Hacker News Reader

Comment by Diederich | Hacker News Reader

I first gained access to a UNIXish system in the fall of 1988, which was some SunOS4.x variant I believe. The shell was csh, the editor was vi, and I had no idea what I was doing, nor was there any help in my immediate area to be found. And no, I didn't know about 'man' yet.

Somehow I discovered that 'vi' was the way to edit things, but I didn't know how to exit, so trial and error led me to pressing control and z at the same time. Bingo! The shell is back.

After some days of building up background suspended vi instances, the system administrator came in and gave me a few pointers.

View on HN

Similar experience, except a ATT 3B2 (and thus no job control). I would have read the man page, but there were none, nor were any books available.

I finally just pulled the plug out of the wall.

It's not much of an exaggeration to say that I became a lifelong emacs user because the start screen even back then explained how to exit.

Holy shit, ATT 3B2, I haven't heard any mention of those in a long time.

So, six years later, in 1994, I found myself as the UNIX system administrator for a large (at the time) datacenter at Keesler Regional Medical Center, Keesler AFB, Biloxi Mississippi, where we had quite a number of 3B2s. They were, in fact, the only machines we had that were rated for classified data.

If memory serves, they ran AT&T System V release 3 UNIX.

We also had, among others, DGUX, SCO, AOS/VS, Sun and Vax/VMS servers. Of course each of AOS/VS and VMS had their own, non-UNIX operating systems.

Funny story: we put word out to AETC (Air Education and Training Command) that we needed more 3B2s, and a few months later, some pallets arrived from Homestead Air Force Base, which had been destroyed by hurricane Andrew a couple of years prior. These servers had green and brown smears on them. Clearly they had been submerged by hurricane flooding for some period of time. So we ran a long extension cord out to the parking lot, plugged them in and stepped back. Much to our surprise, several of them actually worked, and later ended up racked in our datacenter.

Ha. I miss the days when you didn't have to worry about your Unix box blowing away in a hurricane.

For what it's worth, I also used DG/UX, MV/UX and AOS/VS on Data General equipment (MV/10000, MV/4000) at the University of Washington in the 1985ish time-frame.

I remember porting many BSD 4.x utilities to MV/UX. What a learning experience, with the MVs differently-formatted byte vs. word pointers.

The C compiler was really nice, though.

Fun times.

You might be the first person I've run into that has heard of AOS/VS. I've run into one other person over the years that has heard of DG/UX.

Good times indeed.

My first Unix experience (~1993) was in a heterogeneous shop doing electronic pre-presss.

SunOS 3.5 on OEM PDP-11s. SunOS4.1.4 on SPARC Stations. SunOS5 (Solaris) on SPARC Stations. DG/UX on a pizza box. SCO on (I think) a 386

> ATT 3B2

There are a number of papers in pure mathematics authored by Doron Zeilberger and Shalosh B Ekhad. (See https://sites.math.rutgers.edu/~zeilberg/ekhad/papers.html for a list. I don't think it's exhaustive)

Shalosh B Ekhad is Zeilberger's computer, which once upon a time was an AT&T 3B1.

In Hebrew, 3 is "shalosh" and 1 is "ekhad".

(The 3B1 isn't actually much like the 3B2, despite the very similar names.)

I'm an emacs user to this day solely because it told me how to use it when I first opened it.

Wait really? I'm a lifelong emacs user too[0] but it only tells you how to exit if you open it without a filepath. If you're in a file because EDITOR was set to emacs by some other person and you try to exit it just says "C-c" or "C-c C-c is undefined" when it should just tell you how to exit. Or even better why doesn't it just exit on "C-c C-c"?

[0] Well I might write my own emacs-inspired, ruby enabled editor one day, but that's another story.

That's true. If you tried Control-H, you'd get a lead. Unless of course that was mapped to your main backspace key, which sometime happened in the bad old days.

As to why Control-C doesn't just exit the editor, I suppose it was generally considered to be a bad idea. And in the case of emacs, it might take five or ten minutes to restart. :-)

(It's worth noting also that GNU emacs ran on many platforms, some of which for which Control-C was not any kind of "kill my process" character.)

How careless of him, you obviously wouldn't have been the type to free them when you were done ;)

Did you save any?

Yes, I was able to save the file using ZZ rather than :w

It's been a long time, but I was probably hung up on not being able to experimentally figure out : commands, where I was able to figure out control-z and ZZ.

ZZ exits vi(m) though?

Yes, it sure does! Well, after 30 years, my memory of the particulars is a bit hazy. (:

Reply on news.ycombinator.com