[standards-jig] LAST CALL: Service Discovery (JEP-0030)

Joe Hildebrand jhildebrand at jabber.com
Mon Mar 3 21:14:24 UTC 2003


Peter Saint-Andre <stpeter at jabber.org> writes:

> Did we ever address this?
>
> http://mailman.jabber.org/pipermail/standards-jig/2002-December/002266.html
>
> I still see this more as an addressing issue than a disco
> issue. Since we don't have any fourth identifier beyond domain,
> node, and resource, I see no reason why disco must cover something
> that doesn't yet exist. :)

PGM just explained to me what the disconnect was.  Apparently, some
people thought that the router should have knowledge of the node
attribute, and make routing decisions based on the node.  That's
exactly opposite to what I meant in the message above.  The node is
only used within an entity, once the router has routed a stanza to
that entity using the JID.  The receiving entity can then do whatever
internal lookups it wants, to direct that stanza at the correct
internal endpoint.  Different node names might "address" different
internal endpoints.  From that perspecive, a node is a sub-address.
>From the perspective of the router, the node attribute is just another
bit of information that it can route from one JID to another.

The node is analogous to a mail stop within a business.  The post
office delivers the mail to the mail room for the street address on
the letter, and the mail room gets the mail out to the correct person,
based on the mail stop.  The post office doesn't look at the mail stop
and it doesn't count as part of the address, as far as the post office
is concerned.

-- 
Joe Hildebrand




More information about the Standards mailing list