OIDC Deployment Working Group

[Steven Carmody <>]

I think that suggestion makes sense, for several reasons.

This effort could initiate a self-supporting community around using the 
GÉANT OIDC extension. Consequently, we shouldn't limit publicizing this 
effort to just I2; we should tell the world about it.

I think that having documentation that is at least partially organized 
around use cases (eg this is how to use it with mobile apps, etc) would 
be a tremendous help to campuses! Yes, there's a lot of smart people out 
there. But, they don't have enuf time to figure something out by reading 
technical documentation. They'll start by looking for a pattern.

Lastly, this gives the group something very specific to focus on. For 
this group, at this time, that would be a really big help. We might even 
find volunteers from campuses that are working on different use cases. ;-)

On 1/15/19 11:28 AM, Nathan Dors wrote:
> What do folks on the list think about contributing to a deployment guide 
> for the GÉANT OIDC extension to the Shibboleth IdP, as a focus for early 
> 2019?
>
> On the last few WG calls we reflected on possible work directions for 
> the first part of the year, and folks agreed that a practical near-term 
> need exists for guidance for IdP operators testing, deploying, and using 
> the GÉANT OIDC extension for use by various types of relying parties: 
> everyday web-based apps, but also perhaps single page applications, 
> native mobile and desktop apps, and apps with more limited interfaces 
> like command-line apps. Possibly others too.
>
> Sites testing the OIDC extension have an interest in such a guide, and 
> therefore may be willing to contribute to it, either in the form of 
> content or review and feedback. I know the team here at UWash is interested.
>
> The mechanics of the guidemaking need to be sorted out, and coordinated 
> with Janne, Henri, Shib Dev, et al., and the overall shift in focus 
> needs support from sponsoring bodies like InCommon TAC, CACTI, and other 
> stakeholders... but first, what do you think? Is this the right re-focus?
>
> I know there are a lot of other topics related to OIDC and OAuth that we 
> could cover - indeed, almost too many to cover in a reasonable timeframe 
> - so how about we focus on successful integration of OIDC into IdPs, and 
> then see what makes sense to do next.
>
> -Nathan