RE: [opengoop inheritance] Virtual Methods should be called Dynamic Methods

[Jim K. <ji...@ji...>]

Hey Dan,

Thanks for the comments.  Regarding termonology and definitions, there =
is a
lot of flux of ideas right now and things can certainly change.  After =
doing
a little review of my OOP text books, I agree that the terms "Abstract
Method/Class" do have strict OOP definitions and that it would probably =
be a
bad idea to use the word "Abstract" if we didn't adhere to the OOP
definition that it cannot also have its own implementation.  Maybe we =
should
decide exactly what OOP features we want to impliment and how, and then
decide what to call our constructs.  One problem that we will have is =
that,
since the LV compiler isn't enforcing the OOP constructs, we probably =
can't
check/enforce most things until run-time (like checking if a class that
inherits from a purely abstract class implements all of the abstract
methods).  The more constructs we have, the more work that the wizard =
will
have to do.  For example, if an abstract class is special, it might need =
its
own template.

-Jim

> -----Original Message-----
> From: ope...@li...=20
> [mailto:ope...@li...]=20
> On Behalf Of Dan Ponce
> Sent: Friday, November 07, 2003 2:49 PM
> To: ope...@li...
> Subject: RE: [opengoop inheritance] Virtual Methods should be=20
> called Dynamic Methods
>=20
>=20
> I think abstract is the wrong term.  The OpenG definition of=20
> a"Virtual Method" is "A Virtual Method is one which may be=20
> overrided by a descendant class. It is created from a special=20
> Virtual Method template, that has the ability to look for and=20
> dynamically execute descendent class implementations of the=20
> method, or else call its own implementation of the method. " =20
> It is clearly not an abstract method, at least as the term is=20
> used in Java. Abstract methods do not have implementations.
>=20
> I don't think it is confusing to use the term Virtual Method,=20
> as here it is used in the same way as in Endevo's GOOP Wizard=20
> 3.  If there is a need to use a different term, and if=20
> Dynamic isn't quite right, how about something along the=20
> lines of Overridable Base Method or Dynamically Overridable Method.
>=20
> On a slight different topic, I do think there needs to be an=20
> Abstract Method construct.  The way Endevo handled it seems=20
> to be an afterthought and has at least one flaw, according to=20
> their white paper.
>=20
> ----- Original Message ----- --__--__--
>=20
> Message: 2
> From: "Jim Kring" <ji...@ji...>
> To: <ope...@li...>
> Subject: RE: [opengoop inheritance] Virtual Methods should be=20
> called Dynamic Methods
> Date: Fri, 7 Nov 2003 00:52:58 -0800
> Reply-To: ope...@li...
>=20
> Niels and Bj=3DF8rnar,
>=20
> After listening to the arguments you both make, I agree that=20
> "abstract" might be the best term to use, rather than=20
> "dynamic".  Also, I think =3D that "prototype" has an OOP=20
> design pattern (Gang of Four) definition and "template" has=20
> an edit-time connotation, which make these two choices =3D not=20
> very attractive.
>=20
> -Jim
>=20
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.Net email sponsored by: ApacheCon 2003,
> 16-19 November in Las Vegas. Learn firsthand the latest=20
> developments in Apache, PHP, Perl, XML, Java, MySQL, WebDAV,=20
> and more! http://www.apachecon.com/=20
> _______________________________________________
> OpenGToolkit-Developers mailing list=20
> Ope...@li...
> https://lists.sourceforge.net/lists/listinfo/opengtoolkit-developers
>=20