[Standards-JIG] Re: What happened to the ACK proposal?

Stephen Marquard scm at marquard.net
Mon Aug 15 17:51:24 UTC 2005


Peter Saint-Andre wrote:
> David Chisnall wrote:
> 
>> The problem is not with Java, it is with TCP.  
> 
> 
> That's one reason the ACK proposal was not accepted as a JEP:
> 
> http://mail.jabber.org/pipermail/standards-jig/2005-March/007174.html

I must say I didn't really find the Council's reasons very convincing. Labelling 
this an "edge case" rather than a pressing problem is the difference between 
XMPP being a reliable protocol (like SMTP, where your message will either get 
delivered or it will bounce), and an unreliable "best-effort" protocol where 
your message may get delivered, but if you're unlucky, it could get lost without 
  you knowing about it.

Sure, TCP could be a better protocol, but that's not a reason for not addressing 
the undesired behaviour at the XMPP level (via a feature which can be negotiated 
during stream initiation).

Lastly, if it's something that should get fixed in XMPP, why not have an 
experimental JEP which client and server authors can implement to demonstrate 
the best solution for XMPP so that when it gets to the IETF there's some actual 
experience?

The contrast between the response to this JEP and the Apple 
"make-up-your-own-standards-and-then-have-them-accepted-by-the-JSF" case is 
striking.

Regards
Stephen




More information about the Standards mailing list