In article <20041213224357.15e7d17b.jkunz@unixag-kl.fh-kl.de>,
	jkunz@unixag-kl.fh-kl.de (Jochen Kunz) writes:
> On Mon, 13 Dec 2004 09:15:13 -0600 (CST)
> fredb@immanent.net (Frederick Bruckman) wrote:
>> Perhaps we could make this just a warning?
> Don't know. I have no clue about X11, I am a kernel hacker.

I would let herrb make the call. (He added the check in XFree86.)

> [jkunz@Zimbo jkunz]$ xclock -update 1 
> Xlib: Did not parse entire setup message: parsed: 848, message: 1600
> Fontconfig error: Cannot load default config file
> Xlib:  extension "RENDER" missing on display "MrPomeroy:0.0".
> And now I have xclock(1) on my screen. xterm(1) works too, but gvim(1)
> does not. I started top(1) inside the xterm. This worked for some time
> but then, when switching back to the virtual desktop with the xterm, the
> xterm was blank and didn't react to any input.
> [jkunz@Zimbo jkunz]$ gvim tmp.txt                   
> Xlib: Did not parse entire setup message: parsed: 848, message: 1600
> XIO:  fatal IO error 60 (Connection timed out) on X server
> "MrPomeroy:0.0"
>       after 45 requests (39 known processed) with 0 events remaining.

It sounds like there's more wrong than just a spurious warning. Did you
send Matthieu Herrb the tcpdump file?  I am a little curious as to what
SGI would put in there, not that it would necessarily be obvious from
looking at the raw bits.

Are you in a position to open a support ticket with SGI? This does not
look like it would be specific to NetBSD.  If you could reproduce it
with, say, one of those stand-alone boot-from-CDROM linux distributions,
that might demonstrate to SGI that they have a real problem.


Frederick