Subject: Re: How stable is NetBSD on the PPC (specifically G4)?
To: Michael Jessop <mjessop@copyright.com>
From: Henry B. Hotz <hotz@jpl.nasa.gov>
List: port-macppc
Date: 11/01/2000 09:45:08
At 8:42 AM -0500 11/1/00, Michael Jessop wrote:
>I progressed from OS/2 to RedHat Linux and BeOS on my PII, to FreeBSD.  I am
>now running YDL on my Mac and am wondering about NetBSD.

What's YDL?

>So, if you wouldn't mind sending me a few opinions I would appreciate it!!
>Thanks!

I've been running 1.5 alpha 1 for several months on a 7500 (w 604 
upgrade).  The core system has been solid as a rock except for some 
strange networking bug that the ISS "heavy plus" security scan trips. 
It crashes the machine hard and the keyboard is dead so I can't trace 
it.  (ISS incorrectly identifies it as a known SunOS bug and 
recommends a specific patch.  JPL has a trouble ticket open on the 
problem since I would like to know what the UDP packet is which 
triggers the crash.  Maybe I can stick an IPfilter rule in to block 
it.)

X is flakey.  I was able to crash it with minimal effort.  I do leave 
xdm running as a screen saver though since otherwise the text console 
leaves a few uninitialized pixels around the edges.  The machine is 
primarily a server in a different building from me.

NFS works fine with a Solaris 2.5 box both as client and server.

I was not able to get ssh 1...30 to work, but the pkgsrc version at 
1...27 works fine.  MPI 1.2.1 works with only a couple of fixes which 
were easy to invent.  FFTW worked out of the box.  Postgres works 
from pkgsrc; it would work directly except for shared library linker 
options.

YMMV of course.  I'm currently putting 1.5 beta on a machine to see 
if that fixes the security vulnerability.

The major piece of unfinished business for the x500 machines I use 
IMNSHO is getting a way to dual-boot with MacOS on the same disk. 
Bill Studenmund got the code over from port-mac68k so the BSD 
filesystems can live on the same disk with MacOS, and the current 
booting procedure already relies on a minimal Apple Partition Map so 
Open Firmware can find the first-stage booter.  It's really close.

__________________________________________________________
The opinions expressed in this message are mine,
not those of Caltech, JPL, NASA, or the US Government.
Henry.B.Hotz@jpl.nasa.gov, or hbhotz@oxy.edu