MathGroup Archive 2001

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

Search the Archive

Re: OOP

  • To: mathgroup at smc.vnet.net
  • Subject: [mg29087] Re: OOP
  • From: "Orestis Vantzos" <atelesforos at hotmail.com>
  • Date: Tue, 29 May 2001 02:57:17 -0400 (EDT)
  • Organization: National Technical University of Athens, Greece
  • References: <9ertnp$3jg@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

Yes, ofcourse I implement inheritance. Multiple inheritance is a tricky
thing to implement on the other hand, since many OOP operators are naturaly
implemented via recursion,in the sense that an operator of a particular
class calls the corresponding operator of its mother-class. This can only be
done if there is a well-defined (=>unique) mother-class. A friend did notice
nevertheless, that my code is compatible with the notion of
"friend-classes", namely classes that offer methods to a given object
without actualy being ancestors for that object's class. I will limit my
ambitions to that for now, as I am pressed to release a beta version of the
package in a reasonable amount of time. If things go well, I will seriously
consider multiple inheritance as a feature to be realised in a future
version...
Orestis Vantzos

"Hermann Schmitt" <schmitther at netcologne.de> wrote in message
news:9ertnp$3jg at smc.vnet.net...
>
> Hello Mr. Vantzos,
> do you also implement "inheritance"?
> I think it would be useful to implement multiple inheritance too, if
> possible. Most OO languages do only implement single inheritance as it
> is the case with Java.
> Hermann Schmitt
>




  • Prev by Date: User Functions with DisplayTogether
  • Next by Date: problem with graphics export from UNIX to NT
  • Previous by thread: Re: OOP
  • Next by thread: NonlinearFit work but NonlinearRegress, why?