Comment on draft PEP for deprecating six builtins

[James J. Besemer]

Robin Becker wrote:

> map, filter & reduce et al are the fastest ways to do many things....,
> pow and divmod are also extremely useful. I don't use input, but I'm
> sure others do.
>
> Ther are no fast equivalents to the functional forms, list
> comprehensions are as slow as for loops. We need more fast operations
> for functional style programming, not less.
>
> This PEP smacks of political requirements not practical ones. The
> builtin name space is only crowded if we want to use the names for
> something else.
>
> Again this is a backward incompatible change. It's not broken and
> doesn't need fixing.

Hear! Hear!

I don't see a problem that this change solves and it actually detracts
from the language, as Robin so eloquently puts.

Furthermore, divmod() and pow() are natural and common enough and have a
direct and familiar relation with inline operators that they reasonably
deserves to be builtins.

Furthermore the change invalidates a lot of the books published to date
for Python, for no good reason.

I gotta say, for a group who ostensibly is looking out for newbies, I
submit that the number of language changes in the last 12 months may
constitute the largest single barrier to understanding the language.  It's
a moving target.  A number of people HERE have said they can't keep up
with the latest.

I think we should be very reluctant about changes that break old code.
Further we should focus on the important, substantial ones instead of
dinking around with trivial crap like this.

Regards

--jb


--
James J. Besemer  503-280-0838 voice
http://cascade-sys.com  503-280-0375 fax
mailto:jb at cascade-sys.com