IP Editing: Privacy Enhancement and Abuse Mitigation/Persistence

From Meta, a Wikimedia project coordination wiki

Please note that this is a technical investigation. It’s not a declaration of formed plans or a proposal.

See IP Editing: Privacy Enhancement and Abuse Mitigation for context.

The Wikimedia Foundation legal and public policy teams have asked us to investigate the potential impacts of storing masked IPs for a shorter period of time, compared with the current way we store IPs. We have some ideas and a general outline of how this could work, but we can't decide this without the Wikimedia communities in general and the patrollers fighting vandalism in particular – we need you to be part of this conversation. It could mean the masked identity would not have the same semi-persistent identity – tied to the IP, not the user – some IPs have. Depending on how this is handled, it could cause problems as the anti-vandalism workflows look today. We are aware of many of these. Many of them are the same as we already deal with, with temporary IPs.

It would make it more difficult to flag static public IPs like public libraries and schools, which is important to let other editors know to expect certain patterns of behaviour and the cost of blocking them. It lets us recognise good faith-users with an agenda and IPs with a mix of good and problematic editing behaviour, both things that can take a lot of time to investigate and should not be thrown away. Blocking is often a ladder: someone is blocked for a short period of time, then longer, then longer yet, if it doesn’t help. It would also lead to recurring identity loss for long-term unregistered editors. This is, however, part of the price paid by making the decision not to register an account. It’s fine to not actively register a persistent identity, but it comes with the drawback of maybe not having one. And so on. Would it be helpful if we implemented a distinction based on IP and a cookie? Today we have a problem with IPs shared by a very large number of persons: being able to pinpoint one could help us block IPs we can’t block today, or allow others to edit from a shared IP which would previously have been blocked. But it would also make it easier to circumvent for technically minded users requiring slightly more of an investigation before the IP could be blocked.

There are probably things we don’t know. Things we want to know, so that we can do our uttermost to avoid making the work of patrollers and vandal fighters on the Wikimedia wikis more difficult. Please tell us on the talk page. Let us know all the issues and all the examples of what problems not having this information could cause or what different solutions for this would mean for you. You can write in your language if you’re not comfortable with English.

Also, are there reasons this would help you? Anything this would be useful for in addition to better privacy protection, in that it would cause less confusion? Please let us know, too.