[Top] [All Lists]

Re: [Fwd: I-D ACTION:draft-crocker-email-arch-05.txt]

2006-10-25 13:51:39

John C Klensin wrote:
As for silver gorillas, I'm frankly assuming that the world of
X.400 experience is sufficiently marginalized in the world of
the modern Internet to make that concern minor.

Hmm.   I see symptoms of "X.400 experience" -- or maybe actual
X.400 code-- almost every time I see an error message or NDN
from a Microsoft Exchange server.  While there are many people
in the world of the modern Internet who consider that particular
product marginal in other ways, I'd contend that it is fairly
hard to dismiss it and its impact as minor.

1. Yes, Exchange's code was based on X.400. But I do not understand how the nature of an internal software base is relevant to a discussion of Internet Mail architecture or the role of other global email service architectures.

2. I never said that the impact of X.400 was minor. I said that it was marginalized in today's world. There's a difference. If there is a substantial X.400 community that is active in Internet email discussions, please point them out.

It has been years since I've tried to understand its
documentation, and things may have changed anyway, but I think
one should verify whether it uses ADMD-style terminology and, if
so, how, before trying to reuse that term.

Why should one do that?  Who are you worried will be confused?

-email-arch defines the term. Although I have now explained why I believe it is compatible with how the term was used in the X.400 world, I have some difficulty seeing how close inspection of the point alters its acceptability.

(And by the way, I assume you know that the UA/MTA model that is attributed to the X.400 world, since that is where is got popularized, was in fact developed primarily by folks who did Arpanet mail.)

And, no, I don't have better suggestions, partially because I
continue to believe that your document is an attempt to
retroactively introduce structural models that really don't fit
-- about as helpful in the long term as the many efforts to show
how the Internet fits nicely and smoothly into OSI seven-layer
fairy tale.

Since the exercise of trying to specify -email-arch was done as a direct and adaptive attempt to observe what exists in Internet Mail and then describe it -- whereas the Internet-to-OSI effort was a Procrustean exercise to coerce Internet stuff into the pre-existing OSI model that was developed largely without reference to the Internet stuff -- I do not understand how you can compare the exercises.

But I do get the point that you aren't happy with the document.


  Dave Crocker
  Brandenburg InternetWorking

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