Ultrix 3.0 pitfalls; please warn me

Charley Kline kline at tuna.cso.uiuc.edu
Tue May 23 03:58:45 AEST 1989


> >...  On our system (Vax 11/780) we
> >will be building 
> >News/Nntp, 
> Dunno .. but don't expect any problems

We run NNTP here on a 3500 with 2.2 (soon to be 3.0), and serve all the
campus rrn's. Works like a champ.

> >BIND (name daemon),
> The BIND there is 4.7.3 with some "improvements".  I had some troubles
> with that version tho so we put up 4.8 but left the resolver stuff alone.
> (The /etc/svcorder file is a very nice touch)

We're running the new 4.8+ version of bind from the University of Toronto.
Fixes a lot of bugs in 4.8, like the core cancer problem, and so on. Seems
to fit in seamlessly. Agreed that svcorder is wonderful. I like to be able
to tell the lookup routines to pay NO ATTENTION TO YP.

> >and using sendmail to route to
> Use MMDF .. it's good for the soul ;-)

Paul Pomes here is a rabid sendmail addict, so I expect our 3500 will be
running the new 5.61/IDA stuff soon. The hope is that IDA sendmail will
integrate the mail11d DECnet mailer, and some local stuff which does SMTP
over a DECnet connection to a PMDF gadget on one of the VMS machines here.

> The only other thing is that D-CWindows is slooooooooooooooow..

Agreed. But it seems to be fine if you don't run any of the DOGwindows
clients, which are not only slow but HUGE. I run the DOGwindows server
but only native X applications, and everyone is happy, including me,
since the speed is acceptable. This on a 6Mby VS2000.

-----
Charley Kline, University of Illinois Computing Services
kline at tuna.cso.uiuc.edu

"Flaring high or flaring early makes the little prop tips curly."



More information about the Comp.unix.ultrix mailing list