Updating RFC 2368 (mailto: URI)

[Martin Duerst <duerst@xxxxxx>]

This is an issue that I don't think has been raised yet.

RFC 2368 (http://www.ietf.org/rfc/rfc2368.txt), "The mailto URL scheme"
(Hoffman, Masinter, Zawinski) defines the details of the 'mailto:'
URI scheme.

We should think about how this is affected by IMAs.
In particular, with IRIs, I think that users will expect that
they can write
    <a href='mailto:IMA@xxxxxxx'>Comments about my page</a>
(where IMA and IDN are non-ASCII characters). The IRI-to-URI
conversion converts the above to
    mailto:%II%MM%AA@%II%DD%NN.tld
where %II,... is the %-hex-escaped UTF-8 encoding of the original
characters. So we have to extend the definition of mailto: URIs
to include this case, and we have to make sure that this maps
to IMAA. This mapping will turn out to be more or less complex
depending on how IMAA looks. For IMAA-ACE, we will have to
describe the details of the conversion (similar e.g. to those
in http://www.ietf.org/rfc/rfc2192.txt), and they will have
to be implemented.

For 'UTF8ADDRESS' (I changed that from plural to singular,
but I think that there is still room for improvement), it
will be more straightforward.

This of course would not restrain the ACE-initiated to
use e.g. an ACEd IDN directly. But this would confuse
the users.


While we are it, some more comments on RFC 2368 and internationalization.
Here is some text that I found:

   8-bit characters in mailto URLs are forbidden. MIME encoded words (as
   defined in [RFC2047]) are permitted in header values, but not for any
   part of a "body" hname.

This is not completely clear. Is it forbidden to use 8-bit octets
in mailto URIs in raw form? The general URI spec would definitely
forbid this, so this would just be a repetition.

Or is it forbidden to include such octets %-escaped? In that case,
it would be impossible e.g. to create a mailto: URI that sends a
PNG image to somebody (for example, not that that's frequently done).


Regards,    Martin.