Subject: Re: realistic but short and simple LISP examples?
From: Erik Naggum <erik@naggum.net>
Date: Wed, 19 Dec 2001 08:58:08 GMT
Newsgroups: comp.lang.lisp
Message-ID: <3217741085383991@naggum.net>

* Bruce Hoult <bruce@hoult.org>
| I'm sick to death of incredibly nifty things disappearing along with the 
| companies that created them.  Or, often worse, with the company 
| continuing and just the project or product getting cancelled and being 
| sat on, presumably to prevent competitors from getting a leg up.

  What this actually means is that too few people want to pay enough for
  those features to even warrant keeping them listed as products, much less
  pay for maintaining them or pay someone who knows them well enough to
  offer any sort of customer questions, and no competitor thinks it is
  worth paying the original developer for the right to the source.

  As you may be aware, art museums are not just _given_ the artwork they
  display for free, but go out and actively participate in auctions and
  have annual budgets for acquisitions.  Other museums display things they
  funded, or they are part of universities and their continuing obligation
  to present knowledge to the public.  This is stuff that very few people
  would want to purchase for themselves.  Vast numbers of people prefer to
  chip in using taxes and admission fees to be able to study these things.

  I think there is time for software museums.  There are already several
  web sites that effectively amount to being museums, such as the DEC-10/20
  site, which also houses an emulator so the old software can be run on
  modern machines.  The biggest challenge faced by a software collector is
  to keep it running.  Modern machines are fast enough to emulate most of
  the "ancient" hardware, even to the point where they can run operating
  systems on the emulator.  However, one major reason software rots is that
  the rest of the world moves on while it does not.  Operating systems
  change incompatibly, cooperating software changes, library versions
  change, etc, etc.  Making sure that some "ancient" software keeps running
  is no small feat.  Unless you get people to pay for this maintenance,
  whatever values it once had will depreciate to zero very quickly.

  Just having "source" around is no guarantee of anything at all.  People
  do not read source code.  If they did, they would not use languages like
  C++ and Perl.  Source to an application is not enough -- the rest of the
  system with which it interacts is also necessary to understand it.  You
  can ask running code questions and expect answers.  Forensic medicine is
  based on extensive experience with an unchanging basis, but forensic
  computer science is based on having a running computer around to study
  the dead program.  Perhaps the software museum should be called The Code
  Morgue.

///
-- 
  The past is not more important than the future, despite what your culture
  has taught you.  Your future observations, conclusions, and beliefs are
  more important to you than those in your past ever will be.  The world is
  changing so fast the balance between the past and the future has shifted.