[Wikimedia-l] Who invoked "principle of least surprise" for the image filter?

[Tobias Oelgarte]

Am 18.06.2012 13:52, schrieb Thomas Morton:
> On 18 June 2012 08:00, Tom Morris<tom at tommorris.org>  wrote:
>
>> On Monday, 18 June 2012 at 02:44, Tobias Oelgarte wrote:
>>
>>> Every stupid bot could do this. There is no "running out of the box"
>>> solution at the moment, but the effort to set up something like this
>>> would be minimal compared to anything else.
>>>
>>> I would say that Citizendium failed because they did no automatic
>>> updating. What i have in mind is delayed mirror with update control. It
>>> is not meant to be edited by hand. It is a subset of the current content
>>> selected by the host (one or many users) of the page himself. It is
>>> essentially a whitelist for Wikipedia that only contains
>>> selected/checked content. That way a "childrens Wiki" could easily be
>>> created, by not including any unwanted content, while the effort stays
>>> minimal. (Not more effort then to create your own book from a list of
>>> already written articles)
>> {{sofixit}}
>>
>>
>> If all the people in favour of filters had spent their time building them
>> rather than arguing about them, we would have had a wide array of different
>> solutions, without any politics or drama.
>>
>> That said, if people want to filter Wikipedia, a client-side solution
>> rather than a filtered mirror is preferable. If a filtered mirror were to
>> come into existence and become popular, this would mean that people would
>> just filter all of main Wikipedia, which would prevent people from editing
>> Wikipedia. A client-side solution means they are still looking at
>> wikipedia.org just without naughty pics and doesn't interfere with
>> editing. It also reduces the need for any servers.
>
> The technical solution is a fairly trivial part of the problem; a
> client-side filter could probably be put together in a few days IMO.
>
> The *hard* problem is convincing the "not censored" abusers that it's a
> useful feature for our community.
>
> Tom
>
It is not convincing since it interferes with the work of our editors 
that aren't interested in such a feature. If we tag images inside the 
project itself then we impose our judgment onto it, while ignoring or 
separating it from the context it is used in. The first proposal 
(referendum) mentioned various tagging options/categories that would 
have to be maintained by the community, despite existing and huge 
backlogs. Additionally we are a multi culture project with quite 
different view points and which accepts different view points (main 
difference between Flickr and Co). The result will be huge amount of 
discussions about whether to tag an image or not. This leads me to the 
simple conclusion that it isn't worth the effort, especially if the 
filter is advertised to make Wikipedia a save place for children, while 
everyone (including children) can disable it at any time.

Separate projects that only focus on one task (providing a whitelisted 
view, an automatically updated subset of Wikipedia) would not be a 
burden for the community or at least for everyone not interested in or 
against filtering. Additionally it could define it's own strict rules 
and could even hide images and articles entirely depending on it's goal.

But i have to add that the WMF should not be part of this projects. This 
projects define their own rules like Flickr and Co.