PEP 238 (revised)

Paul Boddie paul at boddie.net
Fri Jul 27 11:38:46 CEST 2001


Peter Hansen <peter at engcorp.com> wrote in message news:<3B60B3D5.664A3D7F at engcorp.com>...
> 
> The updated PEP also makes no mention of a utility for scanning 
> existing code to check for potential problems.  Is this considered
> an intractable problem?  Is there any commitment to doing this,
> eventually, as part of the implementation of this PEP?  Or is the
> presence of the "warn" option considered sufficient (to allow for
> testing, rather than scanning)?

I'm really disappointed that the "migration" process hasn't been
addressed in the PEP, since the disruptive nature of the changes to
division constitute my principal objection to the adoption of these
changes. I get the impression that there isn't a good solution to this
problem, and since this situation could hinder "progress", it is being
"swept under the carpet".

But possibly the first thing anyone with any investment in the
language should be thinking upon starting to read the PEP is: "How on
Earth am I going to know what needs changing, and how much time and
money am I going to be spending to adapt to this?"

One could suggest (as I did before) that until such time as effective
type inference/analysis tools are available, this change shouldn't go
through.

Paul



More information about the Python-list mailing list