[standards-jig] JEP-0047 (IBB) Updated

Dave Smith dizzyd at jabber.org
Mon Apr 7 23:12:35 UTC 2003


On Monday, Apr 7, 2003, at 16:41 America/Denver, Justin Karneges wrote:

> Since XMPP guarantees the order of packet delivery, there is not even 
> the need
> for a timestamp or a seq# or anything.  I was not aware of such a 
> guarantee
> until just a moment ago, when I chatted with Peter.

XMPP guarantees order of delivery, but that is quite a bit different 
from sequencing (which provides a measure to also detect MISSING 
deliveries, as DW points out).

We need the sequence number -- but the prevseq attribute, should and 
CAN go away. If we hit the max of a 32 bit integer, we can just 
negotiate a new stream. It's an unlikely edge-case that shouldn't have 
dedicated protocol attributes just to deal with it.

80/20 is your friend.

> Assuming we switch to messages, this is what I propose for the data 
> packet
> format:
>
>   <message to="joe at blow.com/Home" id="inband_2">
>     <x xmlns="http://jabber.org/protocol/ibb" sid="mySID">
>       <data>A1B2C3D4E5F6</data>
>     </x>
>   </message>

Add back the seq attribute on the <x> element, and I'm jiggy with it.

> Also, switching to <message> means some extra rules:
> - If the entity goes unavailable, STOP sending packets

Sure, that's fine.

> - When the entity returns, it may receive some spooled IBB packets.  
> It should
> drop them (this should naturally happen anyway).

No, don't drop them. They're perfectly good packets.

Diz




More information about the Standards mailing list