@levie If you want to avoid your users having to click through a bunch of OAuth consent screens when connecting all these data sources, give this a read and get in touch https://t.co/ORxsblYJck
@tweetsbycolin@zeeg Yes, it's been there forever, but rarely used. There's also step-up auth RFC 9470, not about scope, but about other attributes https://t.co/wZ00n8zuFR
@tweetsbycolin Hey there, sorry I was out of town last week and didn't see this until now! I have some ideas for how to improve this, what's the best way to discuss?
@samuelgoto@Paul_Kinlan@mfosterio oh I bet the browser is blocking it. If you've ever clicked the little X it blocks it from working until you go unblock it. Go to this page and I bet you'll see "automatically blocked third-party sign-in" chrome://settings/content
@samuelgoto Yeah, I've been following that, and think it's a great idea. If that ships, I would then 100% recommend people use a BFF to manage OAuth tokens and a device-bound cookie to their BFF. https://t.co/GZy34yxXUZ
@samuelgoto Yes this is still a gap. Non-exportable WebCrypto keys are a good start, since you can use them to sign access token requests using DPoP, making token exfiltration less of a problem.
@ottokruse That's the TLDR. The longer version is explaining the intricacies in how the implicit flow relates to OpenID Connect and all the combinations of response types and response modes
@jakesloaninak tbh go look at the comments from large channels on related videos that you'd do and see what the viewer community is like.
I should have done that before I made that Remote ID video because 😬😬😬 notice how I haven't posted another video about that since 😬