Wikifunctions:Status
 | Abstract Wikipedia team staff primarily maintain this page's content. |
This page aims to give a reasonably current view on the current status of Wikifunctions, and an overview of the Abstract Wikipedia team's planned work for the current quarter. If something does not work, this page is a good first stop to check if that is a known issue.
Is anything currently very broken?
- No current known P0 or P1 issues!
This section is not the complete source of details. A more immediate view is the "Abstract Wikipedia" Phabricator board. This section just aims to provide a good and more easily understandable overview of major known issues and problems.
See Wikifunctions:Report a technical problem for details on how to report a bug or make a feature request.
What are we working on this Quarter? (October-December 2026)
 | Updates on this quarter's work, and the new quarter's work to come, are first published in the community newsletter. |
The Abstract Wikipedia team's work this quarter supports our annual objective, DE2: Abstract Wikipedia (see the annual plan):
- Demonstrate Abstract Wikipedia's viability as a scalable, human-centered way to create multilingual encyclopedic content.
Key Result: Pilot Wikipedia Integration (Q1/Q2: July–December 2026)
Demonstrate the first integration of an Abstract Wikipedia article into Wikipedias.
Planned work for Q2:
- Lexeme coverage (continued from Q1, led by our colleagues in Wikimedia Deutschland). We're aiming for an average of 80,000 Lexeme edits a month across Q1 and Q2 (baseline for January–March 2026: 66,670), with efforts increasingly focused on Abstract Wikipedia's pilot communities.
- (T440365) Wikifunctions workshops for pilot communities. If we run workshops tailored to the pilot communities, we can support the communities in creating and improving the language-specific functions their languages need to render Abstract articles.
- (T440032) Improving language-specific function creation. If we make it simpler to create a new language-specific function, it should be easier for communities, pilot communities especially, to build the functions they need.
Key Result: Abstract Wikipedia Experience (Q2/Q3: October 2026–March 2027)
Enable editors without in-depth Wikifunctions knowledge to create and improve Abstract articles, and know what to do when the content they want cannot yet be expressed.
Planned work for Q2:
- (T440368) Abstract Wikipedia editing improvements. If we iteratively implement and evaluate focused editing improvements, then we can progressively make editing and article creation accessible to a broader, less technical group of contributors.
- (T440367) Feedback loop between Abstract editors and function creators. If we explore ways for contributors to capture sentences they would like to include but can't yet express on Abstract Wikipedia, and share these with the Wikifunctions community, we can establish a workable feedback loop between content needs and function development.
- (T440366) Exploration: Natural language to function mapping. If we explore approaches that map what contributors want to express in natural language to relevant function calls, we can find out whether this might be a viable way to lower the barrier to creating abstract content.
- (T440369) Cache-refresh for Abstract articles. If we provide a clear and reliable way to refresh stale cached data on Abstract Wikipedia, contributors will be able to see their changes in Wikifunctions and Wikidata reflected, and be more confident that their edits take effect.
On top of the above planned work, we'll also have some "essential work", where we respond to bug fixes and emergency situations following our standard protocols, aimed at reducing disruption for the Wikifunctions community and its users.
The plan for this quarter was published in the 9 October update, and progress is reported in the weekly updates.
You can see our team's Phabricator board for the current quarter for more detailed tracking of how things are going.
Longer-term plans
- These are issues that we hope to work on in the future, as part of the bigger plans for Wikifunctions and Abstract Wikipedia. We will prioritise between them based on your feedback and ideas.
Type creation is locked-down to staff
For now, we only support a limited number and nature of types, and creation is limited to only staff. There are a number of built-in functions, e.g. first element of a list, typed list, and many others, which are currently not well-supported for custom types, which we are looking at addressing. Generic types and generic functions require a bit of development and bug-fixing, and are not ready yet.
Function pages don't show you where or how much they're used
This would be an important way for the Wikifunctions community to decide how to focus effort and warn users of changes, like how the GlobalUsage tool guides the Commons community.
Diffs are ugly, so it's hard to do vandalism patrolling or community moderation
For now, diffs "work" but show ugly blobs of JSON rather than nice, understandable, formatted results. We want an experience like Wikidata's or better.
Search is ugly, so it's a problem to find things
For now, the search "works" but shows ugly blobs of JSON rather than a nice, understandable, formatted result, and you can't filter by type of object (e.g. "show me only Implementations that match my search").
Note: You can include the required type’s reference (e.g. "Z14") prefixed by the literal "Z1K1 " (case insensitive) in your search. For example, : suffix "z1k1 Z14" will tend to find Implementations containing the string "suffix" (because Functions and Test cases are unlikely to contain a string that is equivalent to "z1k1 Z14", whereas all Implementations contain "Z1K1": "Z14" in their JSON representations and this is equivalent to "z1k1 Z14").
