Re: [art] Auto-configuring Email Clients via WebFinger

["Paul E. Jones" <[e-mail]>]

John,

The first query in my draft is the WebFinger one where the client would receive a list of URLs to then query to get various mail server configurations.

Using my wife's use case, the web server is maintained by a different team than the mail server. The only thing the web server team needs to know is where to direct the client for email configuration for each user, thus one static entry in WebFinger.  I'd expect the team managing mail server to deal with the actual mail configuration data and any authentication requirements.

The example in the draft with two addresses is to show that a single account identifier could have multiple named email accounts behind it. I suspect that would be unusual, but the possibility exists. If people don't want that flexibility, we could require that only one link relation be returned of the mail configuration type.

Does that help clarify my point?

Paul


-------- Original Message --------
From: John R Levine <[e-mail]>
Sent: July 18, 2019 12:23:56 AM EDT
To: "Paul E. Jones" <[e-mail]>
Cc: [e-mail]
Subject: Re: [art] Auto-configuring Email Clients via WebFinger

On Thu, 18 Jul 2019, Paul E. Jones wrote:
>> What are the risks if the mail server and HTTP server are maintained by 
>> different authorities?
>
> I expect they would be in some cases, such as service providers that sell 
> simple web hosting services. That's another reason why people several years 
> ago suggested that we separate the initial inquiry from the follow-up that 
> actually returns the mail configuration data. That's how the example is 
> presented in my draft.

But in your draft, the initial response is often specific to the 
particular address, e.g., it has a personal and business version.  How 
does that help?  It still needs to know something about each address.

Regards,
John Levine, [e-mail], Taughannock Networks, Trumansburg NY
Please consider the environment before reading this e-mail. https://jl.ly