Subject: Re: Lisp Jobs
From: Erik Naggum <clerik@naggum.no>
Date: 1998/08/21
Newsgroups: comp.lang.lisp
Message-ID: <3112721759359606@naggum.no>

* Mark Conlin <gt2863a@acmez.gatech.edu>
| Is it possiblt to get a job writing Lisp code?

  yes, but the Lisp code itself is not the real issue -- solving problems
  that are intractable or unaffordable without Lisp is one.  another is
  making functionally sophisticated solutions available for less money than
  the often graphically sophisticated but otherwise trivial solutions that
  the market is choked with.

| Do any of you write LISP code for a living?

  yes.  for the past two years, now, I haven't had to work in any of the
  languages _claimed_ to be the customer's choice, such as C++ or Perl.
  (I have instead replaced half-working solutions in those languages with
  fully functional solutions in Common Lisp.)

| Who do you work for?

  various parts of the financial markets in Oslo, from the Stock Exchange
  to a financial news agency.  I do servers and fundamental technology, not
  user interface stuff.  it would have been called "systems programming" a
  few years ago.  unlike what many appear to believe, traditional "systems
  programming" languages, such as C, C++, Perl, etc, are unsuited for their
  tasks, but it takes a change of focus on your problems to see why they
  actually _fail_ to deliver, despite their overwhelming "success stories".
  systems programming is about providing services to a host of applications,
  but as applications change, they pose new demands on the system, which
  it needs to adapt to provide.  when the application programmer instead
  has to turn to systems-level hackery to get a needed service, we're not
  talking about "powerful" languages because they can do this, we're
  talking about _dirt_poor_ languages that make you _have_ to in the first
  place.  Common Lisp provides a means to make very powerful abstractions
  available as services to people who really care about correctness and
  speed in the right places, and to adapt them quickly while maintaining
  the correctness.

  the markets where this is true are not the markets to which Microsoft is
  selling its shoddy software, and thus it takes a little effort to find it
  and also some experience to answer their cries for help, but only when
  you start to see what kind of problems _vanish_ when you program in
  Common Lisp, can you appreciate why people are still struggling with so
  many _unsolved_ problems elsewhere, employing ever more C++ programmers
  to keep the _real_ problems unsolved and cure only obvious symptoms.  as
  you can imagine, once you get people to believe that _only_ the symptoms
  can be cured and that the problems are fundamentally unsolvable, you will
  have a wonderful marketplace that never runs dry of the trivial problems
  you get more and more experience "solving" over and over again.  if you
  _really_ solve a problem, you won't need to solve it again.  there's no
  market in that.  in fact, there's unemployment in that for at least half
  the software industry, and although most of the people who would be out
  of work are young and smart enough to make a living doing something else,
  they will still resist you with everything they got, and invent problems
  that only they can solve with their particular tools and mindset.

  the funny part is that it isn't hard to find people who will listen, but
  you cannot talk to the masses.  it's got to be a few people at a time,
  because this is about losing one's religion -- the false belief that no
  software can ever satisfy the customer's demands, the ever increasing
  false belief in promises of the next version ("life after death", "for
  our children's children", etc), and the ever increasing attitude that the
  last version was a piece of shit.  losing that religion is hard for many
  people -- it'a too easy to believe that what you promise those who would
  shuffle off the upgrade coil is just another religion, but if you get
  really good software, you don't need to upgrade, and the software stays
  current and valuable for years to come.  Common Lisp is like that.  you
  just have to find people who don't _expect_ to throw your software out
  mere months after they got it.  that's who Lisp programmers work for.


  incidentally, I believe Microsoft has a world-wide campaign these days
  (at least they have a campaign in Norway that looks like those dubbed
  commercials for shampoo and toothpaste and whatever that fit Europe in
  general but no country in particular) in which we are told that "Windows
  98" is a medicine against headaches.  _which_ headache?  wasn't it sold
  to the target audience by Microsoft itself under the name "Windows 95"?

#:Erik
-- 
  http://www.naggum.no/spam.html is about my spam protection scheme and how
  to guarantee that you reach me.  in brief: if you reply to a news article
  of mine, be sure to include an In-Reply-To or References header with the
  message-ID of that message in it.  otherwise, you need to read that page.