Hi gang! Interesting idea, it’s definitely a lofty goal to have “one specifier to rule them all” kind of thing, and I applaud this kind of initiative.
Considering that PEP 725 creates a new standard, and the authors propose the virtual addition upstream to the spec, would that inclusion remove the need to maintain a separate DepURL specification?
If so, would it be an even better ecosystem-wide initiative to combine efforts into making the registry mapping part of the work for purldb.readthedocs.io/ instead of creating another bespoke registry, or asking packaging ecosystems to adopt a registry?
There’s some efforts already in play live on https://public.purldb.io/ - it doesn’t appear to be fully launched yet, but maybe there’s already a bunch of work done there on scanning and matching package-names-to-PURLs that might accelerate adoption of PEP 725 if it satisfies the need? See also More decentralized, distributed PURL metadata collection · Issue #727 · aboutcode-org/purldb · GitHub