[Top] [All Lists]

Re: pseudo LAST CALL - draft-crocker-email-arch-04.txt

2005-03-30 20:15:16

Sorry if there's some duplicate stuff in here.  I did read the entire
thread so far, but didn't memorize it.  Without trying to nitpick
too much, here are some comments.

2.2.4  Gateway

   > Differences between mail services can be as small as minor syntax
   > variations, but usually encompass significant, semantic distinctions.
   > For example, the concept of an email address might be as different as
   > a hierarchical, machine-specific address versus a flat, global name
   > space.

I think you mean "concepts of an email address" not "the concept of"

But then:

   > Or between text-only content and multi-media.  

Here I think you must be referring to "differences between mail
services" and not to "concepts of email addresses" -- because if it's
the latter it really doesn't make sense.  If it does make sense, maybe
it needs to be said differently.

   > In particular, can a Recipient's
   > MUA automatically formulate a valid Reply?

... that will reach the original Sender.   (the reply can certainly
be formulated; the word "valid" doesn't really imply that it will go
to the sender..)

3.1  Mailbox Addresses

   > A mailbox is specified as an Internet Mail address <addr-spec>.  It
   > has two distinct parts, divided by an at-sign ("@").  The right-hand
   > side contains a globally interpreted name for an Administrative Unit.
   > Domain Names are discussed in Section 3.2.

This is better than before, but I still think it's funky.  The right
side of an Internet Mail address is an Internet Domain Name.  That
domain name does directly lead to the "Administrative Unit" but it
really isn't the name for the AU.  There can be a many-to-one mapping
of domain names to AU, but this does not make one a synonym for
the other.  JMHO.

This is compatible with the mention in 3.3 that the right-hand side of a
message-id specifies the AU assigning the id.  (Even that's a a bit
inexact: however it's better, as it merely says that this "specifies"
the AU and isn't the "name for" the AU -- again the difference is
between a mapping-to and a name-of).

4.1  Message

   > Internet Mail has distinguished some special versions of messages,
   > for exchanging control information:


   > Message Filtering (SIEVE):

Other controversies about Sieve aside, this is not a special version of
a message, and doesn't belong in this list.  I suppose it could be a
Relay or Mediator function.

4.3  Mail Submission Agent (MSA)

   > RFC2821.HELO or RFC2821.EHLO
   >    Set by: Source
   >    The MSA may specify its hosting domain identity for the SMTP HELO
   >    or EHLO command operation.

The HELO name is offered to the MSA, not specified by the MSA.  Maybe
I don't get what you mean here.

(repeat this for 4.4 also)

5.  Mediators

   > RFC2821.HELO or RFC2821.EHLO
   >    Set by: Source or Relay
   > The MSA may specify its hosting domain identity for the SMTP HELO
   > or EHLO command operation.

Same comment as above, but I'd also add that this doesn't seem
relevant here.

5.1  Aliasing

   > What is most distinctive about this forwarding mechanism is how
   > closely it compares to normal MTA store-and-forward Relaying.

This is the first mention that this might also be known as a "forwarding
mechanism" -- actually it's confusing because earlier in the draft there
has been other text describing various forwarding actions that don't
resemble this kind of aliasing.  

Also, you might want to mention that this use of "aliasing" is different
from a fairly common use of that word to describe admin-level list
expansion (e.g. the /etc/aliases sendmail model).  Unfortunately there's
a lot of overlap between words commonly in use, such as "alias" and
"forward" and "bounce" and "resend" and "redirect."  Me, I wouldn't use
"alias" here since it already has a lot of other meaning and I think
this use contradicts what a lot of people already think "alias" means.
"Forwarding" is closer since there's the ".forward" model out there that
you seem to be describing here, except that the word "forwarding"
implies different things in different contexts.  Actually your
"re-address" term is a good way to think of it, too.

5.2  ReSending

   > Also called ReDirecting, ReSending differs from Forwarding by virtue
   > of having the Mediator "splice" a message's addressing information,

Mention of "Forwarding" without prior establishment of that term.


<Prev in Thread] Current Thread [Next in Thread>