[Pidgin] #3154: Threading code in jabber makes no sense
Pidgin
trac at pidgin.im
Sat Nov 3 17:28:52 EDT 2007
#3154: Threading code in jabber makes no sense
-----------------------+----------------------------------------------------
Reporter: arivanov | Owner: nwalp
Type: defect | Status: new
Priority: minor | Milestone:
Component: XMPP | Version: 2.2.0
Resolution: | Keywords:
Pending: 1 |
-----------------------+----------------------------------------------------
Changes (by arivanov):
* pending: 1 => 0
Comment:
I believe the patch does exactly what the first sentence in the Thread
section of the most recent draft-saintandre-rfc3921bis. So it looks like
this is what St Andre (I assumed you meant him by stPeter) currently
thinks:
5.2.3. Thread
The primary use of the XMPP <thread/> element is to uniquely identify
a conversation thread or "chat session" between two entities
instantiated by <message/> stanzas of type 'chat'.
[snip]
The value of the <thread/> element MUST be a universally unique
identifier (UUID) as described in [UUID].
[snip]
You are right that the meaning of thread was slightly different before.
At least in my tests the patch does exactly what the draft says - any new
"chat session" is marked with a new UUID and all messages within the
conversation are marked with it. The moment both parties leave the
conversation the thread is considered finished and no further messages are
marked with the same UUID.
I probably need to look how this looks for all use cases especially when
one of the people "leaves" the conversation by closing the window while
the other one stays. There are some obvious ambiguities on
setting/resetting ids for conversations started simultaneously by two
people or restarted by a person while another one has closed it.
We should probably also comment on the draft before it becomes yet another
slightly underbaked RFC to ensure that these cases are covered.
--
Ticket URL: <http://developer.pidgin.im/ticket/3154#comment:5>
Pidgin <http://pidgin.im>
Pidgin
More information about the Tracker
mailing list