This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Section 8 "Implementation Considerations" starts with the following sentence: "Implementations SHOULD generate expirations in subscribe and renew request and response messages that are significantly larger than expected network latency." This seems like pointless advice. Proposal: remove sentence
i guess maybe in the original spec the clock starts on the sender side, so the subscriber has to count the network delay. Now that the spec says the clock starts on the receiver's side, so the network delay, however long, is removed. But still the subscription manager should NOT create a subscription that will expire before the subscriber gets the response, due to network delay. Otherwise, the subscriber may get a positive SubscribeResponse while the subscription has actuall expired. However, in practice, the expiration time is probably sufficiently large already and the above scenario is unlikely to occur.
resolved as proposed