yum-yab
commented
1 year ago Q1: What's the plan for the prototype now?
Currently the plan for the prototype is (as far as I know, @holycrab13 may correct me):
- We (or someone else) sets up a specifically configured lookup, indexing metadata from one specific mod -> these should be publicly available
- These regularly rebuild their index (e.g. cronjob) to reflect the current status of the specific mod
- encourages following standards, e.g. publishing metadata in the frictionless format because there is already a mod that picks it up and a lookup that automatically indexes it
- returns Databus Identifiers based on metadata of the mod
- If someone wants to include certain mod metadata in its Databus search, multiple lookup endpoints can be added/deleted/managed in the settings
- Everyone can write metadata to everything, but by doing it this way the search/view of each individual user does not get cluttered
- There can be a list of top-curated Mods that can be added by everyone, showcased in the Databus itself
- Once someone searches, the client queries the Databus itself, but also every lookup endpoint configured in his settings, merges the results by best effort and presents it to the user
- no more work for the Databus server - all hard work is performed by each individual client
Ideally, we provide a data search layer over the Databuses and Mod data. This was already aptly labelled MOSS. https://github.com/dbpedia/databus/issues/58 and https://github.com/dbpedia/databus/issues/57 tackle data acquisition and discovery and where it is stored.
Q1: What's the plan for the prototype now?
Q2: What are existing generic approaches and standards to combine searches (server or client-side)?