On 4/13/06, Eli Barzilay <xxxxxx@barzilay.org> wrote:
>
> [Please skip this paragraph.]
> [Yes, I know that you're not contradicting your own argument.  I just
> don't think that you're not not contradicting your own argument
> strongly enough for me to not say that you're contradicting your own
> argument.  (And I'm not contradicting myself in this sentence.)]
> [Just ignore the above.]
>

Oh, how exceptionally witty you are, Eli! Charming. Not exactly a
Will Clinger, but you're getting better.

>
> > But this is getting boring, [...]
>
> Right -- so there's this useful bit of functionality that just must be
> implemented over and over and over again.  If I lower my music for a
> second I can actually hear the c.l.l folks laugh.

The c.l.l folks? Oh, they are just trying to understand why nobody
uses their wonderful language (and they are still not getting it - I invite
you to think about it). Perhaps it's a matter with the keyword arguments,
who knows?

> >
> > Configuration-objects could be composed, inherited, modified by
> > accessors, whatever. I claim such an interface is cleaner, possibly
> > less verbose and likely to be more efficient.
>
> And so we get to the point where you give me an example that I
> explicitly talked about why it doesn't work.

No, you were describing optional arguments. That is different
from a (Hint: single) optional configuration object.

> Hint: the summary was
> that while this is possible, I've never seen real code that uses it

Well, I *have* seen real code that uses _configuration objects_ (assuming
we are talking about the same thing, but that assumpton may be wrong),
but perhaps I have just seen more code that was good for me...

Anyway, next question?

cheers,
felix