eduroam to end users round table 09.03.16 Trust and Identity Users' session:

 Challenges/barriers:

• eduroam viewed as wireless provider by end users, and there are alternatives/competitors.
• end users care mostly about getting to the internet, not so much about how that's achieved.
• eduroam is complicated to set up compared to other techs such as SIM cards. Creates a barrier to starting to use eduroam and possible that support is never asked for if an end user just defaults to a "simpler" solution. The impression of eduroam should be that it is of equal or greater ease to set up and use than alternatives.
• Difficult to fix when something goes wrong.
• Removal of roaming fees in the EU; end users less incentivised to use eduroam on mobile devices.
• MulteFire developing an offer of LTE-like performance with Wi-Fi-like simplicity - Could be a challenger to eduroam https://www.qualcomm.com/news/onq/2015/06/11/introducing-multefire-lte-performance-wi-fi-simplicity 
• Assisting end users with identifying who their IdP is and providing them with details of who/how to contact.

 Good stuff:

• Do still get feedback after 10 years that end users are impressed at the range of locations that eduroam is available in.

 Take-away:

• Most difficult: Improving the quality of support offered to end users by their IT/support departments/IdPs.
• Should improve on support tools, to enhance knowledge and enable better support to end users.
• How far should we go in "forcing" IdPs to be helpful? Could provide a document that outlines  best practice for eduroam; question over whether this should be for information only or something IdPs should be asked to agree/sign-up to?

Moonshot moving to service round table 09.03.16:

Current status and plan to launch:

• • Moonshot Pilot ran through GN3+ and GN4-1, and will continue in GN4-2 as service launch is prepared for.
  • The service specification and cost benefit analysis for a GEANT-operated service based on Moonshot were approved.
  • Will be working closely with SA2 in GN4-2 to transition into service. There are a number of dependencies in play. 
  • Earliest possible service launch will be the back end of 2016.

Challenges/barriers to deployment and uptake:

• Finding convincing use cases for which Moonshot is the solution (rather than a "solution without a problem").
• Moonshot technology supported and manintained in Distros, e.g. libraries and packaging for CentOS and Debian. Being quicker at getting into Distros when new versions released. Note: Packages for CentOS 7 being done currently, due end of March (with thanks to University of Murcia).
• Security of protecting users and traceability & audit of provisioned user accounts.
• Performance - slower than an end user will accept in production? With Trust Router the first establishment of a Trust path can take up to 9 seconds, but after this first establishment is quick. But to an end user, waiting 9 seconds may give the impression of a failed attempt.
• Difficulty in deploying the required infrastructure.
• Support for MAC. Researchers predominantly use MAC and support is not in place. Jisc intend to address this by Q3 2016.
• Need to understand what versions of operating systems are being used by potential users, and what is supported by the NREN/organisation.
• Competitors? ECP gathering pace and addressing SSH (one of the key use cases for Moonshot, also).
• Potential loss of interest if not progressed with haste to a MVP production release?

 Use cases:

• Jean Marie - RENATER: Sharing a resource amongst many users and provisioning accounts quickly.
• University of Murcia: Use case for providing outside users to resources at the institution.
• Avgust Jauk - ARNES: using Moonshot as authentication mechanism for email clients.

Thought:

• Potential to make use of underlying eduroam RADIUS infrastructure; can this be exploited?