Community Wishlist Survey 2022/Search/Close the outdated WiViVi and reintegrate its key features into Wikimedia Stats

Close the outdated WiViVi and reintegrate its key features into Wikimedia Stats

  • Problem: WiViVi was discontinued in 2018 after Erik Zachte retired as WMF's Data Analyst and some of its key features are not accessible anymore nor active on Wikimedia Stats. Without this tool, it is currently impossible to assess the outreach of a wiki linguistical community within a country.
  • Proposed solution: Integrate a browsing option that allows to search each wiki-specific pageviews as percentage to language for all worldwide countries. For example: In Spain 72% of the Wikipedia views are in Spanish, while 16% in English —but only 3.8% in Catalan or 0.4% in Basque (as of 2018; Asturian or Occitan are also indigenous languages in Spain but with such low % that cannot even access their proportional views). It is also important to be able to compare it with both native and total speakers, so that it can easily be seen if that wiki receives more or less views than expected by its speakers. The linguistical data can be easily taken by Wikidata, if available. For example: being able to check that only 15% of Wikipedia views in Andorra are in Catalan. Based on this value, if provided by Wikidata that Andorra has a population of 78K people, plus if provided by Wikidata that 40% have Catalan as their mother tongue, it can be seen that Catalan Wikipedia in Andorra is cleary underused (15% views in Catalan vs 40% native speakers of Catalan).
  • Who would benefit: All minoritized languages that really need to know which is the visit share (%) of their wikis as compared to the major or official language that is spoken within the borders of a country. This can help communities to plan their institutional collaborations and language use awareness.
  • More comments: Regional level would be desirable in the future. Percentages should be able to have a low cutoff or filter by specific language, in order to foresee very minoritized languages.
  • Phabricator tickets: phab:T257071, phab:T284294
  • Proposer: Xavi Dengra (MESSAGES) 19:08, 16 January 2022 (UTC)[reply]

Discussion

That is a wonderful wish Xavi, and I like the focus on giving visibility to minoritized languages. I will check in with the Data Engineering team to see if this is something we could get done with one wishlist cycle. KSiebert (WMF) (talk) 13:59, 17 January 2022 (UTC)[reply]
Because WiViVi is outdated I created commons:File:Most popular edition of Wikipedia by country.svg. I used the existing API to do so and this fairly simple piece of code. Unfortunately, it needs to be updated regularly. As you can see, the image is used on several articles in many editions of Wikipedia so I guess this data is considered useful. I opened a Phabricator ticket in July 2020 about this feature request. An interactive map like WiViVi would be better to display countries where a language is not #1 but still often used. For instance French in Canada and Morocco, Romanian in Romania, English in Eastern Europe, etc. As @Xavier Dengra: said regional subdivisions would be great. I opened another Phabricator ticket for that in June 2021. Without regional subdivisions, the map lacks granularity and under-represent "small languages", as mentioned in the ticket: "we have a unit on the map for Malta with 0.4m inhabitants, but not for West Bengal with 180m inhabitants". A455bcd9 (talk) 07:38, 19 January 2022 (UTC)[reply]
That's a great job, thank you for reaching out @A455bcd9! Creating automated maps via Commons to attach to articles is always a good idea, but in terms of basic communication, general awareness, research and data collection is not feasible. If there is a centralized option like Wikimedia Stats, which we all expect to be the "officially mantained" we cannot keep disseminating tools, patches and visualizations along the servers or repositories: neither the general public nor even most of wikipedians will find them. The depart of Erik Zachte and the "still-visible corps" of WiViVi sums up to the many examples of what shouldn't be repeated. There is not even a "last update" of the page and that leads to big confusion, therefore removing it or with a big outdated banner seem the only solution for me. For the interactive map, this is the Wikimedia Stats one that could integrate this Wish. However, it is worth to mention how unacceptable is the WM Stats current map: does not even include, for example, Andorra or San Marino in Europe, and forgets most of the states of the Pacific. At this point, several people in the world cannot even browse the views of Wikipedia in their country. Overall, a very anglocentric point of view not in line with the ongoing equity recommendations. Xavi Dengra (MESSAGES) 08:28, 19 January 2022 (UTC)[reply]
I totally agree that everything should be on Wikimedia Stats, and the map I created is, I hope, only a temporary workaround in the absence of a good and officially maintained interactive map.
I think you could add the two Phabricator tickets I opened in your submission btw ("Phabricator tickets:") A455bcd9 (talk) 10:12, 19 January 2022 (UTC)[reply]
  Done Xavier Dengra (MESSAGES) 23:31, 21 January 2022 (UTC)[reply]

Voting