Hi Eli,
Le 11/01/2016 15:45, Eli a écrit :
On 1/10/16 8:15 AM, Dennis Birkholz wrote:
I would really like to understand the rational behind anonymous voting
in the PHP internals context. Votes for RFCs should be purely based on
technical reasons and whether the language change would benefit the
language in the long run or not. I see no reason why such a vote
should be confidential.
I will chime in my quick thoughts here Dennis, as to a reason I could
see for doing so ... (Not going to argue if 'this reason is good
enough' or not. But it is a valid reason)
If a person does not stand behind his/her opinion for a technical
change, I am not sure if that person should be allowed to decide the
future of the language.
So the reason is not because someone isn't willing to 'stand behind
their opinion'. It's purely about being harassed (perhaps beleaguered
is a better terminology to not confuse this with 'illegal harassment')
for having said opinion. I was one of the people who, due to my vote on
STH, immediately started being beleaguered for holding my views and for
voting as much. My inbox/twitter/IRC/etc filled with how I was ruining
PHP and ruining people's lives. Old friendships were threatened to be
ended. And my entire week ended up becoming full of responding to these.
Instead of getting to be an informed voter, go in and cast my vote, and
await for the results to be displayed ... I become embroiled into the
arguments, back-n-forth, defense, and dealing with the beleaguering
comments.
Yes, I stood behind my opinion. But it has made me gun shy about
voting in the future on any contentious topic, because I know I need to
set aside the time to 'deal with that'. Yet those contentious topics,
are the ones where we should be encouraging as many people as possible
to vote, to make sure that we have a broad spectrum of views and that it
is the 'will of the community' as it were. And (at least in the US) is
against the idea in general of voter confidence. Where you are free to
hold your belief without needing to be slammed for it publicly.
So anyway, that's one reason. Whether it's a good reason or not is up
to others to decide.
... But it may be preferable to hide the Person<->Vote table until the
vote is over. That would provide protection against harassment to win
someone over and change his/her vote.
Unfortunately that won't stop the above situation. While it would stop
the idea of campaigning someone to change their vote (which is perhaps
another reason to do it). It just means all the above issues would be
taking place post-vote, instead of during-vote.
Eli
That's the reason why I suggested anonymous votes once again.
On few occasions, it happened that I didn't vote on an RFC because I
didn't want the RFC author to see how I voted. That may look strange but
one may have lots of reasons not to want his vote to go public.
Sure, in an ideal world, we should stand by our decision and be ready to
defend it with pure technical arguments, whatever relationship we have
with the RFC author. Unfortunately, that's not always the case and the
STH saga proved that, at least in this case, a lot of people, like Eli,
would have felt more comfortable if votes had been anonymous. Anyway,
the course of the vote would have been very different.
I must say I don't understand why people want individual votes to be
public, even after the vote is over. The mailing list is here for
eveyone to expose arguments. IMO, the vote is a place for privacy. I
definitely *don't* want an RFC author to ask me why I voted against his
proposal, and I will certainly *never* do that on one of my RFCs. If I
have something to say, I will say it on the list. My vote is my private
decision and I don't have to justify it against anybody.
Actually, that's nothing else than the rules already applied in every
elections in democratic states. Would you approve everyone to know who
you voted for after a presidential election ? Someone said that
registers are used to keep a track of who voted. That's right but the
information is here to avoid mutiple votes and is not publicly
available. So, even the voters' names are not disclosed. I just suggest
we do the same, ensuring votes cannot be biased by non-technical
considerations.
A case that could be prevented by public votes is the RFC being
massively rejected while almost no objections were done on the ML. In
this case, we could argue that the RFC author could use the vote
information to get more information from voters. Unfortunately, that
case already happened with Stas' (great :) RFC about default class
constructor, which demonstrated that public votes are not an efficient
protection against this. IMHO, the solution to such case is not a
question of public/private vote. It is more due to the fact that many
list members don't read the RFCs before vote starts. So, the discussion
phase is almost empty and, as soon as you start the vote, everybody
wakes up and asks you to stop or change everything. It happened to me
several times and is, IMO, a real issue in our RFC process. That's
another subject but I would like the process to be amended to disable
posting opinions/discussions about an RFC while the vote is open,
considering there was enough time for that during the discussion phase.
Regards
François
--
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php
