[WikiEN-l] Rollback, and now here comes instruction creep

[Gregory Maxwell]

On Jan 11, 2008 4:46 PM, Majorly <axel9891 at googlemail.com> wrote:
> No thanks. I already said I disagreed with that.
>
> I suggest everyone read Anthere's latest post on WP:AN.

It would be helpful to people to include a link:
http://en.wikipedia.org/w/index.php?title=Wikipedia%3AAdministrators%27_noticeboard&diff=183698882&oldid=183693933

(otherwise anyone reading this more than a few hours from now might
end up fishing through the archives)

I thought her final point was interesting:
> My suggestion right now would be to say that the feature turned on by JeLuF is NOT critical. Give it a try. See if it is
> helpful, or to the contrary, bring in more troubles than it solves them. And revisit the issue in 3 months if
> necessary (yeah, vote again). Give it a try, and move on.

If the argument for this is "give it a try", then why are we not
extending rollback to autoconfirmed users?

Doing so appears to also be widely supported (in subsequent polls, it
was not offered as an option the poll used to justify the most recent
change). Moreover, extending rollback broadly is consistent with the
current handling of other features which are slightly prone to
disruptive use (move, edit semi-protected), yet still more
conservative than other parallel features (undo, and edit itself both
cover a lot of the same space as rollback).

Finally, requiring admin approval undeniably has a lot of overhead and
promotes dispute (which we've seen in practice already), it can only
really be justified based on the reasoning that anything less
authoritarian would lead to significant misbehavior or vandalism.
That such things would happen is far from a universally held view.
If we make it automagic, and learn that it causes problems, we could
always back off to admin approval (with its associated overheads). If
we just make it admin approval based we will be subject to overhead
and never learn if the simpler approach would work better.

Clearly there need to be a way to break status quo and test new things
even in the face of uncertainty, but simply taking the path of the
most aggressive parties, as seems to have been done here, is probably
not a good solution.  At the end of the day all the yelling about
consensus seems misguided to me. The overwhelming majority of users
don't know or don't care about this poll.   Most large polls on enwp
these days tell us almost nothing about consensus, instead they tell
us much about which extremest group is larger, better organized, or
more angry.