feat(rfcs): Add RFC to describe the protocol for parsec-auth

[FirelightFlagboy]

No description provided.

[AureliaDolo]

It's cumbersome to read but it would be nice to have the same format as in others RFCs (like here) aka the json5 format

[touilleMan]

How the client knows the auth medium ID that should be provided here ?

[mmmarcos]

I guess a new route is needed to "list" the registered auth mediums methods. That would seem consistent with the proposed routes to "add" and "remove" an auth method.

BTW, based on this comment, I would rename this route to be about the account manifest (and not the auth medium), so i suggest:

• account_manifest_get (this route, to get the account manifest)
• auth_method_add (already described below)
• auth_method_remove (already described below)
• auth_method_list (route to be added below, allows to get the auth medium method ID )

[FirelightFlagboy]

auth_method_list could not be used to get the ID since it will be an authenticated route whose require AuthMediumID to generate the HMAC signature

[FirelightFlagboy]

It was stated in the prior RFC that AuthMediumID is generated from the secret

[touilleMan]

This route is no longer useful since those data should be store in the account manifest

[mmmarcos]

All the encrypted devices would be stored in the account manifest? So is the client that directly stores that in the manifest?

If that's the case there is no need to "list" or "get" devices, but there is also no need to "upload" them either??

[FirelightFlagboy]

We still need the list for the revoked_on field, current the manifest only store symmetric key to decrypt the devices.

Should we change that?

[mmmarcos]

All the encrypted devices would be stored in the account manifest? So is the client that directly stores that in the manifest?

If that's the case there is no need to "list" or "get" devices, but there is also no need to "upload" them either??

[FirelightFlagboy]

Superseded by #10090