This is the item for discussing systems that are historically 
interesting, but were evolutionary dead ends -- they didn't lead 
directly to modern computer architectures or operating systems.  Often 
these are quite fascinating because they represent different design 
philosophies than those we're used to.

12 responses total.

I've been working my way through a website by Jim Horning called "The 
Way It Was: Tales from a life in computing."  It's done in blog format, 
but it's meant to be read serially, starting with the first item.  It 
details his experiences with early computers, in particular the Bendix 
G-15.  The G-15 was a physically small computer for the time -- about 
refrigerator-sized -- and had some interesting features to reduce the 
number of vacuum tubes required.  All arithmetic was done bit-serially 
on one-bit-wide buses, for example, using magnetic drum memory for 
storage.  The stories on the site make fascinating reading if you're 
interested in this sort of thing.
http://horningtales.blogspot.com/2006/05/introduction.html

cadet

 I still mourn the loss of Apollo's Domain/IX, a Unix-like operating
 system that went head to head with SunOS in the mid-80s and got its
 butt kicked.  At the time I started at the University of Michigan
 (1986) Apollo workstations were scattered among the computing labs
 on the engineering campus.

 I remember them as being non-standard (which is probably what killed
 them) but way ahead of their time in many ways.  Back in 1986 or '87
 they had: 

   1)  a completely transparent-to-the-user network filesystem.
   2)  filesystem acls
   3)  a distributed registry
   4)  an interesting and innovative integrated windowing environment
       with powerful text editing and great keyboard support.
   5)  the ability to boot diskless and run from the network file
       system (years before anyone had ever uttered the phrase
       "thin client")

 Most or all of these features (except the edit pads from the window
 manager -- I miss those and the way editing was done through a standard
 window-system widget available to every program) eventually showed up
 in other operating systems, but not until 5, 10, or even 15 years later.

And universes for BSD and AT&T style Unixes!

I started out on those same Domain boxen in 89.  I really loved those. 
I almost bought one of the ones I did a lot of circuit design on when I
saw it at property dispo.  It had been doing duty as a print spooler
after the rest of the lab was converted to Suns.  Poor old "not".  I
wonder what ever happened to it.  Nothing good, I suppose.  Oh, well.  I
guess it's semi immortal as long as I remember it fondly.  :~(

Perhaps not an O/S, but whatever happened to LISP?  I loved all them CAR's
and CDR's...  ;-)

Look up squeak

Well, there were LISPmachines, whose hardware and OS were completely devoted
to supporting LISP.

I don't so much miss LISP, but that's because it didn't serve my niche.

Lisp is alive and well.

Now we call this item "retro-computing". Amazingly, a lot of these "dead ends"
can now be revived in simulators and a lot of the documentation has been
preserved on bitsavers.org and elsewhere. There are even a few of the old
systems running on lovingly restored original hardware (livingcomputers.org,
etc.).

This category includes every non-Unix-based proprietary OS other than Windows,
z/OS, and OpenVMS. And OpenVMS will be leaving the exceptions list soon.

In particular, DEC pre-VMS produced (DEC System-10, -20, RSTS, etc.) or
inspired (ITS, etc.) many systems whose influence is still significant today.
DEC made them all obsolete by fiat when they declared VMS to be the One True
OS.

May be the real dead ends are non-binary architectures.
http://ternary.info/
Will quantum computers ever promote binary hardware to retro?

Response not possible - You must register and login before posting.

- Backtalk version 1.3.30 - Copyright 1996-2006, Jan Wolter and Steve Weiss