Clearly the movement is in a bit of a quandary here.
The Board, some of whom have been elected by the whole community, has decided to implement an image filter, the full details of which have not yet been announced/designed.
The Foundation announced a referendum, but actually ran a consultation, the results of which give many pointers as to what features of a filter would be more acceptable, workable or the reverse.
But it didn't give a clear answer as to the level of support for an image filter as that wasn't a question in the consultation.
Our third largest project has held a referendum with a very clear result, though as we don't yet know how the filter would work, I do wonder which potential version(s) of the filter they were voting on.
An image filter would inevitably involve Commons, our largest project at least in number of mainspace pages; But it could be implemented in such a way that any other project could opt out of it.
One possible solution to the current divide would be:
1 The Board publicly accepts that this system will not be implemented without the support of the community in a referendum.
2 Using the results of the consultation the devs code up a filter and install it on a test wiki. This will enable people to know how it would actually work and what (dis)functionality it would contain. This might need to involve choices in the form of different versions of the filter. A version or versions of the filter only get to be considered for full implementation if they've been tested and there are people who want to commend that version of the filter to the community.
3 The movement commissions some research among readers and potential readers as to their attitudes to this sort of censorship on wikimedia sites. This research would attempt to answer amongst other things, how many, if any people who avoid us now would use our sites if we offered such a filter (for me and I suspect some others there would be no point in progressing this if the people who currently don't use us would not be mollified by such a filter).
4 Decide the electorate(s), question wording and interpretation of a referendum. This includes deciding between a Federal solution, (we have/have not support over the movement as a whole so this will/will not be implemented on all wikis) and a Confederal solution (those wikis that vote for it get it, those that voted against don't). If its a confederal solution we need to remember that some of our wikis are inactive and many are not yet created, so we need to decide whether this is Opt in or Opt out. The electorate also needs to be agreed, this is almost simple for a Federal election, but for a confederal one you have to decide if somebody who is active on three wikis get one vote on the federal total, but can vote in three different wikis as to whether they opt in or out. If the devs can't code all the feedback into one version of the filter and instead offer us a choice of different types of filter then this referendum could start to get complex. Getting one series of questions where we can agree what the questions mean, how the results will be interpreted, and where everyone who can make up their mind on the issue will be able to express their opinion with a particular set of answers, will not be easy. But I think it is possible.
5 Translate the referendum into multiple languages, and then hold the referendum
6 Announce, discuss and if we have a green light, implement the result. If we have a red light then we can stop the process, otherwise:
7 If some or all projects decide to implement this, then we need to tell our readers how this works.
8 Monitor the results
9 After an agreed time review the results. This is the time to ponder questions such as who is actually using the filter, what are they filtering out, are they happy with the result? If we've implemented it in some languages spoken in the Islamic world have we gained readership there?
While I personally probably wouldn't use a filter I'm more than happy that those who want to filter out spiders, penises, artwork banned by their religion and indeed various degrees of nudity can do so. But more important to me is that we find a way to discuss and resolve this that leaves both sides, and especially whoever doesn't get their way, thinking that they've been listened to, and that the process has been fair.
For me it would be better to be on the losing side of a fair and open process that on the winning side of an unfair one.
WereSpielChequers
On 16 September 2011 23:00, WereSpielChequers werespielchequers@gmail.com wrote:
While I personally probably wouldn't use a filter I'm more than happy that those who want to filter out spiders, penises, artwork banned by their religion and indeed various degrees of nudity can do so.
But again this requires no foundation involvement to do. It can already be done through adblock plus with no modifications.
I would like to see a minor clarification to your suggested actions WSC. If the WMF commission any further round of analysis or referendums there must be openly reported direct costs and some reasonable measure of the lost opportunity cost of volunteer time used to discuss, analyse and contribute to these pre-implementation activities that may have no eventual outcome in the projects.
Being told that the costs were "tiny" without WMF (apparently) having the capability of calculating how much donated money has been eaten up in this debate I find increasingly worrying, particularly if we go through another round.
Thanks, Fae
On Fri, Sep 16, 2011 at 11:00:05PM +0100, WereSpielChequers wrote:
Clearly the movement is in a bit of a quandary here.
2 Using the results of the consultation the devs code up a filter and install it on a test wiki. This will enable people to know how it would actually work and what (dis)functionality it would contain. This might need to involve choices in the form of different versions of the filter. A version or versions of the filter only get to be considered for full implementation if they've been tested and there are people who want to
That's fine and dandy.
However, (unless I am very wrong), there is no issue with the filter, but rather with the underlying category system.
Therefore, it might be wiser to look at that category system in isolation. That would also be rather cheaper. ;-)
sincerely, Kim Bruning
wikimedia-l@lists.wikimedia.org