[SciPy-user] matrix mult operator

Robert Kern robert.kern@gmail....
Mon Jun 25 14:18:58 CDT 2007

David Warde-Farley wrote:
> On 25-Jun-07, at 2:25 PM, Robert Kern wrote:
>> Because there isn't enough support for such a construct when we  
>> already have
>> if/elif and other ways of dispatching code. Guido asked the  
>> community and PyCon
>> 2007, and there really wasn't much of a favorable response.
>>   http://www.python.org/dev/peps/pep-3103/
>  From what I've read it was a quick, informal poll during a talk,  
> which is a painfully poor basis for making design decisions.

Of course, all of this happened after a long, drawn-out series of discussions
and designs, which is *also* a painfully poor basis for making design decisions
if it's the only basis. Why? Because every participant is self-selected. After
investing a lot of time sitting in the weeds, working out the details, it can be
difficult to take a step back and see if the feature you are working on is
really important enough. That's why Guido asked. We already have good ways to
dispatch code. Is adding another going to generally improve people's use of
Python? The answer was fairly clear. The community in general really doesn't
feel a need for a switch statement. For adding such a large feature as new
control flow syntax, you really should have broad support for it.

> But as  
> usual we are at the mercy of Guido and his fanboys and their  
> arbitrary notions of what's "Pythonic" or not.

You are entirely mistaken if you think PyCon is solely attended by Guido's
fanboys. Really, getting several hundred Python users across a broad spectrum in
the same room and asking them questions is actually probably one of the better
ways to keep perspective and determine how much of an impact your design
decisions are going to have.

Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco

More information about the SciPy-user mailing list