[Ironpython-users] feedback on PEP 421

[Eric Snow]

On Sun, May 6, 2012 at 4:20 PM, Jeff Hardy <jdhardy at gmail.com> wrote:
> On Sun, May 6, 2012 at 12:05 PM, Eric Snow <ericsnowcurrently at gmail.com> wrote:
>> Hi,
>>
>> I'm looking for feedback on PEP 421: "Adding sys.implementation".  The
>> idea came up in 2009 and garnered positive feedback, but didn't go
>> anywhere.  I've revived it and am hoping to get it worked out in time
>> for Python 3.3.
>
> First off, I like it. It's long overdue, but better late than never.

Agreed.  Better cognizance of the alternate implementations has been a
growing sentiment among the CPython core developers, especially Guido
and Brett Cannon.  Hopefully I'll be able to make a bigger push on
this front once sys.implementation is all sorted out.

>>
>> Any feedback would be very helpful, particularly with regard to the
>> decision on the type of sys.implementation and the constraints on
>> sys.implementation.version.  Thanks.
>
> I'd strongly prefer the named-tuple approach, like version_info. The
> metadata attribute being a dict is fine (although the PEP doesn't
> actually call that out in "Required Values" section; I had to search
> to find out if it was actually a dict). I have some stuff I want to
> put in there already.

Good to know.  Is your preference for a named tuple instead of a
normal class?  I'd rather not expose sys.implementation as a sequence
type, but mostly I want to end up with something that's meaningful for
everyone.

> I'm not really sure there's much value in having sys.version_info and
> sys.implementation.version be different, but I believe PyPy works that
> way, so I have no objection to it. They'll be the same in IronPython,
> though.
>
> Thanks for pushing this through. I'll probably add it the day after
> it's approved.

That's just what I wanted to hear.  <wink>

-eric