Re: Real OO

[Bob Hathaway <rjh@geodesic.com>]

In article <4l92n3$1g6@gaia.ns.utk.edu>,
Matt Kennel <kennel@msr.epm.ornl.gov> wrote:
>>...
>...
>I'm an Eiffel fan but I'd disagree here.  "polymorphism" is not like alms
>for the poor.  Per se more 'polymorphism' doesn't mean 'better', it just 
>means 'more polymorphism'. 

I have to disagree.  Having used multiple-polymorphism in a class-based
language I have found it to be far more powerful.

I haven't been following this thread and its been a long time since I've
programmed in Ada (before 9x), so could some kind sole remind me if
class wide types are essentially true types, meaning that the true types
of arguments are not lost, for future calls for instance.  I seem to
recall class wide types are not lost and multiple-polymorphism is provided
in this case, at least for the class-wide parameter types.

Take a look at my quick but hopefully convincing argument for a true type
system and multiple-polymorphism in http://www.sigs.com/objectcurrents,
issue 3 (I think) on True Types And Multiple Polymorphism as it presents
my argument on this issue.  I see single-polymorphism as a static efficiency
measure or for an approach to enforcing strong static typechecking (although
it's certainly not the only one) and part of the object message paradigm
which I see as best for MP too, at least for objects.

I welcome comments.

Best Regards!
Bob