Page MenuHomePhabricator

Snapshots for saved queries
Open, Stalled, MediumPublic

Description

(This is basically a scaled-down version of T157231/T192870, though it does not require developing a new type of query engine)

Users may not only want to see the results of a query, but also see how the result sets evolve over times (see T290961#7352669 for a particular use case). This is a general idea containing two possible versions:

In both cases, a scheduler is useful, for example we want a query to be run once every day/week/month.
Both are depending on softwares/services that do not currently exist, so this task is currently not actionable.

Event Timeline

MPhamWMF moved this task from Incoming to Feature Requests on the Wikidata-Query-Service board.

I imagine a world where in the query service there is a time slider: and you can change it to query old snapshots instead of the current data. So, we can visualize the Wikidata situation as today, or as in January 2022, 2021, 2020 for whatever query. To T192870: Provide a mechanism to get historic results of a SPARQL query on a WDQS instance is really the solution.

That's possible deploying some query services starting from old RDF dumps.

This is not something totally impossible; it is something we already know how to do. Already doing an experiment starting with yearly snapshots would be a very powerful killer feature, from scientific to community purposes.