dotgnu-general
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [DotGNU]C@


From: Peter Minten
Subject: Re: [DotGNU]C@
Date: Sun, 11 Aug 2002 10:28:58 +0200

BioChem333 wrote:
> 
> On Sat, 2002-08-10 at 12:03, Peter Minten wrote:
> > C@, pronounced cat
> 
> I still prefer C_ (C flat).

The flat symbol doesn't really give an impression of what the language is IMHO.
The at symbol is related to the net and so people who see C@ will think it's a
net language, and it is.

> What about private inheritance? Inheriting implementation without
> inheriting interface can be quite useful, and avoids most of the
> problems of ambiguity with multiple inheritance. The order in which the
> classes/interfaces are listed could be used to determine which classes
> have priority. The first class listed is public, and any classes listed
> after it are privately inherited. In the case where a class isn't
> supposed to publicly inherit from any other classes, it must still be an
> Object, so Object could be explicitly listed as the first class (eg.
> SomeClass : Object, PrivateClass). Additionally, base could still refer
> to the first class, but base[SomePrivateClass] could be used to access
> privately inherited base classes.

Private inheritance is a good idea, but the syntax you suggest is very unclear
to most people. I have to think a bit on how to integrate private inheritance in
the language, but it will almost certainly involve using a keyword to make a
clear distinction between public and private inheritances.
 
> Another feature to add to the list: friend methods and classes. The
> internal access modifier is too broad for most situations. Limiting the
> access to a particular class or method also limits the potential for
> problems with changing implementations.

Good ideas, now we just need to figure out how to weave them into the language.

Greetings,

Peter



reply via email to

[Prev in Thread] Current Thread [Next in Thread]