From: ···@gfa-genesis.com
Subject: Real World Common Lisp Development and Developer Position
Date:
Message-ID: <uaflx8pf9.fsf@gfa-genesis.com>
Considering the type of discussion that has been raging
here on why Common Lisp is not used more in the market
place, I think I may be forgiven for posting the
position announcement below.
From my view, as a software developer, the factors that
have made Common Lisp a viable alternative today are:
1. Available on Windows NT
2. Interactive GUI builder
3. Interface with database, like SQL via ODBC
4. Foreign language interface, to C at least, for those
numerically intensive functions.
5. Deliverable runtime system.
6. Support and training material, such as books and
on-line documentation.
I submit that these are minimum requirements for most
"real world" applications. The commercial systems
available from Franz and Harlequin now provide 1 - 5 (to
different degree, but enough) and there are enough books
that 6 is satisified.
Personally, I think Standard ML is a beautiful language
(at least semantically, syntactically I must confess I
prefer Lisp syntax), but all the implementations so far
lack some of the above requirements. For the windows
platform, the announcement by Harlequin of their Dylan
system is exciting, but is not here yet.
So, for now, I think Common Lisp is the best development
environment.
-- Young-il Choo
==========================================================
Common Lisp Developer Position
Participate in the design and development of the Company's
core software product in financial analytic tools.
The main development will be in Common Lisp, though, more
important than knowledge of any particular language or
system, we are looking for persons who are able to rapidly
become proficient in any programming language, have solid
computer science background and strong mathematical
aptitude, and interested in computational aspects of
finance.
Knowledge of data base systems, quantitative financial
methods and numerical methods for PDE's definitely useful.
The company, Gifford Fong Associates, is located in
Lafayette, CA, 15 miles east of San Francisco in the East
Bay.
Please send your resume (in plain text) to
Young-il Choo at ···@gfa-genesis.com.
==========================================================
···@gfa-genesis.com wrote in article <·············@gfa-genesis.com>...
> From my view, as a software developer, the factors that
> have made Common Lisp a viable alternative today are ...
You left out one item. The simple fact that the standard
computer purchased by just about anybody from
your mother to the largest commercial enterprises
has at least a 100mHZ 32 bit processor with
a billion bytes of hard disk, and from 16 to 32
megabytes of random access memory and comes
bundled with a operating system which is not
antagonistic to running lisp.
Managers believe that Intel will survive, they believe Microsoft
is good. Therefore they have little reason to reject
a mature software environment like a commercially
supported Common Lisp because "those guys who hack
in C/Unix" have a time-to-market advantage when some
new processor is being pushed by some hardware company.
How strong is the "it must run in Pure Java" movement?
Frankly I've been shocked to here "Pure Java" advertised
on local Talk Radio in the Boston area.
When is the last time you heard an advertisement for
a programming language on the radio?
In article <·············@gfa-genesis.com>, ···@gfa-genesis.com wrote:
I'm playing devil's advocate.
> From my view, as a software developer, the factors that
> have made Common Lisp a viable alternative today are:
>
> 1. Available on Windows NT
If you're into NT. Others use Macs, Unix, ... CL is available there.
> 2. Interactive GUI builder
Nothing really cross platform / cross Lisp available. Nothing
ommercially available that really does impress me.
> 3. Interface with database, like SQL via ODBC
Available - but not very good supported. Small user base.
> 4. Foreign language interface, to C at least, for those
> numerically intensive functions.
No standard in sight.
> 5. Deliverable runtime system.
Still to large. Implementations deal very differently with that.
> 6. Support and training material, such as books and
> on-line documentation.
No (good) book about real-world programming (graphics, database,
web, games, ...) in Lisp has been published in the last five years.
Few books about CL have been published in the last five years
(less then ten?).
> So, for now, I think Common Lisp is the best development
> environment.
The competion is SmallTalk - widely used in business environments.
Libraries and third party products are available.
Sadly CL (which I also use due to lack of alternatives - and none
in sight) has a lot to improve.
I think users of Lisp/CL have to do more to ensure its success.
Writing real world code and talking about that is a good way.
Don't blame anybody else for the lack of progress in the last five years.
Better education, better documentation, better books, better examples,
dedication to real world delivery, etc. How can one company
release a Lisp on NT without actually having a plan to support
networking (TCP/IP)? In which world do they live? How can
it be that MCL easily beats PC-based Lisps? This is where *I*
would expect to be a market.
It's about time to do mainstream development with Lisp. Don't waste
your time trying to reinvent the wheel.
Support the LispOS on PC approach. Make people interested in writing
*applications*.
--
http://www.lavielle.com/~joswig/