[Foundation] More Questions...
mikelin at MIT.EDU
Wed Jul 31 01:36:23 CDT 2002
Since noone else except stpeter seems to be posing debate questions
thusfar, I've pulled some more out of my hat. Obviously, they've been
questions I've been doing a lot of thinking about, and I am in the
election, so the de facto overseers of this election can use them or
not; but I would really like to know everyone's answers to them.
There has been a lot of talk -- and 4 separate JEPs -- concerning
building a pub/sub system into Jabber. I would really like to know: why
do you want pub/sub? I don't mean this in any way sarcastically; I am
really interested to know exactly what kinds of systems and applications
you see being built on top of a Jabber pub/sub system, and I would
especially like to have concrete examples rather than nebulous concepts
- particularly from anyone who has had direct experience in building
large pub/sub systems.
In particular, it seems that there should be some useful synergy among
pub/sub, message queueing, and presence. How do you see this fleshing
out in specific applications?
How do you think ownership of the trademark on the Jabber term should be
handled, particularly in light of the fact that the goals of the JSF and
the commercial interests of Jabber, Inc. seem increasingly divergent?
What steps should the JSF take, if it should take any such steps at all,
to wean itself from its dependence on Jabber, Inc.?
3. PASS, DSPS, etc.
As you probably know, I am in favor of building JNG into a messaging
system that will fully allow the in-band transport of large payloads.
What do you think of this? If you believe an out-of-band solution is
needed, what (technically) differentiates it from SIP?
It looks as if we are moving forward with SASL for client/server
authentication, and Paul Lloyd has done some great work with end-to-end
message security and XML Encryption. The major issue that has yet to be
fully addressed is inter-server trust. There has been some talk of using
SASL to authenticate between servers; this is fine from a protocol
standpoint, but the larger issue is, how do we establish trust between
servers on an Internet scale, especially when we introduce multi-hop
routing, where dialback no longer offers whatever assurances it does
More information about the Members