CLIM mail archive
[Prev][Next][Index][Thread]

Re: The cost of CLIM and its future

In message entitled "The cost of CLIM and its future" on Sep 16, 
Scott McKay writes:

> Could somebody suggest a way that us Lisp companies get to stay in
> business?

At LUV-92, during the "CLOS vs. C++" Panel, one of the points that 
struck me was that the "bundledness" of LISP, although better for
us users, was much worse for the market. The argument was that with
C or C++, 12 different vendors might supply libraries that you use
with your compiler from a 13th vendor. This gets a lot of companies 
into the business, taking out lots of ads, and all promoting C usage.
In LISP, the same vendor supplies all the math libraries, object-oriented
stuff, formatting libraries, error trapping, etc., etc., because it all
is part of the language. This makes being a LISP vendor an all or nothing
proposition, and few people want to make such a big plunge.

Perhaps CLIM is more of the same: another large system that takes a big
investment, and that only the vendor who sold you the LISP can provide.
Now, there are plenty of counterexamples around, where very large, high 
investment systems (eg UNIX) have caught on. But I still wonder if there
is any way to provide a role for small LISP addons, rather than only the
people who can provide a big LISP plus an editor plus a development 
environment plus CLIM. Perhaps CMU/CL will become more successful and start
giving their sources away for free/cheap to commercial developers, so that
there will be plenty of good baseline compilers. Then it will be a lower
investment to get into the market, and vendors may try it for relatively
small products as well.

I don't know the answers, and certainly I greatly prefer a more powerful
language with everything built in, and want CLIM to become widely accepted.
My company is still willing to invest in LISP and CLIM (from multiple
vendors) because I can convince them on the return on their investment.
Perhaps we've ridden out the worst of the storm, and machines are finally
fast enough that managers will be willing to do what we've claimed they
would for a long time, and trade initial investment and usually some
runtime speed for power and programmer productivity. But I'm not so sure...

						- Marty
(proclaim '(inline skates))


0,,
Follow-Ups:
Main Index | Thread Index