CALENDAR INTEROP UWs View on Objectives Status Obstacles

CALENDAR INTEROP UW's View on Objectives, Status & Obstacles 03 June 2010 Terry Gray

MEETING OBJECTIVES Shared view of current pain, opportunities, and challenges Brainstorming on options and Best Ways Forward Foundation for future progress 2

AGENDA Introductions Setting the Stage -Terry Google Perspective -Chris Microsoft Perspective -Paul UW Perspectives -Erik, David, Brad Update on Cal. Connect -Paul Converging on Key Issues -All Identifying Best Approaches -All 3

THE PROBLEM Outlook/ Exchange User IT Staff http: //www. loc. gov/exhibits/bobhope/images/vcvg 20. jpg Google Calendar User 4

CONTEXT: Research Universities http: //liu. english. ucsb. edu/wiki 1/images/4/4 c/Collaboration. gif Mission: discovery & innovation Means: extreme collaboration – Globally, at scale Culture: decentralized; diffuse authority – Collections of many independent businesses – A microcosm of “the Internet” “Industry turns ideas into money; Universities turn money into ideas. ” --Craig Hogan 5

COLLABORATION VISION the illusion of simplicity and coherence! • Seamless & simple collaboration across multiple platforms & orgs – – – Cal/Scheduling “just works” Doc sharing invitations “just work” User & resource discovery is easy no matter where data is hosted. • Robust federation replaces “Multiple Account Madness” 6

THE PLAYING FIELD UW Exchange Servers Other universities Microsoft Live@edu + BPOS UW Share. Point Servers Other cloud services Google Apps UW IMAP & Web Servers The IT challenge: make collaboration work in this context! 7

CLASSES OF USERS (both MS & Google) Within a given collaboration platform: Full suite of apps Basic email/cal services Calendar only Collaborator only Across collaboration platforms Entitled users Invited collaborators (should not need local credentials) 8

INTEROPERABILITY ELEMENTS Access control to calendar data; account assumptions Data format (e. g. i. Cal, . ics) Transfer/Access protocol (e. g. Cal. DAV, Web Svcs API) Autodiscover for authoritative data (per user, per group) Client design re profiles, individual v. group context/domain System-wide vs. per-user configuration 9

IDEALLY Exchange & Gcal users can be freely intermixed – At both department and individual level Existing tools work the same for scheduling all users Don't need accounts for everyone on each system System admins can easily configure things for everyone Delegation, Resources, and Recurring meetings work Scheduling works across institutions, not just across depts Don't need to keep an Outlook client running to sync Solution works Live@edu, not just on-prem Exchange 10

THINGS WE'VE TRIED Google server-to-Outlook sync tool (deprecated) Google server-to-server sync tool (permissions issues) Google apps sync plugin for Outlook (separate Profile) DIY Exchange web svcs to web page tool(s) DIY One-way Google → Exchange sync tool 11

KEY ISSUES Is Cal. DAV still important (vs. Web Svcs APIs)? Server-to-server vs. client-to-other-guy's-server approaches Autodiscover vs. configured precedence lists Account & access control assumptions Client design assumptions (“native” vs. “foreign” users) Scheduling resources (rooms) Departmental domains vs. selective group access 12

Lifting the fog at UW. . . DISCUSSION 13