MathGroup Archive 2002

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Re: OO in Mathematica

  • To: mathgroup at smc.vnet.net
  • Subject: [mg37503] Re: [mg37473] Re: [mg37458] OO in Mathematica
  • From: "Hermann Schmitt" <schmitther at netcologne.de>
  • Date: Sat, 2 Nov 2002 03:30:21 -0500 (EST)
  • References: <200211010642.BAA11239@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

Mathematica is primarily a system for doing interactive computations, OO is
not suitable for those cases, at least directly, but classes could be a
substitute for packages. If you store above menhtioned computions in order
to avoid reenter, this is not a case for OO, either.
But Mathematica is - in my eyes - one of the best programming languages and
also applicable beyond Mathematical computations. There are no facilities
for building GUI's and working with databases, but J/Link makes it possible
to use these facilities in Java. It makes not much sense that invent the
wheel again, for theses facilities.
As I consider Mathematica a very good programming language, it makes sense
to program larger applications in Marthematica, and then OO comes into the
play.
Hermann Schmitt

----- Original Message -----
From: "Andrzej Kozlowski" <andrzej at tuins.ac.jp>
To: mathgroup at smc.vnet.net
Subject: [mg37503] [mg37473] Re: [mg37458] OO in Mathematica


> But you are yourself describing precisely this "contrast"!  Moreover,
> the "contrast" between function oriented programming and object
> oriented programming was not invented on this list but is  well known
> and quite often discussed. See for example the discussion in Part II of
> "Mastering Mathematica" by J.W. Grey, particularly the section entitled
> "The Duality between Functions and Data", or various discussions in
> books on CLOS (object oriented common Lisp). What you are saying is in
> fact exactly the approach taken in CLOS. But it is also pretty sure
> that not everybody using Mathematica  needs the kind of data and
> program structuring that object oriented programming provides, in fact
> I am pretty sure  the majority of Mathematica users do not (the
> majority of users do not even need the existing "package" mechanism).
> It does not mean of course you should not develop your OOP package; in
> fact I am quite keen to try it when it becomes available.
>
> Andrzej Kozlowski
>
>
>
> On Thursday, October 31, 2002, at 06:40 PM, Hermann Schmitt wrote:
>
> > Hello,
> > There is a tendency to see a contrast between OO and Mathematica. In my
> > opinion this makes no sense:
> > OO descibes a method for structuring programs and the data related to
> > the
> > programs. Mathemtica describes the instructions with which programs
> > may be
> > built.
> > Hermann Schmitt
> >
> >
> >
> >
> >
>
>



  • Prev by Date: Re: When does Integrate give incorrect results?
  • Next by Date: Fwd: Re: Direct tensor algebra
  • Previous by thread: Re: OO in Mathematica
  • Next by thread: Re: OO in Mathematica