ietf-openproxy
[Top] [All Lists]

Dynamic Rules (Re: Some Enquiries on IRML)

2001-04-24 07:32:42
This is really a philosophic point, but it can have significant consequences
for the policy architecture.  Policy, I believe, is a reflection of the
intended behavior of a network.  When a person logs on, the intent does not
change, it's the network state that changes.  This taxonomy implies that
policy is static (changes on human time scale not computer time scale), it's
updated out of band with very limited requirements for synchronization and
the rule owners and admin servers need not be involved in network state
changes.

For the example given, isn't the 'normal' case going to be that the service
provider is going to have some page at which the user registers their
preferences and pays for the services?  That application, then, could have a
transaction that updates the rules on behalf of the client.  If instead, we
think that browser configuration (or set-top box or whatever) is going to
express rules (e.g., language preference implies translation service) then
the proposed taxonomy needs a closer inspection; but I wouldn't think that
to be the case.  Am I missing something here?

Lee M. Rafalow
Voice: 1-919-254-4455; Fax: 1-919-254-6243
IBM Internet Technology Management
IBM Corporation
P.O. Box 12195, BRQA/502
RTP, NC 27709 USA
Alternate email: rafalow(_at_)us(_dot_)ibm(_dot_)com
Home email: rafalow(_at_)mindspring(_dot_)com


<Prev in Thread] Current Thread [Next in Thread>