home *** CD-ROM | disk | FTP | other *** search
Text File | 1989-10-12 | 1.6 KB | 39 lines | [TEXT/GEOL] |
- Item 9442518 12-Oct-89 00:09
-
- From: MUYSVASOVIC1 ER&D - J-D Muys-Vasovic
-
- To: D2086 Efficient Field Svc, C Faith,PRT
-
- cc: MACAPP.TECH$ MACAPP Tech
-
- Sub: Re: Changing MacApp Source
-
- Curtis,
-
- Your are right: we would like to be able to change the behavior of whatever
- class in the hierarchy without the problems you raise. However, this problem
- does not belong to MacApp, but more to any object-oriented approach, with an
- directed acyclic inheritance graph for data types. This is so in Smalltalk for
- example.
-
- In fact this is an instance of a more general class of problems: you have some
- working code that you want to customize for your needs. Without OOP, it's close
- to impossible. OOP introduced a clean way to do that: subclassing and
- overriding. Of course, this rely on the fact that the implementer of the class
- you subclass divided its behavior into methods with sufficiently small
- granularity. In fact the smaller the granularity, the more customizable the
- code. But then you reach an efficiency problem. That's a good example of trade
- off. That reminds me of a joke about specialists: a specialist is a person who
- knows a lot about a small part of human knowledge. The ultimate specialist is
- therefore a person who knows everything about nothing.
-
- The real life solution to the problem, introduced by Smalltalk and taken over
- by MacApp, is to modify the source code in some (undefined) circumstances, with
- all the associated problems.
-
- Such is life.
-
- Jean-Denis Muys-Vasovic.
-
-
-