[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [oleg@xxxxxxxxx: Interface view of dictionaries]



Bear wrote:
>> It argues for the implementation of those operations *EVEN ON
>> DICTIONARIES WHERE THEY'RE NOT PARTICULARLY EFFICIENT*.  That was, in
>> fact, my whole point.  Only if the operations exist on all
>> dictionaries will they be used in "generic" code.  Only if they are
>> used in "generic" code will the benefits, where available, be
>> realized in general systems.

scgmille@xxxxxxxxxxxxxxxxxx wrote:
> I see your point entirely.  My main argument is against operations
> that don't make any sense for some collections.  For ones which do
> generalize but are implementable in terms of the more basic
> operations, there is little point *except* efficiency.

That's not a "little" point.

> In that case, the operators should be defined over a superset of
> collections where they're both defined and can be implemented more
> efficiently.

Which ties users to implementation details of the containers they use,
even when that isn't necessary. You make a big deal about constraining
collection implementors to specific implementations, but you don't think
it's a big deal to do it to collection users?

> For the general classes of SRFI, neither point holds.  

That's because you are, in fact, missing the point.
-- 
Bradd W. Szonye
http://www.szonye.com/bradd