The current draft allows clients to cache the document for at least ttl seconds. Does this imply that the client can assume the location of the document won't change? Or does the client have to start with the WebFinger request each time it wants to update the cached copy?
On the last call we agreed that the client should try to reload the document using the cached location. If that fails with 404 it should do a WebFinger request to retrieve the new location.
The current draft allows clients to cache the document for at least
ttl
seconds. Does this imply that the client can assume the location of the document won't change? Or does the client have to start with the WebFinger request each time it wants to update the cached copy?