[Biococoa-dev] why placeholder classes was: BCSequence class cluster

[Koen van der Drift]

On Jan 9, 2005, at 12:51 AM, Charles PARNOT wrote:

>> On Jan 8, 2005, at 8:20 AM, Alexander Griekspoor wrote:
>>
>>>> Again, the reason why I came up with the idea of some public 
>>>> headers for placeholder classes for typed sequences was to propose 
>>>> the user BOTH OPTIONS! (but maybe we should not).
>>> That's perhaps the most important question we have to answer first 
>>> indeed.
>>>
>>
>> I still don't understand the construction with these placeholder 
>> classes. Why do we need them, next to the subclasses we already have? 
>> (this is an educational question, not a critical one ;-)
>>
>> - Koen.
>
> This only applies to the class cluster design. Because the public 
> superclass declares all the methods as valid, then all the subclasses 
> are known by the compiler to also respond to all the methods.
>  So we can't just use the headers of the subclass. We have to provide 
> separate headers with a restricted set of methods.

Why a restricted set f headers?


> But alloc-ing these objects will actually return objects of the class 
> cluster (hence 'placeholder' classes). The compiler is fooled. The 
> runtime gets what is needed.
> I hope the shortness makes it clearer. I can make a longer version if 
> you want ;-)


Please do, because it is still very confusing to me.


- Koen.