On 11/17/25 20:20, Murray S. Kucherawy wrote:
> On Mon, Nov 17, 2025 at 11:27 AM Daniel K. <[email protected]> wrote:
> 
>>> At this point I don't think the validity of an unencoded '+' is
>>> disputed. The problem seems to be that the consumer of the DMARC record
>>> does not do any pct-decoding at all, leading to the misdirected reports.
>>
>> Or, reading over and thinking about it again, maybe the consumer is just
>> being selective about what to decode, since '%' itself is in 'atext'.
>>
> 
> I think "+" as part of an email address, when used in a "mailto" URI, must
> be encoded because it's a reserved character in both RFC 3986 and RFC
> 6068.

I'm not so sure.

Per RFC 3986, [email protected] is a path, and quoting the relevant
ABNF from that document:

      path includes path-rootless as an alternative

      path-rootless = segment-nz *( "/" segment )

      segment-nz    = 1*pchar

      pchar         = unreserved / pct-encoded / sub-delims / ":" / "@"

   sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"
                 / "*" / "+" / "," / ";" / "="

As sub-delims is different from pct-encoded, I conclude that '+' must
not necessarily be pct-encoded.


> Interoperability, then, is only guaranteed when all consumers do the
> decoding.

Agreed, any char *could* be pct-encoded, not just the chars outside of
unreserved, reserved, etc. and the implementation should be prepared to
encounter and handle, e.g.

        %6c%6f%63%61%6c%2b%70%61%[email protected]

just as well as the more human readable version at the start of my reply.


Daniel K.

_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to