Summary of PEP 308 Vote for a Ternary Operator

Tim Hochberg tim.hochberg at ieee.org
Tue Mar 11 05:51:18 CET 2003


Raymond Hettinger wrote:
> The PEP 308 vote is summarized at:
>   http://tinyurl.com/763f
[Snip]

First off, Raymond, thanks for going to the work of putting together 
this vote.


> * The highest ranked constructs were:
> 
>   235 for        (if C: x else: y)
>   206 for        C ? x : y 
> 
> * The 235 breaks down to 177 accepting and 58 rejecting.
>   If the RejectAll votes are attributed entirely to that
>   syntax, the ratio becomes 177 favoring to 140 opposing.

This certainly isn't how I would interpret this. Options that I didn't 
rank I considered as rejected. I suppose that might be arguable in the 
case where someone ranked three options and accepted all three: in that 
case an argument could be made that they might have accepted some other 
syntaxes given the opportunity, but in cases where at least one syntax 
was rejected, and for example, syntax C was not ranked I would consider 
it rejected.

So, my initial reading of this, without crunching the numbers to get the 
fuzzy votes is that option C was closer to 177 accepting and 341 rejecting.

Alternatively, I think the fuzzy votes come to ~100, so if we just throw 
those out, the most accurate thing I can think of given the constraints 
of the vote, one has 177 accepting to 241 rejecting.

I can't recall the details for the complimentary vote, but perhaps the 
two weren't so far off after all.

[SNIP]

> * Though the results lean towards accepting the PEP as
>   proposed, it is not decisive.  Some of the no-change
>   votes included strong pleas.  This will certainly be
>   a consideration.

[SNIP]

Again, I'd disagree. The results seems to slightly leaning toward some 
ternary syntax, but no change seems to be preferred almost 2:1 (or 1.5:1 
if you toss the fuzzy votes) over any particular syntax. I certainly 
wouldn't characterize that as leaning towards accepting the PEP as proposed.

Regards,

-tim





More information about the Python-list mailing list