Re: methods on make

• To: <info-dylan@ai.mit.edu>
• Subject: Re: methods on make
• From: Scott Ribe <scott_ribe@killerbytes.com>
• Date: Tue, 8 May 2001 17:57:39 -0600
• In-Reply-To: <3AF877AE.2BA9DE4@oracle.com>
• References: <3AF877AE.2BA9DE4@oracle.com>
• Sender: owner-info-dylan@ai.mit.edu

On Tue, May 8, 2001, Roman Budzianowski <Roman.Budzianowski@oracle.com> wrote:

>What problem are you trying to solve ? If we assume that only one
>implementation
>can be active at one time (a reasonable assumption) the just reloading
the new
>method will replace it, thus replacing an implementation.

No, I'm not assuming that at all; that's the point, that there can be
multiple concrete classes and that make on the abstract base classes can
choose between them at runtime. Think various "factory patterns".


Scott Ribe
scott_ribe@killerbytes.com
http://www.killerbytes.com/
(303) 665-7007 voice

• Re: methods on make
• From: Gabor Greif <gabor68@yahoo.com>

• Re: methods on make
• From: Roman Budzianowski <Roman.Budzianowski@oracle.com>

• Prev: Re: Shared libraries and ObjC on MacOS X
• Next: Re: methods on make
• Index(es):
  • Main
  • Thread