Ron at USMedRec.com
Mon Nov 14 20:58:34 UTC 2005
I'd buy that, but since new is implemented in both the superclass (object)
and its subclass (ArrayedCollection), and since the class is an abstract
superclass and shouldn't respond to new and doesn't currently implement it
and since it also solves the problem, I'm not sure I would characterize it
as outright wrong. Another benefit of implementing new on an
abstractSuperclass is that it gives someone a clue that it is abstract. We
implemented an abstractSuperclass method that was like subclassResonsiblity.
Does squeak have that?
From: Andreas Raab [mailto:andreas.raab at gmx.de]
Sent: Monday, November 14, 2005 12:22 PM
To: Ron at USMedRec.com; The general-purpose Squeak developers list
Subject: Re: Squeaking trouble
Ron Teitelbaum wrote:
> We should fix this problem with something like overriding new to return
That would be outright wrong. It is not #new that is a subclass
responsibility it is #size.
More information about the Squeak-dev