[FoRK] Fwd: Where to put the frosh (... on a VAX!)
Joseph S. Barrera III
joe at barrera.org
Wed Apr 3 22:07:15 PDT 2013
Remember when computers were expensive?
Remember when emacs had a primary value as an EDT emulator?
Remember when it mattered how many terminals you could connect to a
Remember when 4MB was a luxurious amount of RAM for a multiuser computer?
Remember when you had to dedicate one of your serial lines to UUCP?
Remember when you would send email to recipients without @ signs?
Remember when you would send email to recipients using @, %, *and* !
signs? (not shown below)
-------- Original Message --------
Subject: Where to put the frosh
Date: Sun, 16 Feb 86 14:41:17 pst
From: brian (Brian Zill)
To: david.joe.maria.mike, scgvaxd!mike.mueller.purves.rhall.rick
The topic under discussion seems to be the following:
All the frosh are going to get on some Unix machine for a little while
(3-4 weeks?) in CS6 this year. (Actual dates I don't know, sorry).
muddcs is currently overloaded as it is, and the frosh would kill the
system. Such poor system response would not give the frosh a good
impression of Unix, which is just the opposite of what we'd like.
Engineering's HP9000 is currently sitting unplugged in the new
machine room as the link-a-bit is dead in the water. Maybe we should
try to borrow the HP9000 for the frosh to use. (Extremely remote
possibility that we'd be able to keep it for ourselves).
Purve's comments (selected):
> There may be some nuisance getting adequate terminals. One
> possibility is to move half or more of the frosh terminals to
> the HP9000 during the relevant time period of 3-4 weeks. We
> would need to leave some on hmcvax though, for their word
> processing and other purposes.
I'm all for putting the frosh somewhere other than muddcs as I
agree the load would be terrible. The only alternative I see
that would allow the frosh to use muddcs would be for cs101 to not
have any programmming homeworks during that time.
However, there are serious problems with putting them on the HP9000
as well. A few that come to mind:
The HP9000 has a current maximum capability of 6 terminal
lines (not including the console). No ethernet capabilty.
UUCP unrealiable without a dedicated line (cutting lines
down to 5). The disk isn't the largest thing in the world
It has never been testing under a large load (i.e. more than
2 or so users at a time). Remember, we only have the single
It runs System V, not 4.2bsd. General preferences aside,
the fact is that 4.2 has a bigger presence on this campus
than System V and probably will for quite some time. It
would be better for the frosh to learn about 4.2. Also,
(and this is a biggie) our staff is not trained to support
a system V machine.
The HP9000 is not set up well. Unless great advances have
been made since the last time I logged in, things on the
HP9000 filesystems are not all arranged nicely. It took us
all summer of '84 to arrange muddcs (and we're still not
perfect). There is no emacs on the HP9000 (and thus no
EDT emulator). I'm not even sure there is a Pascal compiler.
So what to do? Well, why think small? Let's go for a VAX! Depending
upon when in the semester we need the extra machine we could go for
one of the following scenarios:
(I know, we'd never get it, and it'd be over Andy's dead body,
but it would be fun to try :-). But actually, until it gets
more memory it isn't much use to anybody anyway. Frosh
needing to use VMS could use the old FROSH 750 (which is
also currently sitting unplugged in the new machine room).
The FROSH 750:
It's currently a very expensive table in the new machine room.
By my count, Andy should have a spare RA81 or RA80 lying
around. I doubt the administration is planning on taking
possession in the near future by the looks of the thing.
The NASA 750:
In a post-memory arives for the 8600 timeframe, we could take
the NASA 750 during the period of time between NASA's moving
to the 8600 and 4CCVAX's moving to the 750.
Advantages to a VAX are basically that it doesn't have all the
disadvantages that the HP9000 has. Plus we could rlogin, rcp, rsh,
etc. over the ethernet between the machines!!! Getting one to run
Unix is simply a matter of wheeling a disk over to our machine,
loading it up, and moving it back. Or if Mike wants to do a demo
for his operating systems class, they could load it from the boot
tape... Oh well, I suppose everyone gets the idea. I hope people
don't think I'm way out of line for suggesting this, but seeing the
frosh machine sitting there unplugged got me thinking...
P.S. We have loads of accounting stats collected that show increased
usage on muddcs if anyone wants to use them to argue.
P.P.S. In my logic design class, Goldstein recommended that we all
use muddcs to write a C program for his class because the HP9000
wasn't up!! Seems to me it should be up for engineering...
More information about the FoRK