Subject: Re: market research--what would be "cheap enough"?
From: Erik Naggum <erik@naggum.net>
Date: Mon, 10 Sep 2001 18:28:22 GMT
Newsgroups: comp.lang.lisp
Message-ID: <3209135302101586@naggum.net>

* Nels Tomlinson <tomlinso@purdue.edu>
> This kind of trouble doesn't exist with libre software.

  That is not true at all.  Software rot is real: When the environment
  changes, you still need to keep it running.  Your Osborne example would
  just as much out of luck with source as he is without.

> Odds are that someone will port linux and the bsd's to every new platform
> to come. along for a long time to come.  Given that, porting my favorite
> appication becomes thinkable.

  I doubt that.  Have you see how much changes from release to release in
  Linux?  Anyone who wrote a real application would have to work hard to
  stay on top of these changes.  Duane Rettig of Franz Inc has written
  several good articles about Allegro CL on Linux here about these problems.

> Black box methods just don't fit well with the scientific method.

  I do not think that statement fits well with the scientific method.

> To top it all off, we see the sort of mess that the city of Virginia
> beach got into; they lost some of their receipts for some of their
> software, and so had to pay for it again, and had to pay some large fines.

  From a single data point you can extrapolate in any direction.  People
  have lost the only sources to their software, too.  It is important to
  remember that accidents are just that.

> Overall, closed source software has a LOT of problems, and the fact that
> the libre stuff is often better is the least reason to prefer it, in the
> long run.

  If the free Lisp stuff was better, it would probably kill the commercial
  offerings, but the free offerings for Lisp are very, very poor compared
  to the commercial software.  This is why Franz Inc. can survive at all.

///