[WikiEN-l] Community ready for Pending Changes (nee Flagged Protection)?

[Risker]

Despite the fact that I do have reservations about several aspects of this
trial deployment, I do recognise that this is indeed a *trial*, and that the
purpose of a trial is to learn, and to try out new ideas to see whether they
work in whole or in part.  The opportunity to learn is the reason that I
feel we should proceed at this time.

Having said that, I feel extremely strongly that we need to be willing to
take extra steps to protect the editors (particularly the non-admins) who
are putting their editorial reputations on the line to help us find out if
this is a workable process.  We need for the participants to make mistakes
(so we can figure out how to fix them), to give us honest feedback from the
"ordinary editor" perspective, to make sure that the tool works in the way
it's expected to, and to figure out whether the parameters we've set for
tool use are realistic and viable.

I think one thing that is missing from this entire trial process is that
there are no broadly accepted objectives, and conflicting supposed
measurables for determining whether or not the tool has made a difference.
These are some of the things I've heard bandied about:

*Pending changes will encourage more non-editors to try to edit, and these
new editors will become part of our community.
---Just because someone edits an article doesn't mean that they actually
helped. We will have no realistic way of measuring how many new editors made
useful edits and how many made vandalistic ones. Nor, if they are IP
editors, will we be able to say with any certainty whether they actually
stick around to become contributing editors.
---I'd like to hear from someone in the know whether or not we will be able
to determine if new accounts created during this period had their first edit
on an article under pending changes protection. If we can't tell that, then
we cannot attribute any higher-than-usual number of new editors to the use
of this tool.

*Pending changes will protect more BLPs.
---The same criteria for protection continue to apply. If the article does
not qualify for semi- or full protection, it does not qualify for pending
changes either. Pending changes is being billed as an alternative to semi-
or full protection and is explicitly not to be used as a means to extend
protection to articles that would not otherwise qualify.  At the end of the
trial, there should be no significant difference in the total number of
articles covered by one of the three forms of protection than there is at
the time we start the clock.

*Pending changes will help stop edit wars
---Edit wars are content disputes, and need to proceed through our normal
content discussion process; pages that have been protected because of edit
wars are not eligible for pending changes. The only exception is if an
article is semi-protected to keep anons/unconfirmed users from repeatedly
adding the same vandalistic or BLP-violating material, and that is vandalism
control as opposed to edit-warring.

*Pending changes will reduce visible vandalism
---Um, no. If every review of a pending change is carried out correctly,
there should be no difference in the amount of vandalism viewable by the
general reader. That's because otherwise the articles would have been semi-
or fully protected, and almost all vandalistic edits would have been
rejected by the software.

*Nobody's being prevented from editing in the way they always have
---We won't know until we try this part. If we see autoconfirmed editors
having their edits caught in the pending review queues of articles on Level
1 pending changes, then this is patently false; their edits have always been
publicly visible from the time they hit "save".  This is data that would be
really valuable to capture, if there is a way to do so.
---As well, editors who take on "reviewer" permissions will automatically
have their edits accepted, even on formerly fully protected articles (should
we decide to try Level 2 pending changes). The reviewer permission goes with
them everywhere, so they will now have to review any pending changes before
making their own edits to articles that may be part of the trial.

*Anonymous editors will now be able to edit the [[George W. Bush]]  and
[[Barack Obama]] articles
---No they won't. This was actually one of the first, and easiest decisions
made by the on-wiki team looking at trial implementation processes. There is
no reasonable chance that the number of useful edits will make up for the
incessant vandalism and BLP violations in what are already a {{good}} and
{{featured}} article respectively. They certainly won't be part of the
trial, and even if the community decides to continue using this tool, almost
every other BLP in the entire project would be a better candidate for
pending changes than these ones. Even the German Wikipedia still has some
protected articles.


So....now that I have deflated everyone's expectations....We really do need
to think about what we would consider to be a useful outcome in this trial.
It's time to stop thinking pie-in-the-sky, and get down to what we'd
consider a sufficiently positive outcome to proceed.

Incidentally, I think it's important that we reinforce repeatedly that this
is a trial. Trials end, and this one ends in two months. Unless there is a
newly minted community consensus to keep this trial deployment going, I
fully expect it to be turned off on August 15th, along with all the other
bells and whistles that go with it (such as deactivating the reviewer
permission). If there is no intention at this time to stop the trial and
deactivate the extension on August 15th, I'd like the WMF and the developers
to say so now. Because if that is the case, then this isn't a trial, it's a
seat-of-the-pants deployment, and the very large section of the community
that is already concerned about how this tool will be used will have every
reason to believe they have been handed a pig in a poke.

Risker