Re: [dane] case sensitivity and draft-ietf-dane-smime / draft-wouters-dane-openpgp

[Paul Wouters <[e-mail]>]

On Mon, 9 Sep 2013, Viktor Dukhovni wrote:

> On Mon, Sep 09, 2013 at 10:52:26AM -0400, Paul Wouters wrote:
>
>> Mail servers and mail clients do not treat email addresses as
>> case-insensitive. When encoding an LHS with base32, the case matters.
>> Using the wrong case will cause you to not find the SMIMEA / OPENPGPKEY
>> record.
>
> This is not accurate.  While domain part of an email address is
> case-insensitive, the case-sensitivity of the local-part  is a
> local matter.  SMTP relays MUST NOT change the case of the local
> part of envelope addresses.  On arrival at the destination domain,
> that domain's mail system may choose to treat local-parts case
> insensitively, but remote systems must not rely on this.
>
> Therefore, <[e-mail]> and <[e-mail]> are a-priori
> distinct addresses except perhaps in the hands of the MTAs that
> handle example.com mail.
>
> [ The fact that most systems are in practice case-insensitive is not
>  sufficient to invalidate the right of some to be case sensitive as
>  they see fit. ]

The _last_ thing I want to do is having an email client not encrypt a
message because someone mailed [e-mail] instead of
[e-mail].

So I think our choices are:

1) do a lookup for the case "as is" and a lookup for lowercased (serial
    or parallel)

2) lowercase and lookup once.

What I do _not_ think is a valid choice is "only do the as is case
lookup".

Paul