Re: [EAI] I-D Action: draft-ietf-eai-rfc5336bis-12.txt

[Chris Newman <[e-mail]>]

--On August 18, 2011 1:21:10 +0200 Frank Ellermann 
<[e-mail]> wrote:

> On 18 August 2011 00:04, Chris Newman wrote:
>
>> So we need:
>
>>  esmtp-value  =/  UTF8-non-ascii
>
>> added to the ABNF.
>
> My personal EAI 2011 experiment, can I grok what folks are talking
> about if it is not related to Message-IDs...
>
> RFC 5321 says that <esmtp-value> follows the RFC 3461 <xtext> syntax
> for addresses, and I-D.eai-rfc5337bis specifies <utf-8-addr-xtext>
> for this job.  That is hex. encoded UTF-8, or in other words ASCII.

That comment in 5321 is confusing and unnecessary. If we were to ever 
revise 5321, I'd support removing it.

The "utf-8-addr-xtext" in RFC 5337 has a misleading name as it is unrelated 
to the 3461 xtext construction. I suggested renaming it to 
"utf-8-addr-7bit" in 5337bis (on July 28).

> With <UTF8-non-ascii> in <esmtp-value> that has (obviously) to be
> limited to states when UTF8SMTPbis is already negotiated.  And if
> that is the case using <utf-8-addr-xtext> outside of ORCPT could be
> confusing.  If it is not only me who is confused, please add also
> an explanation in I-D.eai-rfc5336bis when you update <esmtp-value>.

I see no confusion on this matter in 5336bis. When UTF8SMTPbis is 
negotiated, the ABNF in 5335bis/5336bis applies. When it is not negotiated, 
the ABNF in 5321/5322 applies.

> Examples in the direction of "do this if outside of UTF8SMTPbis" vs.
> "do that in UTF8SMTPbis" maybe followed by a BUT for ORCPT could be
> good.

5336bis has to describe what is allowed in ESMTP parameter values when 
UTF8SMTPbis is negotiated. But discussion of what goes in ORCPT belongs in 
RFC 5337bis.

		- Chris