Re: Linux and multi-user options

Top Page
Attachments:
Message as email
+ (text/plain)
Delete this message
Reply to this message
Author: Alexander Henry
Date:  
To: plug-discuss
Subject: Re: Linux and multi-user options
On Fri, 7 May 2004 18:29:59 -0500, Jerry Davis <> wrote:

>> How does one truly utilize Linux in a multi-user configuration?
>
> In linux it is ALREADY there. I do it all the time.
> And as usual there are several ways.
>


War stories from when I was getting my BS in Computer Science (1992 thru
1997):

There were Sparc stations, older DECs, and DECs with oooo... color
monitors in the main lab, mixture of older Sparcs and 386's in another,
and Sparcs and Macs in a distant one for more 'public' use, and dumb tty
terminals in the library that were either open to the main accounts
through telnet or hard-wired to run lynx in the library's search system.
I got my first account for a CSC101 class, which I used for my entire
accademic career. I could log in to any one of the Sparcs or decs, and
start working. Any lab, any building, log in, 'cd homework5', and go.
Actually, first thing I would do was an rwho, which would display everyone
that was logged into the network. So if my friends jhantin or blondie
were in another building, I could zwrite them a whassap, or send them a
talk session. who would display just logins of people at my machine,
which was useless. I could configure X with a certain window manager,
these settings were preserved when I logged into a different machine. Now
I know this is because each computer was linked to the same NFS server.
They all knew to look at the NFS, find /home/a/h/ahenry, and hook me up.
If I wanted to use a program that was only available, say from the
engineering lab's server, I'd telnet to the server, xhost+, and when I ran
the program on the foreign machine, the window would display on the
machine at which I was physically sitting. Later, I learned about nohup.
I exploited this in my advanced classes, say when I was learning about
random number generators and how to test their effectiveness. I would
compile several slightly different versions of a test, and at say 11pm
telnet over to other machines and run one on each, nohupped, then logged
off and went home to sleep. When I woke up, the results were there,
except for maybe one or two instances when one of the Indians who
frequented the labs from 12mid to 4am bumped into the power button. Oh
ya, did I mention macs and pc's? Macs were there to run Wordperfect,
Printshop, and paint programs. They did have NCSA Telnet, which gave you
a dumb terminal into one of the 'real' computers, and you could do
whatever work you could muster, as long as it only involved command line.
PC's were toys. You'd learn how to write a terminate-stay-resident
program, maybe write an assembly routine that did ray-tracing graphics,
then move on for real work. In 1995, we got a donation from Silicon
Graphics which included I think eight Indys and a $250,000 Onyx. We
learned about 3D graphics in OpenGL. That was the first time I ever had
do think about the physical computer at which I was sitting; 3D displays
were 0.05 times the speed over the network than at the monitor physically
hooked up to the screen.

Then I went to work at Honeywell, 1997. First thing I wanted to do is get
to know the system, where the apps were, which system I needed access to
to get my apps, my co-worker's logins so I could send them whassaps and
ask questions when I had problems and so I could cd over into their public
directory into their account and share files. First roadblock... The
entire system was... toys. PC's. I found the system admins, and asked
them how I could get a UNIX prompt. They said the only way was with
something called Exceed, which was $250/seat and managers didn't ever sign
off on it. Okay... so how do I telnet to my co-workers' PC's and get
their files? You don't... Oh... So where is the central NFS, where the
work is done? What do you mean? I said the network? They went, oh,
here, and showed me the Netscape icon in the start menu.

Finally, in 2001, I FINALLY have all this functionality back. After they
got rid of the X terminals in all labs at ASU, I was forced to install
Mandrake on my PC, and did a ssh/xhost+ to ASU's research cluster to run
Matlab. AHHHh..

Linux hackers are not extremists. They're reactionaries. We're trying to
get things BACK to where they were when real work was done.

--

--Alexander
---------------------------------------------------
PLUG-discuss mailing list -
To subscribe, unsubscribe, or to change you mail settings:
http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss