Re: Trivia: Processor MHz

["srichter " <srichter@xxxxxxxxxxxxxxxxxx>]

>From: Chris Rehm <javadisciple@earthlink.net>

>Which has nothing to do with the issue. I didn't say to not use OO
>techniques. I said to implement the object methods using MI calls rather
>than CPU brute force. I only suggested that because you indicated that
>this new "modern" approach was less efficient than the old approach.

Have who implement the method using MI?  The compiler ?  If so, how does the 
compiler know to do this ?  Or code MI within the "copy" method ? If so, than 
you still have the overhead of a function call to reach the efficient machine 
code.


>What? Are you saying that OOP methods are dependent on being implemented
>in some form? Isn't that exactly the opposite of OOP theory? So if I
>went through the standard libraries and replaced string object methods
>with methods that used the MI calls, then C++ wouldn't work any more?

The standard library is more C than C++. In C you say "strcpy( ToString, 
FromString)". In C++ you say "ToString = FromString" where the "=" operator is 
overloaded by the string class. So you are back to an inefficient function call 
to reach your optimized method.  Result being an inefficient ( on a CPW 50 
system ) operation.  You are mixing OOP and RPG like coding, ending up with 
code that is not efficient and I would argue slightly not OOP.


Steve Richter