|
|
|
@ -274,7 +274,7 @@ includes device ID and keys:
|
|
|
|
|
}
|
|
|
|
|
```
|
|
|
|
|
|
|
|
|
|
16. If doing E2EE then existing device then waits for up to X seconds for the
|
|
|
|
|
16. If doing E2EE then existing device then waits for up to X seconds for the
|
|
|
|
|
`device_id` to become visible.
|
|
|
|
|
|
|
|
|
|
17. If the device is visible within the time period then the existing device
|
|
|
|
@ -348,7 +348,7 @@ It is anticipated that a variant of this would work with OIDC in future
|
|
|
|
|
identified by a different `protocol` value.
|
|
|
|
|
|
|
|
|
|
To be clear, it is not proposed that a OIDC protocol is defined at this time,
|
|
|
|
|
but instead this is included here to help get comfortable that this proposal is
|
|
|
|
|
but instead this is included here to help get comfortable that this proposal is
|
|
|
|
|
reasonably future proof.
|
|
|
|
|
|
|
|
|
|
Whilst not fully thought through it could work like this:
|
|
|
|
@ -473,7 +473,8 @@ the rendezvous checksum; the OIDC device authorization grant code.
|
|
|
|
|
There are two scenarios in mind:
|
|
|
|
|
|
|
|
|
|
1. Showing a "widget" within an existing Matrix client embedded as an iframe/WebView
|
|
|
|
|
2. "Popping out"/launching a new Matrix client from inside of an existing Matrix client
|
|
|
|
|
2. "Popping out"/launching a new Matrix client from inside of an existing Matrix
|
|
|
|
|
client
|
|
|
|
|
|
|
|
|
|
It could be that the two are go hand-in-hand: a widget could be visible in
|
|
|
|
|
embedded mode and a button offered to "open in new window". When clicked a new
|
|
|
|
|