Re: From Plato to Nato: "Platonism" and programming

From: Edward G. Nilges (spinoza1111_at_yahoo.com)
Date: 06/08/04 

Date: 7 Jun 2004 22:45:55 -0700

Mike <m.fee@nospam.irl.cri.nz> wrote in message news:<MPG.1b2fead65225388b9896c5@newshost.comnet.co.nz>...
> In article
> <f5dda427.0406050148.11b56eab@posting.google.com>,
> spinoza1111@yahoo.com says...
> > Programmer Dude <Chris@Sonnack.com> wrote in message news:<qq70c0pbmgrkr5c9kcar5padacj813dpfm@4ax.com>...
> > > Edward G. Nilges writes:
> SNIP
> > > > Actual programs have in fact many actual and potential bugs when
> > > > considered as a class of solutions but the mental picture is that
> > > > they must be bug free.
> > >
> > > Stop right there. You're so close to making sense, but let's think
> > > about this a second. Remember that we AGREE Platonism is a belief
> > > in an *unreachable* World of Forms.
> > >
> > > Are you saying (any/some/most) programmers believe their programs can
> > > NEVER be perfect?
> > >
> > The best would say this.
> >
> > > Consider a simple form: a circle. No perfect circles in this world.
> > > Consider a simple form: Hello, World. Many perfect instances extant.
> >
> > Nope. If I want a Hello, World program that works, one that fails to
> > use a GPS sensor to detect whether it is running on a computer in
> > France or China, and says bonjour, monde, or ni hao, all-under-heaven,
> > using in the latter case ideograms, then I claim the program is not
> > "perfect".
>
> So in this, are you suggesting that for any program to
> be perfect it must meet _your_ personal expectations? By
> the same reasoning I could claim that a functioning
> version of your GPS_inclusive, construct still fails to
> be perfect because, although it prints a culturally
> suitable "Kia orana" to me, it fails to determine
> whether I am literate, sighted and/or looking in the
> right direction at the time. And when, after these
> problems are fixed, if it can't tell me the best way to
> get to the Maniopoto river, or can't fill in a tax
> return for me, or can't define existentialism without
> referring to tertiary sources, I can still claim it is
> imperfect. Seeems to me that you don't want a programme;
> you want a god. Surely a perfect program can be simply
> defined as one that exactly and precisely meets and
> matches its specifications. If that is the case, then,
> for a limited range of specifications, perfect programs
> do exist.

 Excellent insights: but, you fail to see what the phrase "for a
limited range of expectations" does to "perfect program". It cancels
the program and makes it imperfect.

Didn't Socrates reject limited perfection of artisanship at the
beginning of the Republic? His search for "the good" or "justice"
admits that an humble artisan, whether a carpenter or programmer, can
be good-at or even perfect-at, or a programmer's program can be
"perfect for a limited range" but then Socrates, in of course a
Platonic gesture, rejects limited "perfection" as not "perfection" or
even "justice" at all.

The idea that any given implementation of "hello world" can be
"perfect" as opposed to merely good and thus prove the thesis that
Forms exist inside computers is based on the vision of language that
Wittgenstein deconstructs in Philosophical Investigations: for in fact
if language is a form of life, an hello world program would have to be
prepared to say ni hao or to give the Heimlich maneuver.

The only Platonic, perfect, program, would be a single program that
solved all problems...as you say, God 1.0.

This simply reduces the claim that ANY program is bug free to
absurdity, of course.

Relevant Pages

  • Re: Note to all the world.
    ... of your ranting, since both SR and LET only ... have a limited range of applicability. ... falls apart as soon as there is gravity. ... LET fails as SR fails when grvity is ...
    (sci.physics.relativity)
  • Re: Taking Ownership
    ... > have been able to determine is that the user id has the appropriate rights ... ostensibly on the language itself. ... When a function fails (and only if it ... fails so you need to check return codes or look at the EAX register if you ...
    (microsoft.public.dotnet.languages.vc)
  • Re: Need Help Creating an Image Slideshow
    ... ;' fails? ... is not part of a programming language ... You know I am an advocate of feature detection - I'm using it all the ... time - but not in absurd cases like this. ...
    (comp.lang.javascript)
  • Re: Taking Ownership
    ... I have indeed been checking my return codes. ... The only one that fails is the ... > ostensibly on the language itself. ...
    (microsoft.public.dotnet.languages.vc)
  • Re: gcc failure
    ... FORTRAN is ... > an old language, but there is still a lot of scientific software written ... Often fails for one reason on another. ... At one time it was unable to bootstrap itself. ...
    (comp.unix.solaris)