[standards-jig] disco, x:data, etc...
alexey at sevcom.net
Thu Feb 20 19:13:36 UTC 2003
On 20 Feb 2003 09:04:53 -0800, you said:
MAM> I request that you again read JEP-0050 before discounting it. Your first
MAM> argument against x-commands is not valid,
Agreed, but "The payload can be any elements in an extension namespace" still
says that <command/> can't contain all namespaces,
MAM> and the second is vague.
MAM> Now for the second argument. What exactly in x-commands is "hard
MAM> processing from both client and server side"?
Not hard, it "more hard" that it can be to solve one simple task: to address IQ
MAM> So far, you have brought forth concrete issues with "sessionid" and
MAM> 1) "sessionid" is currently required. It's for tracking command
MAM> execution, in much the same way that web application servers (e.g.
MAM> Jakarta-Tomcat, WebLogic) use a similar variable. Allow me go further
MAM> into this for the requester (aka "client-side"; 1.A) and the responder
MAM> (aka "server-side"; 1.B).
I not say that x-commands are bad, but they are bad to solve task of addressing
to node, when we not want to track session, or to know status, etc... This is
separate things, so I prefer to see how your JEP solves its task, but not
addressing to node. IMHO this is separate task.
More information about the Standards