Hi Dave/Peter Thanks for replies
Justification: One of use cases that I am facing currently that requires this feature (Publish multiple events within multiple nodes in an atomic fashion) is Say, a user performs an operation in an application - underneath this could result in multiple events getting broadcasted (e.g. A new patient getting admitted to a hospital could trigger an event for the "Bed management" to arrange for a hospital bed, an Order event for "Radiology department" to arrange for a scan, an event for the "billing dept" to create a bill etc). We want these events to be published in an atomic operation. So only if all the events succeed we can bubble up and return the single (user's) operation as success or failure. Also w.r.t Performance - Having this feature will prevent multiple server trips to XMPP server for publishing related set of events. If it can be handled in the server (like the current batch operations present in XEP 0060) then it becomes efficient. Can this request be considered as an improvement over the existing spec ? Thanks Vijay -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Peter Saint-Andre Sent: 2008 Oct 28 1:36 AM To: XMPP Extension Discussion List Subject: Re: [Standards] XEP 0060 :Publish multiple events within multiplenodes Dave Cridland wrote: > On Sat Oct 25 05:27:49 2008, J, Vijayananda wrote: >> Hello >> >> I have a use case where I need to publish multiple events in multiple >> nodes in a single publish call (to avoid multiple server trips). >> Whereas the batch processing feature present in the XEO 0060 allows >> publishing multiple items within a single node, I need multiple items >> in different nodes. >> >> > I see lots of mentions of "the publish element", etc, whereas the same > text mentions multiple items within it. I think we probably need to > assume that the spec is meant to restrict this to a single publish > element, at this point, and clarify it accordingly. > > You can of course send multiple <iq/> requests for publication > simultaneously, although they will not be atomic. Right. That use case is not addressed by batch publishing. >> Is this allowed ? since it was not very clear to me when I looked at >> the schema 'http://jabber.org/protocol/pubsub' It has only minOccurs = >> 0 for publish and nothing specified for maxOccurs > > I'm afraid I believe our schemas are not normative - or at least, the > text is considered authoritative in the case of conflict. That schema could be restructured in a more helpful way -- I'll do that in the next version of the spec. Peter -- Peter Saint-Andre https://stpeter.im/ The information contained in this message may be confidential and legally protected under applicable law. The message is intended solely for the addressee(s). If you are not the intended recipient, you are hereby notified that any use, forwarding, dissemination, or reproduction of this message is strictly prohibited and may be unlawful. If you are not the intended recipient, please contact the sender by return e-mail and destroy all copies of the original message.
