Subject: Re: When to recompile/reeval?
From: Erik Naggum <erik@naggum.no>
Date: 17 Oct 2002 22:35:47 +0000
Newsgroups: comp.lang.lisp
Message-ID: <3243882947499328@naggum.no>

* Thomas F. Burdick
| I don't mind provocative /per se/, but what you were saying gives the
| impression that SBCL is willfully bad, as opposed to in development.

  If you could at least have recognized that it gave /you/ that impression,
  I could have expressed some regret at it, but when you pretend that it is
  some objective fact, I have to disagree very strongly.

| But lumping it in with willfully noncompliant systems for this reason,
| given it's version number, is inappropriate.

  No, /you/ think it is inappropriate.  As long as you have no room for the
  fact that I think it is appropriate as far as user expectations go, I only
  want to voice my opinion stronger to make you see my point of view, which
  you actually deny.

| People who want a better development environment should check back on
| SBCL later, to see if it's grown one, yet.

  What part of what I wrote could possibly be interpreted otherwise?  When
  I do not even close the lid on people who so richly deserve it, how could
  you think I could possibly have closed the lid on SBCL?  When I say it
  sucks, that is the observation there and then.  People change their mind
  and do something better, products get updated and improved with user
  input, and "it sucks" therefore means "if you do something about the
  things I have criticized, it would only suck if it sucks for more reasons",
  which is often a distinct propability, but you should at the very least
  acknowledge the possibility that it would cease to suck.  I really wonder
  what this "eternal damnation" thing is all about, because it is so very
  irrational and /primitive/.  It may have worked on the savannah where
  things stayed the same for the duration of an entire human life (all 25
  years of it), but we no longer live in a world where /anything/ can be
  expected to remain unchanged through your life (all 80+ years of it).

| I wouldn't imagine people who make heavy use of meta-object programming
| would check in on CLISP periodically to see if it had grown a full CLOS,
| because its lack thereof is a design decision in a mature product, not a
| missing extra-standard feature in a work-in-progress.

  When is the decision made not to include something in a future relase?
  This is not knowable.  Even the CLISP crowed could evolve to get rid of
  their annoying "we know better than the standard" attitude, for which
  they think I am being unfair towards them every time I bring up, but they
  keep insisting on being smarter than the standard.  But just like people
  who keep insisting on being smarter than the law occasionally wake up and
  want to obey it for no better reason than that it is there and that people
  expect them to before they want to deal with them, even they could wake
  up one day and exclaim "hey, we could gain something by conforming to the
  standard!".  However, far be it for me to impute intent to people -- for
  all I know, CLISP may be bogged down in implementation decisions that
  make it intractable to move closer to conformance without backing out a
  fair bit to use a different strategy that would take a lot of time to
  make as good as what they have now.  When you want to climb mountains,
  you usually have to cross valleys to climb the next higher one.

-- 
Erik Naggum, Oslo, Norway

Act from reason, and failure makes you rethink and study harder.
Act from faith, and failure makes you blame someone and push harder.