Re: [OAUTH-WG] OAuth 2 delegation flow names

I'd strongly prefer them to be named based on where they're largely
intended for us. "Polling" has become a bad word especially with the
advent of PubSubHubbub and doesn't describe that it's meant for us on
hardware devices.

--David

On Tue, Jun 8, 2010 at 11:16 AM, Marius Scurtescu <mscurtescu@google.com> wrote:
> Hi,
>
> I find the names of the user delegation flows a bit misleading. These
> flows are currently named: "User-Agent", "Web Server" and "Device".
> The names are pointing to the typical client for these flows, but
> these are not the only use cases, and this is where it can be
> misleading.
>
> For example:
> - "User-Agent" can also be used use by native apps and web apps
> - "Web Server" can also be used by JavaScript based clients and native apps
> - "Device" can also be used by native apps and web apps
>
> Instead of naming them after a typical client, maybe we can name them
> based on some technical characteristics of the flow.
>
> The "User-Agent" flow is characterized by the fact that the access
> token is returned directly to the client, no verification code step is
> used.
>
> The "Web Server" flow is characterized by the fact that a verification
> code is first returned which then needs to be exchanged for tokens
> with a direct call from client to authz server.
>
> The "Device" flow is mainly characterized by the fact that a polling
> mechanism is used to retrieve the tokens.
>
> How about the following names:
> - "Web Server" -> "Verification Code"
> - "Device" -> "Polling"
>
> Not sure about the User-Agent flow. Since this flow does not have any
> direct calls from client to authz server, everything is passed through
> the browser, "User-Agent" could be the right name?
>
> Thoughts?
>
> Marius
> _______________________________________________
> OAuth mailing list
> OAuth@ietf.org
> https://www.ietf.org/mailman/listinfo/oauth
>