Attending: Michael Gettes, Ted Hanss, Mark Johnson, Chris Phillips, John O’Keefe, Klara Jelinkova, Ron Kraemer
With: Ann West, Kevin Morooney, Dean Woodbeck
Today’s meeting continued the review of the Trust and Identity portfolio and business model. Last month’s conversation included eduroam and the InCommon Certificate Service. Today will cover the InCommon Federation and software development.
Almost 970 participants
500 research, colleges and universities actively use the federation - that is, they publish metadata
About 210 sponsored participants actively use the federation
The remaining participants (~260) are either organizations coming up to speed on federation or are Certificate Service subscribers who only intend to use the Certificate Service
The InCommon participation fees support InCommon services and activities
Fees were increased in 2017 to close gaps in operational maturity, service maturity, scaling, and security. This generated $750,000 in new revenues for InCommon operations
The Steward Program long-term business model is to be determined. Currently, MCNC is the only Steward at this point, but there are other regionals interested and it is critical to recruit move into the program. There is also interest from Kevin Thompson (NSF) in the Steward Program’s potential to provide access to scientific instrumentation and databases for K-12 educational purposes.
Kevin reviewed the current InCommon fee model (www.incommon.org/fees.html). He presented several pie charts showing the mix of InCommon participants (Level 1, Level 2, etc.) and revenue sources.
Ann mentioned two threats to InCommon participation. One is eduGAIN, since InCommon charges for membership and most other national federations do not. This could result in service providers leaving, since they could join another federation at no cost. So far this has not happened. The other threat is the implementation of Baseline Expectations, which will likely cause some attrition, particularly among sponsored partners.
Overall, InCommon is financially stable, but like most organizations that do technology service provisioning, we take on a service debt over time. For example, we will adding per-entity distribution of metadata and are looking at support for OpenID connect. There is also some demand for additional support for adoption support (glossies to use on campus to explain federation as one example). One decision could be that InCommon fees increase each year for a certain length of time.
Another gap is a place for CIOs and technical people from InCommon participants to gather. There is a perception that the Internet2 meetings aren’t for those folks.
Software Engineering, Development, TIER
The core software components of the Internet2 community software engineering and development efforts (Shibboleth, Grouper, COmanage) all originated with funding from the National Science Foundation (over $10M over the years)
As it is with NSF-funded software projects, maintenance, security and new features are more than a challenge once the grant funding is exhausted
The core components are connected, yet independent, open source software projects. Internet2 helps with the connections but not exclusively.
There was discussion about the TIER software components, funding provided by Internet2 member dues and the one-time TIER investor funding, and thoughts about what will happen next.
There was discussion about an analysis of the investment/cost of supporting individual components and the adoption/deployment of those components. Might this change the dynamic in what the community would support financially?
There was also discussion about communicating the progress that has been made through the TIER investment - to tell the story to the non-investors. The TIER architects and developers have developed a document that shows the initial TIER requirements, which have been met, and what is yet to be done. This is a long, dense document, but the information could be used to help tell the story.
One focus for 2018 is TIER adoption, with the Campus Success Program as the centerpiece.
There was also discussion about what is happening in the commercial sector that could potentially be a threat to InCommon, particularly if we leave it to those entities to control the conversation. We are in a good position to “own” the conversation. The PAG will continue this discussion next month.