[Edu-sig] re: Does edu-sig extend to Jython

[Kirby Urner]

At 12:40 AM 8/20/2002 -0400, you wrote:
>Kirby writes -
>
> >If you have just 3 parameters, that's six permutations right there.
> >Four parameters means 4!=24 possible function signatures.  No Java
> >programmer want's to provide 24 ways of calling the exact same
> >function with the exact same parameters -- just in different order.
> >That would seem insane.
>
>If there is no good reason, its insane.
>
>If there is a good reason, it's 4 hours work that stands there
>as long as you want it to. (Or at least until 3.0 comes out).
>Not exactly like building houses out of soda cans -
>whatever the hell I mean by that.

Well, perhaps there's an elegant solution out there.  In most
cases, I think the suggestion to make the operator commutative,
regardless of type, simply postpones to case/switch logic to a
next level.

Whatever the solution, it shouldn't add the overhead of slower
speed.  As a user, I don't want a programmer bending over
backwards to accommodate my ignorance of programming, at the
price of giving me slow-moving bloatware ala some Microsoft
products.

>But as I say, I don't see what the good reason would be if
>I were working in Java.

Why would you feel different about your user if it were a Java
app?  Don't you want the same freedoms to pass parameters in
any order, no matter what the language?  Just curious.

>But I do think I know why I would be wanting to accomplish
>it in my exact circumstances.  Will look closer, and back-off of it if it
>proves truly unwieldy.

Well, you're the ultimate judge in your own exact circumstances.
As a general rule of thumb though, I think it's not unreasonable
to have clients of a class pay attention to parameter order.
They have to think a bit like programmers because they're
programming at this point.

Kirby