Replies (1)

• @christin@lsbt.me 2026-09-18 14:47

  @thanksstevenkim@social.vivaldi.net Hi Steven, Thanks for reaching out, and for the work that clearly went into building this registry — normalized software IDs, family groupings, daily-refreshed data, that's a lot of ongoing effort. We've looked at the registry and the docs, and decided not to integrate it into FediSuite's instance/software detection at this point. Briefly, our reasons: Every platform FediSuite supports needs its own tested adapter work regardless — character limits, API quirks, composer behavior, and so on. A name/family mapping alone wouldn't save us that work; we'd still have to build and maintain the actual adapter ourselves.For a feature that affects how FediSuite classifies and talks to other people's servers, it matters to us that we can fully trace and control the classification logic ourselves. Pulling in an external data source that updates itself daily, and whose contents we don't directly control, doesn't fit that — regardless of how carefully it's maintained.An additional runtime fetch to an external source would also be one more potential point of failure (availability, future schema changes) that we'd rather not introduce into the core product. That doesn't mean the project isn't useful to us — I'll happily keep it in mind as an informal reference if we ever add support for a newer or less common platform and aren't sure which family it belongs to. For a firm technical integration, though, it's not something we're looking to do right now for the reasons above. Thanks again for the outreach, and good luck with "The Hitchhiker's Guide to the Fediverse"! Best, Christin

  Open ##4749793