Internet Engineering Task Force                     J. Stracke
INTERNET DRAFT                                            eCal
draft-ietf-calsch-crisp-01.txt                     August 2001
                                        Expires: February 2002


           CAP Realtime iTIP-based Scheduling Profile (CRISP)


1.  Status of this Memo

   This document is an Internet-Draft and is in full conformance with
   all provisions of Section 10 of RFC2026.

   Internet-Drafts are working documents of the Internet Engineering
   Task Force (IETF), its areas, and its working groups.  Note that
   other groups may also distribute working documents as Internet-
   Drafts.  Internet-Drafts are draft documents valid for a maximum of
   six months and may be updated, replaced, or obsoleted by other docu¡
   ments at any time.  It is inappropriate to use Internet-Drafts as
   reference material or to cite them other than as "work in progress."

   The list of current Internet-Drafts can be accessed at
   http://www.ietf.org/ietf/1id-abstracts.txt

   The list of Internet-Draft Shadow Directories can be accessed at
   http://www.ietf.org/shadow.html.

   Distribution of this document is unlimited. Please send comments to
   francis@ecal.com or to the ietf-calendar@imc.org discussion list
   (subscription address ietf-calendar-request@imc.org; "SUBSCRIBE" or
   "UNSUBSCRIBE" in the body).


2.  Copyright Notice

   Copyright (C) The Internet Society (2000).  All Rights Reserved.


3.  Abstract

   This document sets forth a restricted profile of [CAP], one which
   supports no operations beyond the scheduling functionality of [iTIP].
   The motivation is to permit use of CAP's real-time iTIP functionality
   without exposing the calendar access functionality (which may require
   stricter security controls than iTIP).





Stracke                                                         [Page 1]


CRISP                                                        August 2001


4.  Introduction

   [iTIP] defines a scheduling protocol based on exchanging specially
   formatted [iCalendar] messages.  iTIP is defined to be independent of
   transport protocol.  At present, there is one standard binding of
   iTIP to a transport protocol, [iMIP], which carries iTIP messages in
   email.  This is a useful base level capability (email can reach vir¡
   tually any user on the Net), but can involve considerable latencies.
   A real-time binding for iTIP would be useful; it would permit appli¡
   cation developers to give users better feedback on the progress of
   the iTIP operations.

   Since CAP includes full iTIP functionality, one option would be to
   permit full access to CAP; to schedule an event with a remote user,
   one would then make a CAP connection to their CS.  The problem is
   that such a connection may be considered a security risk in some
   organizations; even though the CS has ACLs to prevent the client from
   performing non-iTIP operations, it would be better if the client sim¡
   ply could not attempt such operations.  (It's as if mail administra¡
   tors were told that an SMTP server outside the firewall had to
   include IMAP functionality as well.)  Thus, this document defines
   CRISP, a profile of CAP, a subset which does not support non-iTIP
   operations.

   This document does not specify the relationship between a CRISP
   server and a (full-powered) CAP server.  They may be implemented
   together, with the CRISP server being nothing more than the CAP
   server responding in CRISP mode (e.g., based on source IP address);
   the CRISP server may act as a proxy for the CAP server (see Firewall
   Application, below); the two servers may feed into the same database,
   but not know about each other; or there may be no CAP server, only
   the CRISP server, used for interdomain scheduling, but not for calen¡
   dar access.  Or, of course, there may be other modes of operation.
   These are implementation details, which do not need to be included in
   a protocol spec.

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in this
   document are to be interpreted as described in RFC 2119.

   Note that CRISP is a replacement for the former proposal known as
   iRIP (no reference is available, because the Internet-Draft has long
   since expired), which was abandoned when it was realized that the
   functionality of CAP was a superset of the functionality of iRIP.







Stracke                                                         [Page 2]


CRISP                                                        August 2001


5.  Profile Definition

   A CRISP server is a CAP server with the following capabilities:

      * ITIPVERSION=1.0
      * CAPVERSION=1.0
      * CAR=NONE
      * QUERYLEVEL=NONE

   In addition, various AUTH capabilities are expected.  Anonymous
   authentication SHOULD be supported, since the point of CRISP is to
   permit iTIP communication across security domains.  However, other
   authentication mechanisms may make sense in some cases; for example,
   if CRISP is being used for scheduling between cooperating companies
   (that is, in an extranet), then one company's CRISP server might be
   able to authenticate users from the other company.

   Other capabilities which apply to iTIP operations MAY be specified;
   e.g., MAXDATE and MAXICALOBJECTSIZE.

   Note that NONE is not a legal value for CAR or QUERYLEVEL in the cur¡
   rent draft of CAP.  This will have to be resolved.

   A CRISP server MUST NOT accept any iCalendar component which is not a
   valid iTIP component.

   In effect, the statement that a server is CRISP is a statement about
   the server's current advertised capabilities.  It is conceivable that
   a CAP server might be CRISP under some conditions and not others.
   For example, the server might offer a CRISP capability set on initial
   connection, but upgrade to full CAP if the client uses STARTTLS and
   provides an appropriate certificate.  It's not clear, though, whether
   there's any good way to advertise this fact.  For the rest of this
   document, we will assume that a CRISP server is always CRISP.


6.  Possible Firewall Application

   This section is non-normative.

   Clearly, it would be undesirable for an organization with a CAP
   server to have a CRISP server implemented completely separately, but
   having access to the same database.  Such duplication would increase
   development costs, maintenance costs, and security exposure.  On the
   other hand, it would be possible to build a CRISP server which han¡
   dles all operations by proxying them to the CAP server.  Such a proxy
   could be placed within the "no-man's-land" common in firewalls; the
   firewall would permit CAP connections from the outside to the proxy,



Stracke                                                         [Page 3]


CRISP                                                        August 2001


   and from the proxy to the internal CAP server.  The proxy would
   review all incoming iCalendar components and validate that they were
   legitimate iTIP operations; no non-iTIP components would be forwarded
   to the CAP server.  Similarly, if necessary, the proxy might censor
   the iTIP replies coming from the CAP server.

   Naturally, this is not the only approach possible; this section is
   merely illustrative.  The CRISP client does not know or care how the
   CRISP server gets at the underlying calendar store.


7.  Security Considerations

   CRISP is a subset of [CAP], and accordingly inherits all of CAP's
   security analysis.  However, new analysis does need to be done for
   the subset, especially since the whole point of the subset is to
   address security concerns.



8.  Author's Address:

   John Stracke
   Chief Scientist
   eCal Corp.
   Email: francis@ecal.com



9.  References

   [iTIP] Silverberg, Mansour, Dawson, Hopson, "iCalendar Transport-
   Independent Interoperability Protocol (iTIP)", RFC 2446, November
   1998

   [iMIP] Dawson, Mansour, Silverberg, "iCalendar Message-Based Interop¡
   erability Protocol (iMIP)", RFC 2445, November 1998

   [CAP] Mansour, Dawson, Royer, Taler, Hill, "Calendar Access Protocol
   (CAP)", draft-ietf-calsch-cap-05.txt, July 2001.  Work in progress.

   [iCAL] Dawson, Stenerson, "Internet Calendaring and Scheduling Core
   Object Specification (iCalendar)", RFC 2445, November 1998








Stracke                                                         [Page 4]