Re: Lisp considered unfinished / marketing strategy

[reti@cambridge.apple.com (Kalman Reti)]

At 16:24 6/5/95, kr wrote:
>At 19:40 6/4/95, Ken Anderson wrote:
>>   These issues have been untreated emergencies for some years now.
>>
>>It is easy to say "we did that 10 years ago in Lisp".  However, it doesn't
>>mean much if we can't deliver it to the people who want it now.  We need to
>>collect and focus our capabilities so we can.
>
>I suspect that this is largely due to a marketing issue. As LISP was
>(correctly) perceived as a sophisticated environment, it was traditionally
>sold at price points designed to appeal to customers with tasks so complex,
>that they knew they could not do with anything less than LISP. So this
>language has always been confined to a fairly elitist circle of users (many
>even lavishly funded by the DoD :-).

I suspect being perceived as sophisticated is a hindrance, not a help, in
mass acceptance.

>
>This policy risks being run over at some point by (lower quality) mass
>market products, just by virtue of sheer volume (see Microsoft, etc.)
>
>The only way to counter this, is by changing the strategy and to start
>mass-marketing Common LISP too. This may feel unappealing at first, but
>will likely win out in the long run. The only way to push a language is to
>make sure that it will have a huge user base. If Common LISPs were sold at
>$100 (instead of several times more), so that many people would buy it
>without even thinking much about it (and then getting addicted only later
>on), the potential market could easily be ten times larger than today,
>which still would lead to a higher total profit despite the lower price
>point.

[turn on soapbox mode]

I agree, but I think a language is too abstract to mass market.  I believe you can
only mass market something that solves a real need.  My suggestion (originally
to Symbolics, then to Apple re MCL and more recently to Apple re Dylan)
has always been to provide a mid-level tutorial (book) + basic toolkit that
shows in a cookbook, step by step fashion, how to create custom applications
useful in the business arena, e.g. 1) how to programmatically collect information
from online/network/disk/tape sources, process it a little to find interesting/
relevant/meaningful issues, and present it in a readable/graspable way and 2)
how to easily/quickly set up application-specific client/server networks and
distributed processes. 

I base this suggestion on having personally seen just how much hay several
average programmer types (by no means Lisp or AI wizards) with lisp machines and the right training/coaching made in their organizations by being able to respond to ad-hoc requests for information and analyses from top management
in hours to days (rather than in the days to weeks which the traditional MIS departments would take to do the same thing) and by being able to quickly
construct specialized client/server systems giving upper management instant
access to very specific information needed for decision-making.  In most of
these cases, abilities which only a Lisp-like dynamic language has (being able
to patch running applications while they are still running, sophisticated
condition handling enabling round-the-clock applications to stay up
in the face of unexpected failures, automatic collection of useful data when
such failures occurred [e.g. annotated backtraces sent as mail, with DESCRIBEs
of key data structures included]) were crucial to the success of the endeavor.

Unfortunately, none of this is "sexy" or terribly exciting, it is mostly bit
pushing and formatting and a lot of mundane stuff, which is one of the
reasons I suspect it hasn't been done yet.   It also requires really having
a fairly stable and featureful set of tools (programmatic terminal emulators,
access to various network protocols, etc.) out of which these types of
applications can be built.   (I've participated in several instances of building
such tools, but they were all under consulting contracts and hence the property
of the contractors.)

However, IF you could build such a toolkit and IF you could market a readable
book which showed how you could put them together guickly to get really useful results,  I think you'd have a crack at reviving interest in Lisp.

[turn off soapbox mode]


>
>But the more important, indirect effect will the larger user base, which is
>an advantage that grows over time.
>
>
>Greetings
>Markus Krummenacker