On Nov 24, 2007 1:51 AM, Waldemar Kornewald <[EMAIL PROTECTED]> wrote:
>
> Do you at least agree that Smalltalk has more potential for precedence
> errors than, say, Python?

Of course not, why would it?  In Smalltalk the syntax is uniform in
every situation.  I never have to worry about "oh, is this a statement
or an expression?  Or wait, it's math so that's another case" as I
would in C/C++/Python and so on.

And besides, even if people who do heavy *do* feel more at home with
mathematical precedence rules, what percentage of actual programmers
are they?  The very small minority I would say.

But just to put this issue to bed:

56 * 18 + 3 / 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18
+ 3 / 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7
* 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 +
191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 + 191 -
33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 + 191 - 33 / 11
+ 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 + 191 - 33 / 11 + 15 *
2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 *
30 / 15 * 56 * 18 + 3 / 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15
* 56 * 18 + 3 / 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 *
18 + 3 / 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3
/ 7 * 182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 *
182 + 191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 +
191 - 33 / 11 + 15 * 2 + 1 * 30 / 15 * 56 * 18 + 3 / 7 * 182 + 191 -
33 / 11 + 15 * 2 + 1 * 30 / 15

Ok, now lets go through and read off the steps needed to evaluate the
above expression.  You have to use mathematic precedence and I will
use Smalltalk binary message passing precedence.  Care to wager which
of us will finish first?  It can't be you because you have to make
multiple passes over the expression while I can just start from the
left and read to the right saying each atom I see.  And if we actually
evaluate the expression the difference gets even wider.

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

Reply via email to