[Foundation-l] Possible solution for image filter

[Andreas Kolbe]

--- On Thu, 22/9/11, Tobias Oelgarte <tobias.oelgarte at googlemail.com> wrote:

From: Tobias Oelgarte <tobias.oelgarte at googlemail.com>
Subject: Re: [Foundation-l] Possible solution for image filter
To: foundation-l at lists.wikimedia.org
Date: Thursday, 22 September, 2011, 23:06

Am 22.09.2011 23:55, schrieb Andrew Gray:
> On 21 September 2011 14:14, Jussi-Ville Heiskanen<cimonavaro at gmail.com>  wrote:
>
>> The real problem here is that if there was a real market for stupid
>> sites like that, they would already be there. And they are not, which
>> does seem to point to the conclusion that there isn't a real market
>> for such sites. Doesn't it?
> Not really.
>
> There are basically no major WP-derivative sites of any kind in
> existence - the ones that exist are either plain dumps studded with
> ads, or very small-scale attempts to do something good and innovative.
> As far as I can tell, it's just very hard to get a fork or a
> significantly different derivative site up and running successfully;
> it requires a large investment on fairly speculative predictions.
>
> Given this, it's hard to say that the absence of a particular kind of
> derivative site is due to there being a lack of demand for that *kind*
> of site - there might be demand, there might not, we just can't tell
> from the available evidence.
>
> (To steal David's analogy, it's a bit like saying that unicorns can't
> be trained, as there are no trained unicorns. Of course, there are no
> unicorns at all, and their trainability is moot...)
>
Given the situation that we would provide a filter, as described in the 
referendum as a reference, it would be relatively easy to set up 
something like live mirror. It could work like a proxy (possibly with 
own caches) that could enable specific filtering as the default, without 
the option to disable it. One might provide it as a service for 
institutions that would simply redirect access to Wikipedia over such a 
proxy and therefore enforce the hiding of the images.

Currently you would have the need to create a live mirror and to feed it 
with tagging data. The proxy isn't money intensive, but the tagging is 
very expensive if you would need to do it alone. Thats the main reason 
why no such pages/proxies exist.

If *we* provide the tagging, then it would be much easier to do things 
like that.

And where would the problem be? If a user prefers to go to a Bowdlerised site like that,
rather than wikipedia.org, where they will see the pictures unless they specifically ask not
to see them, then that is their choice, and no skin off our noses.
A.