[Edu-sig] re: Types and true division (was Re: strange output

[Kirby Urner]

At 11:47 AM 10/11/2002 -0400, Arthur wrote:

>Why clutter up the list with quotes and links. It's all there on the record.

OK, thanks for clarifying.  I've not followed the literature
as closely as you have.

I think if newbie confusions had been the only motivation for
the change, at the expense of language coherence, then it'd be
as you suppose, a politically expedient maneuver that dumbed
down the language but broadened its appeal to the masses.

My view is the vetting process, including PEPs, whereby a whole
community gets to examine proposed changes, is critical, and
changes with a shaky technical basis could and should be shot
down, or at least exposed as shaky, during this stage -- even
were they to come from Guido.  This vetting process is what the
community is engaged in currently, around adding rational
numbers as a new type (plus other PEPs).

If the problems with div were originally connected to the problems
of newbies, and if this were the *only* justification for the
making the change, then I think you'd have a point.

As I understand it, you wish there'd be some acknowledgement that
at some point in the evolutionary trajectory, there was a time
when newbie concerns seemed to weigh too heavily in the balance.
You resent being portrayed as irrational merely for wanting to
insist on this historical point.

As I said above, I haven't followed the literature closely enough
to know what the original PEP said or didn't say in detail, but
were this point to be granted in some way, then you'd be
satisfied.

I take it you are also, for your part, willing to acknowledge
that the final motivations for the change, after vetting and
input from a broader community, ended up having to do with
purely technical issues not of concern to newbies.  So that, in
fact, newbies are only happier with this change as a matter of
coincidence, whereas in other cases (e.g. not including copy()
as a built-in) newbie concerns are not being addressed, once
again for technical reasons (i.e. technical reasons have so
far carried the day in both cases).

>I quite believe Tim that that was not the substance of *his* discussions
>on the matter with Guido.
>
>So?
>
>Art

I don't think it's irrelevant that technical justifications for
the change do exist, irrespective of newbie concerns.  The lesson
to draw from this is, if you can't argue your position on technical
merits, it has less chance of surviving.  Just championing the
"new user" is not going to carry the day when it comes to the
Python community's assessment process.  And yes, sometimes newbies
and pros will find themselves in agreement (other times not).

Kirby