pollack@nmsu.CSNET (05/09/88)
I had a suspicion that my $30K 1186 was obsolete when it was delivered in November 1986. Confirmation came when my $2000 AT clone was much faster. Further confirmation came when Xerox announced a 50% discount. Now that Xerox has decided to move Interlisp-D into the Sun environment, we are faced with the problem of what to do with our pretty worthless Xerox hardware. Here are the ideas I thought of: 0) Downgrading the software to Viewpoint. Since the Interlisp software added $10,000 to the price of the equivalent office workstation, Xerox should do the downgrade and throw in the 4045 laser printer for free. Fairly useless, unless you give it to your secretary. 1) Trading it in for the software that will run on a Sun. There are plenty of Sun's around, but who will be able to afford the price of the "new" xerox software? Especially if it includes a useful file system and is maintainable from UNIX. 2) Using it as a PC. Unfortunately, the PCE software is fairly useless as it stands. One can't even run Ventura on it. One can't run anything on it at all, if any Interlisp code is running. Perhaps someone can redo the PCE to use the entire bitmapped screen. Perhaps an 80386/80387 replacement board is in the works at Xerox/OSD. 3) Using it as a Chat terminal to a Sun. That's what I do now. A 24 by 80 VT100 window is too small a bandwidth, so I reshape the window, screw around with the termcap entry, and reset the bottom-margin field of the chat.state to 0. Perhaps someday somebody will write a news or X.11 protocol for Chat. 4) Using it for spare parts. Certain hardware pieces of the 1186 might be useful. For example, the 360K floppy disk is valued at about $70 right now and would fit into any PC; The 80M disk is worth about $600, and would probably also fit into any PC, and work a lot faster than it does in Interlisp. The monitor would be useful if it could be attached to a Macintosh. The optical 3-button mouse is quite a bit nicer than any I've seen on the PC market, but the combination of the raw graycode output and the strange connector make it not worth bothering with, when a serial mouse can be had for $80. And the keyboard is rather nice, once you switch the EXPAND and LOCK keys by cutting traces and soldering in soft wires. But since it also has a strange connector and only one-way communications, it will not serve on a PC. (Is anybody homebrewing these days?) That's all I can think of. If anybody else has good ideas of what to do with an 1186, besides charitable donation, please post them. And if anybody has a more useful chat or full-screen PCE, let me know... Jordan Pollack
SCHMIDT@SUMEX-AIM.STANFORD.EDU (Christopher Schmidt) (05/10/88)
I had a suspicion that my $30K 1186 was obsolete when it was delivered
in November 1986. Confirmation came when my $2000 AT clone was much
faster. Further confirmation came when Xerox announced a 50% discount.
Now that Xerox has decided to move Interlisp-D into the Sun environment,
we are faced with the problem of what to do with our pretty worthless
Xerox hardware.
Since your measure of utility is based on speed and price (rather than,
say, quality of software, programming environment, accessibility and
extensibility of network architecture, virtual memory, or customer
support), you are indeed stuck with an inappropriate workstation.
[I'm sure you also discovered the hard way that d-machines far prefer
networked operation to standalone. And owning only one instance of
any kind of hardware makes for hard-to-diagnose problems.]
Since you paid way too much for your 1186 in the first place you never
had a chance of recovering your costs by resale. In particular, the
nominal value of Xerox Lisp--which can't be resold--was about $8000
at the time. (You should have paid more like $20,000 for an 1186, BTW;
$7000 unit cost for the hardware, $8000 for the software, and $5000
"overhead" for buying quantity one.)
If the absolute speed of an 1186 is no good to you, I think your best
chance of seeing it put to good use would be to find someone else on
your campus who is working on a problem that doesn't require so much
speed; eg. someone working on user-interfaces. Since you have only
one machine, your inheritor may not be able to justify the time in
learning such a unique piece of equipment, though.
Good luck-- Christopher
-------