HELP: restore my faith in Python

[Ken Seehof]

Okay.  You are right.  Better yet let's make sure floats Never work
correctly.  The more they work right, the more people will think they
work right.  We should have a special case for 1/2.  Obviously, 1/2
should be 7.392.

Also, there should be a limit on the size of lists of 100 items (beyond
which
the program should dump core).  If people can have over 100 items in
a list, they might think they can have several trillion, which would
overflow
memory.

Mirko Liss wrote:

> Ken Seehof wrote:
> >
> > Seems to me an improvement would be to have python artificially
> > add epsilon to floats before converting to int.  This would still
> > suck but IMHO it would suck less.
>
> It's not just the conversion to int. As soon as you tell the
> programmer not to expect the worst from floats, he'll invent
> other ways to make a mess with those ugly floating point numbers.
>
> Want an example?
>
> >>> pennies1, pennies2 = 3.0 / 20 , 3 * 0.05
> >>> menu = { pennies1 : "spam" }
> >>> menu[pennies2] =  "lobster"
> >>> menu
> {0.15: 'lobster', 0.15: 'spam'}
>
> Ooops !
> So the division 3.0/20 is just a bit inaccurate, isn't it?
> But who cares, as long as I get at the sea food.
> So I count my pennies and ...
>
> >>> menu[0.15]
> 'spam'
>
> Spicy regards,
>
> Mirko
>
> --
> M Liß, n89553 at squid.upb.de