On 1/13/16 8:13 AM, François Laupretre wrote:
yeah, that(discussion only seems to happen after introducing the voting
phase) is frustrating for the rfc author, but that is the last phase
where
complaints can be voiced and most people have a tendency to defer stuff
until the last minute, that sucks, but not specific to our project, and
don't know what can we do about it.
some of those last minute feedback are actually useful, so if we are
looking for the best solution it i better to have those concerns to be
raised and heard even if late than ignored and voted on a flawed
proposal.
I think one possible countermeasure can be to start the voting as
soon as
the discussion dies down (while still keeping the minimal discussion
period) instead of waiting for more feedback arbitrarly and getting
frustrated that it only comes after one puts the rfc up for votes.
What do you think of the opposite solution : merge the discussion and
voting phases, e.g. allow voting as soon as discussion starts ? This
discussion/vote phase would be required to last at least 1 month.
People could either vote early and, maybe, change their vote if the
discussion makes them think otherwise, or wait for others' arguments.
The big difference would be that the RFC could be amended during this
discussion/voting phase. It could be a way to avoid having to restart
the whole process from the beginning.
Regards
François
Voting should never, ever be on something that's still in flux. Voting
should be on a fixed, unmodified text so you know what you're voting
on. Otherwise, something you're in favor of in its initial form may
drift to be something else entirely that you don't like by the time the
vote period ends, but since you already voted you're not paying close
attention anymore.
Which is the other problem: Once someone votes, they'll wander off. They
won't wait to see the discussion, see if anyone has good points to raise
(pro or con), wait to see if the text changes in a way that would change
their vote (pro or con), or whatever else. Knee-jerk votes are
generally uninformed votes, which is the last thing we want.
--
--Larry Garfield
--
PHP Internals - PHP Runtime Development Mailing List
To unsubscribe, visit: http://www.php.net/unsub.php
