Merge pull request #2610 from matrix-org/rav/proposal/remove-oauth2-auth-type
MSC2610: Remove `m.login.oauth2` User-Interactive Authentication type from the specificationpull/977/head
commit
a96a014512
@ -0,0 +1,35 @@
|
|||||||
|
# MSC2610: Remove `m.login.oauth2` User-Interactive Authentication type from the specification
|
||||||
|
|
||||||
|
The client-server API specification
|
||||||
|
[defines](https://matrix.org/docs/spec/client_server/r0.6.1#authentication-types)
|
||||||
|
a number of "authentication types" for use with the User-Interactive
|
||||||
|
Authentication protocol.
|
||||||
|
|
||||||
|
Of these, `m.login.oauth2` is underspecified and of no
|
||||||
|
real use. This MSC proposes removing them.
|
||||||
|
|
||||||
|
## Proposal
|
||||||
|
|
||||||
|
The definition of
|
||||||
|
[OAuth2-based](https://matrix.org/docs/spec/client_server/r0.6.1#oauth2-based)
|
||||||
|
authentication is incomplete. [OAuth2](https://oauth.net/2/) is best considered
|
||||||
|
as a framework for implementing authentication protocols rather than a protocol
|
||||||
|
in its own right, and this section says nothing about the grant types, flows
|
||||||
|
and scopes which a compliant implementation should understand.
|
||||||
|
|
||||||
|
A better candidate for OAuth2-based authentication of matrix clients is via
|
||||||
|
[OpenID Connect](https://openid.net/connect/), but this has already been
|
||||||
|
implemented in Matrix via the `m.login.sso` authentication type.
|
||||||
|
|
||||||
|
The `m.login.oauth2` section is therefore unimplementable in its current form,
|
||||||
|
and redundant. It should be removed from the specification to reduce confusion.
|
||||||
|
|
||||||
|
## Alternatives
|
||||||
|
|
||||||
|
It would be possible to extend the definition so that it is complete: as
|
||||||
|
mentioned above, a likely implementation would be based on OpenID
|
||||||
|
Connect. Matrix clients could then follow the standardised OpenID Connect flow
|
||||||
|
rather than the matrix-specific `m.login.sso` flow. However, this would require
|
||||||
|
significant design work, and development in both clients and servers, which
|
||||||
|
currently feels hard to justify when a working solution exists via
|
||||||
|
`m.login.sso`.
|
Loading…
Reference in New Issue