Comments on draft-fielding-uri-rfc2396bis-07

I'm concerned about some provisions of the draft which seem to
contradict one another and existing practice, specifically
regarding mailto URIs and the (RFC [2]822) special character
'@', which is also a URI reserved character.

Draft section 1.1.2 gives an example:

      mailto:John.Doe@example.com

whereas mailto URIs encountered in the wild tend to look like:

    mailto:uri%40w3.org?Subject=Re%3A%20editorial%20suggestions%20for%20RFC%202396%20bis&In-Reply-To=<0I5U00G08DFGCR%40mailsj-v1.corp.adobe.com>&References=<0I5U00G08DFGCR%40mailsj-v1.corp.adobe.com>

i.e. because '@' is a reserved character and the mailto URI specification
requires that "all URL reserved characters in 'to' must be encoded" (RFC
2368 section 2), one sees "%40" rather than '@'.  In practice, the two
forms -- literal '@' and encoded %40 -- are currently treated as
equivalent by many implementations. [The example above was taken
from the text/plain body of a message posted to the ietf-822 mailing
list; as plain text, it probably shouldn't have "&" in it, but that's
another matter.]

The draft, however, says:

   URIs that differ in the replacement of a reserved character with its
   corresponding percent-encoded octet are not equivalent.
   Percent-encoding a reserved character, or decoding a percent-encoded
   octet that corresponds to a reserved character, will change how the
   URI is interpreted by most applications.

and

   URI producing applications should percent-encode data octets that
   correspond to characters in the reserved set.

and it reaffirms that '@' is reserved (via gen-delims).

If indeed '@' is reserved and URI producers should encode reserved
characters, then the example in the mailto URI should be 

      mailto:John.Doe%40example.com

The statements regarding non-equivalence and change of
interpretation appear to conflict with existing practice.

Received on Monday, 1 November 2004 16:41:00 UTC