Re: [apps-discuss] Mail client configuration via something, maybe WebFinger

John C Klensin <> Wed, 10 February 2016 18:07 UTC

Return-Path: <>
Received: from localhost ( []) by (Postfix) with ESMTP id 6AD4C1B2E5B for <>; Wed, 10 Feb 2016 10:07:09 -0800 (PST)
X-Virus-Scanned: amavisd-new at
X-Spam-Flag: NO
X-Spam-Score: -1.901
X-Spam-Status: No, score=-1.901 tagged_above=-999 required=5 tests=[BAYES_00=-1.9, RP_MATCHES_RCVD=-0.001] autolearn=ham
Received: from ([]) by localhost ( []) (amavisd-new, port 10024) with ESMTP id Lkd43ZsluahQ for <>; Wed, 10 Feb 2016 10:07:07 -0800 (PST)
Received: from ( []) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by (Postfix) with ESMTPS id A0E001B2E59 for <>; Wed, 10 Feb 2016 10:07:07 -0800 (PST)
Received: from [] ( by with esmtp (Exim 4.82 (FreeBSD)) (envelope-from <>) id 1aTZAQ-000DX9-Av; Wed, 10 Feb 2016 13:07:06 -0500
Date: Wed, 10 Feb 2016 13:07:01 -0500
From: John C Klensin <>
To: Doug Royer <>,
Message-ID: <>
In-Reply-To: <>
References: <20160210025935.98561.qmail@ary.lan> <>
X-Mailer: Mulberry/4.0.8 (Win32)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-SA-Exim-Scanned: No (on; SAEximRunCond expanded to false
Archived-At: <>
Subject: Re: [apps-discuss] Mail client configuration via something, maybe WebFinger
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: General discussion of application-layer protocols <>
List-Unsubscribe: <>, <>
List-Archive: <>
List-Post: <>
List-Help: <>
List-Subscribe: <>, <>
X-List-Received-Date: Wed, 10 Feb 2016 18:07:09 -0000

--On Wednesday, February 10, 2016 10:05 -0700 Doug Royer
<> wrote:

> (1) I just checked GMAIL, YAHOO, AOL, HOTMAIL, and my ISP.
> Only GMAIL seems to -mostly- provide the correct information.
> Gmail SRV returns:
> 86400 IN SRV 5 0 587
> Yet I have to use port 465 (not 587 in the SRV record). 587
> simply fails for me - I did not care why.
> (2) Some ISP's want your email address in all lower case when
> authenticating, others want it as you entered it. When I setup
> with my isp, it asked what email address I wanted: I entered
> "DougRoyer", so that is how I entered my account name. I have


I learned a few things from your note, most of which, sadly, I
had already guessed at.

One is a reminder about where at least some of the complexity of
ACAP comes from (I think there is more than is necessary, but
that is another question).  Unless we are going to change a lot
of habits, conventions, and operational decisions that make
perfectly good sense in their own environments, we need a system
that is flexible enough to deliver configuration information to
different users (and their MUAs/ Clients) based on location
(sometimes "normal", sometimes "right now"), on user or client
identifier (which may or may not be connected to a mailbox name
in an obvious way), on properties of the client software, etc.
In some, probably most, cases, the server from which the
information comes will make the decisions and send out only the
specific information needed to configure the client but there
will be cases where it would be better to send out information
that gives the client choices that only it can evaluate.  For
some situations, it will need to receive the request and send
the response in encrypted form; for others that may be
unnecessary, undesirable, or worse.  To the extent to which the
location or client software are involved in the decision about
what to send out, whatever is done will have to work despite our
best efforts to obscure the information in the interest of
privacy, leading to some of the use cases for sending out a lot
of information and letting the client choose.

Now, while I still think ACAP is too complex and probably
includes fundamental design flaws, the above fairly strongly
suggests that a DNS record or two, or even invocation of
WebFinger or something else popular, might be part of the
solution, but are, at best, only a small part... and might
encourage sloppy thinking that fails to cover a lot of important
cases.  If we do want to include those cases, this isn't going
to be an easy problem, and that brings us back to John Levine's
question about inertia (among others).