Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:60544
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: error (pb1.pair.com: domain nebm.ist.utl.pt from 193.136.128.21 cause and error)
Content-Type: text/plain; charset=utf-8; format=flowed; delsp=yes
To: "Pierre Joye" <pierre.php@gmail.com>, "Nikita Popov"
 <nikita.ppv@googlemail.com>
Cc: internals@lists.php.net
References: <CAF+90c-Xfs0n-E=+s3g9GE=UZHGvPYdagCho0ma65u6YB+0Tgg@mail.gmail.com> <CAF+90c-_78TufE8+L8UNfuXLwPW9QgrSWyQaHvS8TOjj40-_cQ@mail.gmail.com> <op.wd8z16uqidpuyk@damnation.mshome.net> <CAEZPtU7BvZEHh4wQErQh2E+=REV0W=RXx-iPxUAiesTS39neRg@mail.gmail.com> <CAF+90c-N1Sj9TwT8WFdsrjCbY32g5NJnLzQcaUzfJocd1Nmn0Q@mail.gmail.com>
Date: Sun, 13 May 2012 16:43:43 +0200
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Organization: =?utf-8?Q?N=C3=BAcleo_de_Eng=2E_Biom=C3=A9di?=
 =?utf-8?Q?ca_do_I=2ES=2ET=2E?=
Message-ID: <op.wd84i5cxidpuyk@damnation.mshome.net>
In-Reply-To: <CAF+90c-N1Sj9TwT8WFdsrjCbY32g5NJnLzQcaUzfJocd1Nmn0Q@mail.gmail.com>
User-Agent: Opera Mail/11.62 (Win32)
Subject: Re: [PHP-DEV] Re: [VOTE] Vote change for empty() RFC
From: glopes@nebm.ist.utl.pt ("Gustavo Lopes")

On Sun, 13 May 2012 16:15:43 +0200, Nikita Popov  
<nikita.ppv@googlemail.com> wrote:
>
> I'm not sure I can follow. The vote has three options, of which two
> are quite similar. I don't see how the 2/3 rule for votes with two
> options can be applied here, in such a black&white fashion.

The fact there are more than two options should not alleviate the majority  
requirement. Let's say everyone who voted for empty()/isset() would prefer  
none to empty() only. Should this vote now pass just because you added an  
extra options that dilutes the votes against the most voted option?

For instance, debian requires a supermajority for certain proposals  
independently from the number of options (see e.g.  
http://www.debian.org/vote/2008/vote_003 ).

> We had at least one precedent of a vote with three options, where the
> option that was implemented in the end had only 59% of the votes. That
> was the vote for the callable typehint [1]. The three options were:
>
> * Callable: 34
> * Callback: 18
> * Neither: 6
>
> I think that vote was very similar to this one. Two of the options
> were "in favor", but differed in the exact implementation, and the
> last option was "against". The only difference is that in this case
> one option actually has 65%, not 59%.

This vote is an aberration and your description is disingenuous; some  
people that voted for callable also voted for callback, but your  
description suggests that the votes for callback should all count against  
callable. So someone that would vote for callable and callback would  
actually be casting a vote against whichever of those two options that  
would gather more votes.

The following votes were cast:

callback
callable
neither
callable, callback
callback, neither

If there is anything that can be learned from here is that this lack of  
clarity should not be repeated.

> Additionally I want to note that in this case there was a more general
> Yes/No vote before the more precise one, which ended with 12:2 (86% in
> favor).

Irrelevant. That vote happened before the pertinent objections had been  
raised (for instance empty(UNDEFINED_CONSTANT)/isset(UNDEFINED_CONSTANT)  
now returning false/true). And in any case, that voting was superseded by  
the new one.

-- 
Gustavo Lopes