Re: CONPERM/CONNEG

Tamura-san,

Sunday, February 9, 2003, 6:10:35 PM, you wrote:
>> However there is an important question about the complexity of
>> implementing this specification.
HT> Yes, very complex.
HT> This might be one of methods that are satisfying to mail experts.
HT> How about cc'ing to ietf-smtp ML?

Yes, the specification needs review from more people in the IETF SMTP
community.  I thought that it would be better to obtain a review from
the iFax working group, first.  Then we can ask the SMTP community to
take a look.

(We need to add changes that have already come from some additional
comments, before we seek this wider review. I will try to do the
revision in the next few days.)

I am hopeful that the current revisions will be acceptable to the email
community. In doing the current revisions, the comments from Freed-san
and Moore-san were studied very carefully. Also, the guidelines from the
OPES effort (for intermediaries) was studied.

I believe the current specification responds to all their concerns.


HT> Questions.
HT> I do not know well what contents are transmitted
HT> when SMTP Relay server converts images.
HT> Does it transmit Content-Converted header and
HT> converted images in the same MIME part?

I will change the text to make this more clear.

The specifications is intended to direct that each MIME body-part has
its own set of Content- headers (Convert, Converted, Current).  So there
is one set for each MIME body-part.

HT> What if the conversion happens several times?

Excellent! You have anticipated an enhancement that is in the draft I am
currently revising.  It provides for a sequence of conversions.

The Content-Converted headers are designed to be similar to the Received
header, partly for this reason.

I think you are the first person to suggest that.


HT> I think Appendix A, which is the original proposal,
HT> should go to the main body, taking the section 8 or 9.
HT> It is easier to implement than the relay-case.

Yes, it is certainly the mode of operation that is most likely for iFax
devices.

However the previous concerns from the email community require making
the specification work for "regular" email, I think.  This means that
the "direct mode" operation must be specified as a simplified case.
That is why it is in the Appendix.

Because we know that the Appendix will be popular, we should make sure
that it is especially easy to understand.  Please suggest any
improvements.

d/
-- 
 Dave <mailto:[email protected]>
 Brandenburg InternetWorking <http://www.brandenburg.com>
 t +1.408.246.8253; f +1.408.850.1850