pair of machines, the
> Xserver a 486dx2-66 w/24mb, the loghost a 486dx4100 w/48mb, also
> XF86_S3, using wscons on the Xserver. The only difference is, my
> rc.local calls "sh /etc/Xserver0 &", and /etc/Xserver0 has only "exec
> /usr/X11R6/bin/X -query handy.". (The exec is to get rid of the
> intervening subshell. With that, "ps" shows the PPID as "1") I never
> saw the problem you describe--I tested the script from the console
> before putting it into rc.local. Maybe your Xserver is just getting
> tangled up by the job control magic.
>
> Also, XFree-3.3.5 (release snapshots) is considerably more stable than
> the version that shipped with 1.4.1, even though I'm still using _S3.
> ISRC XF86_S3 is deprecated in favor of _SVGA (in -3.3.5), but why mess
> with something that works?
>
>