On Nov 29, 2007 6:36 PM, Igor Stasenko <[EMAIL PROTECTED]> wrote:

> And from this point, ask yourself, why precedence rules, which we
> learn in school have defined in that way? Is there any other ways, how
> they can be defined? Is there any other math operators, except +,*,-,/
>  exists? What is math operator at all? How many of them we can define
> and introduce? Should we always tend to define a precedence rules for
> these new math operators? This is the first questions which you should
> teach and answer before proceeding further.


Reverse polish notation is a great example of math code without
operator precedence.  Yes, it takes time to learn.  It took me several
days to learn -- actually, to "unlearn" what had been drilled into me
at school.  But now I am *much* more productive (there's that word)
on a RPN calculator than I ever was on an algebraic one.  And my
five-year-old daughter needed to be shown only *once* how to use
RPN before she got it.

Simplicity and generality trumps complexity every time.

Waldemar: Remember that this is a tool for building tools.  If you
really want operator precedence, then it'll be easy to create it in
your own code.  But please don't force the rest of us to.

Ian: Please try bring as little as possible of the mainstream software
world's baggage into *OLA.  Otherwise the future you guys are trying
to create will end up looking an awful lot like the present.  I for one
am longing to see something that's far better (and different)  than
what we have currently live with.

Andrew
_______________________________________________
fonc mailing list
[email protected]
http://vpri.org/mailman/listinfo/fonc

Reply via email to