On Tue, Nov 01, 2005 at 09:05:49AM -0500, Jimmy Wales wrote: > We aren't talking about bank-level security. If some vandal does manage > to vandalize a Wikipedia page, well --- uhhhh, actually, that's quite > easy to do, anyone can do it -- and we just revert it. It seems likely that this solution has been discussed before, but my understanding is that Wikipedia works by having a team of volunteers periodically scour the submitted pages for signs of vandalism. Why should they do this ex-post? Is there some critical need for edits to Wikipedia pages to be published as quickly as possible, even if this means that there is potential abuse? I question the need for publishing first, asking questions later. Perhaps your utility function is construed such that for cases in which you feel that you can use routing information and rely upon the ISP for a backchannel for complaints, the benefit of getting edits out right away outweighs the cost of vandalism. However, in the case of Tor nodes, it is clear to me that you believe that the cost of vandalism is too high for this policy. Therefore, I propose screening edits from Tor nodes. Aside from the delay, this method carries no additional cost, since your editors are accustomed to scouring articles anyway. As long as vandalized pages are not published successfully, the value of this method for trolling the world will be neutralized. Even if the world magically turns into a utopia of location-independent sources, you will have a means of dealing with abuse quickly. and provided that your editors are sufficiently effective, you will never publish a vandalized page. Geoff
Attachment:
signature.asc
Description: Digital signature