Talk:Requests for comment/Hiding the number of Russian/Belorussian/Kazakh contributors on the statistics map

• The WMF staff is aware of the RfC and is working on a reply. I know that it doesn't solve the problem, but in the meantime here's some data for Russia in 2022 - 23.--Victoria (talk) 01:37, 21 October 2023 (UTC)

Signpost coverage

FYI: We briefly covered this RfC in the new issue of the Signpost (the English Wikipedia's community newspaper). Regards, HaeB (talk) 07:17, 6 November 2023 (UTC)

Possible Solutions

I am involved in the development of both Wikistats and the country protection list. The privacy team is helping come up with a good long term solution. I was just wondering, in the meantime, what everyone here thinks would work in the short term. One possibility that came to mind was to identify the countries on the list with a special shading/color. On hover, some text could explain that "data from here is restricted". This seems useful regardless of how we change the list or the dataset, and it seems to address a UX problem we've missed until now. Milimetric (WMF) (talk) 18:41, 27 November 2023 (UTC)

This sounds reasonable as a short-term solution. (Russian-speaking admin) Victoria (talk) 09:32, 28 November 2023 (UTC)

This is now done via this change. See linked task T333716 for tracking. Milimetric (WMF) (talk) 22:02, 6 December 2023 (UTC)

Title

Should we move this page to Requests for comment/Hiding the number of Russian, Belorussian, and Kazakh contributors on the statistics map as now every slash makes this RfC actually a subpage? A09 (talk) 21:22, 5 December 2023 (UTC)

Since there are no upper pages, excluding "RFC" itself, it isn't needed, I think. MBH (talk) 18:18, 7 December 2023 (UTC)

Update - the happy end

As a Trustee, I contacted the Tech department and found out that:

The dataset for ru-wiki editors released in February was created manually, utilizing differential privacy techniques that aren't part of the usual data processing pipelines. This data has been manually processed and added to the geoeditors_monthly dataset, which is published monthly here.

Integrating this into the WMF's regular data processing pipelines would require significant engineering effort. Currently, this work is not scheduled because the teams capable of doing it are focused on other priorities outlined in the WMF annual plan. One of their key tasks is addressing the urgent issues with the “dumps” process, which produces publicly available data at [1] and supports the monthly metrics pipelines, among other functions.

WMF analysts suggest that if WMF begins releasing differentially private datasets in Wikistats, they would also need to adjust the user experience to help viewers interpret the data accurately. Such datasets can display characteristics that differ from the plain datasets currently available. This would be a new undertaking requiring experimentation to determine the best approach, making integrating into the existing system a non-trivial task.

Good news:

The Enterprise team has allocated resources for differential privacy pageview work for FY 24/25. The ru-wiki dataset will be maintained and expanded as part of a broader enterprise integration covering pageviews across all language projects. The work is scheduled for completion by Q4 FY 24/25 as part of an Enterprise product. Following Enterprise’s operating principles, the dataset will be made publicly accessible, with access details to be provided once the Enterprise team finalizes their project plan. Victoria (talk) 08:35, 21 August 2024 (UTC)