You are both right.

The SPARC lies behind my hybrid firewall/proxy config which makes running
exploits difficult. ( Notice I did not say impossible )
99.9% of script kiddie exploits are written and run on x86 hardware.
So, for people running anything but the x86 it increases the probability
that you won't see a succesful crack due to differences in memory and CPU
structure.
Remember, sometime a buffer overflow or other type of exploit is architecture 
independant or application specific and no platform changes can protect you.

I sleep better knowing that any attacks thrown at my Sparcs are usually:
1. Script Kiddies on x86 boxen
2. directed at my firewall first

The four Sparcs are only in rotation for serving web pages via the proxy and its
cache so the weakest link in my scheme is still the firewall/proxy.

See for yourself.

Go to www.netcraft.com and lookup www.magusnet.com 
on bothe the secure and non-secure servers pages.
or
telnet magusnet.com 80
and type
GET <enter>

My questions: What is wrong with the 2.0.35 kernel?
              For that matter, what is wrong with 0.99pl53?

If the kernel works, why fix it?

That comment on not being an up to date kernel in the response smacks
of the upgrade mentality of those "other" operating systems.

HTH. HAND

Jean Francois - JLF Sends...
President & CEO - MagusNet, Inc., MagusNet.com, MagusNet.Gilbert.AZ.US
Director Of Managed Services - OpNIX,Inc., www.opnix.com
OpNIX - Simply Better Bandwidth
602-770-JLF1 - Cellular, ICQ:  8137851
See My Certifications:
http://www.brainbench.com/transcript.jsp?pid=1214021


It seems like on Sun, Oct 15, 2000 at 10:46:47PM -0700, plug@arcticmail.com scribbled:
Orig Msg> 
Orig Msg> But for a packet to reach JLF's 2.0.35 kernel, it
Orig Msg> must first survive the 37 layers of minefields,
Orig Msg> flamethrowers, anti-personnel^H^H^H^H^H^H^H^Hacket
Orig Msg> devices, machine gun emplacements, nerve agents...  :)
Orig Msg> 
Orig Msg> 
Orig Msg> D
Orig Msg> 
Orig Msg> * On Mon, Oct 16, 2000 at 12:48:36AM +0000, George Toft wrote:
Orig Msg> > 
Orig Msg> > "J.L.Francois" wrote:
Orig Msg> > > 
Orig Msg> > > The next time someone says Linux can't hang show them this.
Orig Msg> > > This is a SparcClassic with  32MB RAM serving WWW pages
Orig Msg> > > and my Sparc test platform.
Orig Msg> > > Obsolete hardware to everyone but me.
Orig Msg> > > 
Orig Msg> > > I wanted to get this out before I have to start turning equipment
Orig Msg> > > off to mount in imy refrigerator converted to server cabinet.
Orig Msg> > > 
Orig Msg> > > Woops, gotta go and reboot my wifes Win98 desktop, again.
Orig Msg> > > 
Orig Msg> > > ==============================================
Orig Msg> > > Linux heirophant 2.0.35 #1 Wed Feb 17 20:37:03 CET 1999 sparc unknown
Orig Msg> > >   5:55pm  up 406 days,  6:57,  1 user,  load average: 0.04, 0.01, 0.00
Orig Msg> > 
Orig Msg> > Uptime is nice, but I hate to see uptime over 150 days - it shows that
Orig Msg> > the kernel has not been updated against various exploits.  The 2.0.35
Orig Msg> > kernel is about three years old, and I'm pretty sure there are some remote
Orig Msg> > compromises up to the 2.2.13 kernel.  Maybe the sparc kernel is different?
Orig Msg> > 
Orig Msg> > Feel free to correct me if I am wrong.
Orig Msg> > 
Orig Msg> > George
Orig Msg> > ---
Orig Msg> > Compared to Linux, working with the Windows desktop is like performing
Orig Msg> > gymnastics in a strait jacket.
Orig Msg>