Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.misc > #3692

Re: [CM] Two interesting articles on Symbolics/Lisp

Cross-posted to 2 groups.

Show all headers | View raw

On 6/15/2014 6:03 AM, Robert Swindells wrote:
> On Sun, 15 Jun 2014 10:36:19 +0000, RS Wood wrote:
>
>> On 2014-06-13, Kaz Kylheku <kaz@kylheku.com> wrote:
>>> On 2014-06-13, Paul Rubin <no.email@nospam.invalid> wrote:
>>>> RS Wood <rsw@therandymon.com> writes:
>>>>> Ergonomics of the Symoblics Lisp machine - reflections on the
>>>>> developer productivity
>>>>> http://lispm.de/symbolics-lisp-machine-ergonomics
>>>>
>>>> That article was interesting but I don't understand why it said
>>>> "Symbolics Lisp Machine" in almost every sentence.
>>>
>>> Search engine spam. It wants to be high ranked in search engine indexes
>>> for that keyword.
>>
>> A reasonable guess, but there can't be a huge number of webcrawlers out
>> there desperately looking for that keyword combination.  If they're
>> hoping to sell ads, they've not chosen a topic that will make them rich.
>> Could just be a matter of sloppy writing, too.
>
> I don't think it is search engine spam, the hosting site is a fairly low
> power machine with no ads.
>
> I don't see anything that is incorrect in the article, stuff that is
> specific to Symbolics is identified clearly.
>

I think the key issues of the time are somewhat misconstrued.  The
market in the early 1980s included a host of Motorola 68000-based
Unix workstations at a price point of about $10,000.  (There were
some other specialty CPUs as well, e.g. Xerox, VAX, CMU). Not perhaps
as nice as a Symbolics system but good enough for many
programmers, even Lisp programmers, to work comfortably. So-called
3-M machines ... 1 megapixel, 1 megahertz, 1 megabyte proliferated.

Of course now such machines would be considered terribly weak and
$10k would be considered outrageous,  but that was then.

Consider, how many users of Symbolics machines really took advantage
of the fact that the Lisp source code of the operating system was
hackable? Maybe it was overkill and consequently overpriced? The
Symbolics-and-no-substitutes acceptable market was, I thought, primarily 
DARPA contractors. Sort of like the Connection Machine
business, I suppose.

 From a business perspective, one might seriously consider the advantage 
of programmers writing an application based on
Lisp (or C or ...) running on a Sun-3 which you knew
could then be distributed on other Sun-3  $10k systems to
your customers. And you might choose C over Lisp because of any
number of reasons (not worth arguing about here).  Using a
Symbolics machine for C was presumably possible but not very
attractive.

While Symbolics eventually got to the point of
making/selling lower-cost application delivery systems,
maybe it was already too late.

Note that one Lisp program, considered important at the time,
Macsyma, was purchased from MIT by Symbolics, but Symbolics was loathe
to distribute it on anything but Symbolics machines.  Did they
make a strategic mistake by not selling the software on those
other machines? (Especially VAX) Until too late..

I agree with the notion implicit in the cited remarks that
Symbolics provided a productive environment, and many subsequent
systems (e.g. IBM PC, MS-DOS, Windows) represented huge backward
steps. But Symbolics' user interface was not without rivals from
Sun, Bell Labs (BLT), Xerox, Apollo, DEC, Tektronix, IBM (RT), HP, 
Apple(Lisa), CMU and probably others
in that time frame.  All of them ran Lisp. Most of them could run
Macsyma, starting circa 1979.  I suspect that it became
increasingly difficult to argue that Symbolics' "special sauce" was
so special or even was the right sauce (cf. various AI winters)
for what people thought they needed.

Richard Fateman

Back to comp.misc | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[CM] Two interesting articles on Symbolics/Lisp RS Wood <rsw@therandymon.com> - 2014-06-13 11:37 +0000
  Re: [CM] Two interesting articles on Symbolics/Lisp Paul Rubin <no.email@nospam.invalid> - 2014-06-13 09:44 -0700
    Re: [CM] Two interesting articles on Symbolics/Lisp Huge <Huge@nowhere.much.invalid> - 2014-06-13 19:06 +0000
    Re: [CM] Two interesting articles on Symbolics/Lisp Kaz Kylheku <kaz@kylheku.com> - 2014-06-13 21:03 +0000
      Re: [CM] Two interesting articles on Symbolics/Lisp RS Wood <rsw@therandymon.com> - 2014-06-15 10:36 +0000
        Re: [CM] Two interesting articles on Symbolics/Lisp Robert Swindells <rjs@fdy2.co.uk> - 2014-06-15 13:03 +0000
          Re: [CM] Two interesting articles on Symbolics/Lisp Richard Fateman <fateman@cs.berkeley.edu> - 2014-06-15 16:11 -0700
            Re: [CM] Two interesting articles on Symbolics/Lisp Paul Rubin <no.email@nospam.invalid> - 2014-06-15 16:34 -0700
              Re: [CM] Two interesting articles on Symbolics/Lisp Robert Swindells <rjs@fdy2.co.uk> - 2014-06-16 08:48 +0000

csiph-web