PEP scepticism

[Robin Becker]

In article <3B471314.8B179A92 at engcorp.com>, Peter Hansen
<peter at engcorp.com> writes
>Robin Becker wrote:
>> 
>> The language is evolving as all languages do because of the needs of its
>> users. Five years ago the leadership said integer division was good,
>> today it's bad; the audience changed is all.
>
>Actually, maybe it's just which audience members are vocal which 
>has changed recently.
>
>I happen to use Python frequently (almost exclusively) in situations
>where it is very valuable to have integer division work the way 
>it does now.  Floats are only infrequently used.
>
>If I were constantly forced to jump through hoops to make sure 
>my integers divided like integers (as in, like the CPU designers 
>intended... not like CS people believe it should work), I would have 
>a lot more work than I do now.
>
>But until now my voice was silent on the matter.  Perhaps the
>reason you think the needs of the users has changed is because
>Python suits the needs of some users so well that they don't
>feel the need to constantly whine about the fact.  Maybe only
>those who have a problem with it are going to make their voice
>heard.  (As someone said a few messages back: Well, duh! :-)
>

You infer too much as I prefer to keep the / operator as it is. My
preferred solution would be a module level configuration that allowed
the old style to be used. So then people such as yourself could do

from __past__ import int_divide

and in such a module the '/' operator would be treated as it always has
up to now. This is a comparatively easy way to do things, but I fear
that the purists amongst the python team would reject any such laziness
and demand that we do indeed jump through hoops.

I believe that the python developers have an interest in promoting the
language to a wider audience and so they keep changing the language to
enhance its popularity. 

If the MDFL was perfect and got python 'right' the developers would have
little or no reason to continue developing python. Experience suggests
they ought to hand the code to a new team and do something else.

I often wonder why these little languages now seem to evolve as
singletons rather than species like algol and pascal did. 
-- 
Robin Becker