-
Notifications
You must be signed in to change notification settings - Fork 5
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
No way to wildcard subscriptions or subscribe to subtrees #42
Comments
This proposal could be extended to support topic wildcard subscriptions following MQTT conventions ( E.G:
|
Dear @bdamm and @mattaylor, sorry for the late response to this. As I mentioned in other issues, the draft has now undergone many changes. In relation to The issue is at least twofold: 1. Finding all topic_data URLsI htink this part is solved, there are many ways to find the topic resources and if your broker had a specific hierarchy you could design wildcards. For example a collection with some topics under it on a predefined path like As @mattaylor points out, you can also subscribe to the collection resource itself The usual discovery mechanisms apply https://core-wg.github.io/coap-pubsub/ps-rev/draft-ietf-core-coap-pubsub.html#name-discovery 2. Subscribing to them in one single messageAssuming that under the topic collectiong there are 100 resources like ps/data/ (they can also be labeled with rt=core.ps.data). How does a client observe all of them with one GET message? This part that is not there yet. You could craft specific FETCH + obs messages that contain topic data URIs and subscribe to them in one go. @cabo for example mentions these two potential payloads: |
Suppose there is a topic structure such as /${tenant-id}/${device-mac}/${event-id}
Suppose a subscriber wants to receive all events with a specific ID. There is no way to do that in this protocol.
Suppose a subscriber wants to receive all events from a specific device-mac.
Suppose a subscriber wants to receive all events from a specific tenant-id.
How could this be efficiently realized at all, even with complete freedom to structure the topic tree in any way? With the current protocol I don't see how to do it. Subscribers would need to subscribe to millions of potential topics, even though there may not be millions of actual topics. Even worse, subscribers would need intimate knowledge of all possible topics (all possible IDs and MACs) which may not even be possible.
The text was updated successfully, but these errors were encountered: