We have implemented and are using the above extensions to PubSub here, and have 
a few comments.

XEP-0248 I would like to get in un-deferred (what ever that word is) and I can 
take authorship of it.
  It is a working part of the Tigase PubSub component. I would request it is 
moved on in the std. track?

XEP-0253 Is being implemented here. I would request it is not deferred. I can 
take authorship of it.

XEP-0254 Is implemented and in testing here. I would also request that it is 
not deferred and I can take authorship of it.
  Along is a diff of some changes to the current document I have done.

Index: xep-0254.xml
===================================================================
--- xep-0254.xml        (revision 3550)
+++ xep-0254.xml        (working copy)
@@ -175,11 +175,12 @@
   </section2>
   <section2 topic='Deleting an Item from the Queue' anchor='delete'>
     <p>When the subscriber that received the item has successfully processed 
it (whatever that means in the context of the queue), the subscriber deletes 
the item from the queue.</p>
+    <p>Note: Whatever the retract is successful or not, the component MUST 
subtract from worker's lock count unless this is already 0.</p>
     <example caption='Entity deletes an item from a node'><![CDATA[
 <iq from='[email protected]/foo'
     id='delete1'
     to='workflows.shakespeare.lit'
-    type='get'>
+    type='set'>
   <pubsub xmlns='http://jabber.org/protocol/pubsub'>
     <retract node='a290fjsl29j19kjb'>
       <item id='ae890ac52d0df67ed7cfdf51b644e901'/>
@@ -237,6 +238,7 @@
   </section2>
   <section2 topic='Unlocking an Item' anchor='unlock'>
     <p>The subscriber might determine that it cannot process the item 
(whatever that means in the context of the queue); if so, the subscriber 
unlocks the item.</p>
+    <p>Note: Whatever the unlock is successful or not, the component MUST 
subtract from worker's lock count unless this is already 0.</p>
     <example caption='Entity unlocks an item'><![CDATA[
 <iq from='[email protected]/foo'
     id='unlock1'

Reply via email to