Closed yunapotamus closed 6 months ago
would like to verify that hipcamp would be able to satisfy 1 b/c they have the user id and anonymous user id (luid stand-in) and snackpass relies on us for user auth event logging so would fall into the unspecified 3rd option which is we do the user event logging.
i'm trying to figure out 2 a bit more. why would the userId be set alone w/o luid? when would this happen if 1 is true?
would like to verify that hipcamp would be able to satisfy 1 b/c they have the user id and anonymous user id (luid stand-in)
Yes, in theory. In practice, we are only sending the anonymous user ID.
and snackpass relies on us for user auth event logging so would fall into the unspecified 3rd option which is we do the user event logging.
We do the user event logging for Snackpass.
i'm trying to figure out 2 a bit more. why would the userId be set alone w/o luid? when would this happen if 1 is true?
userID
would not be set by itself. If you set logUserID
by itself, then any existing userID
would be set to nil.
Right now, setting the user ID requires a call to
startSessionAndLogUser(userID:)
, which also sets thelogUserID
and logs a User event. Proposal is to allow the following:userID
andlogUserID
at the same time. If the user setsuserID
then we force them to also setlogUserID
. This is because we don't want to let the two get out of sync.logUserID
, which would then clear anyuserID
already set.Sign off
Work begins when sign-off is received from all of the following: