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

> given the rfc2821 language, it is tempting to change the document to use the 
> term "errors".  the problem, then, is that it might be confused with 
> "errors-to".
> but, as I say about all the content of the email-arch document, the deciding 
> factor is any indication of community preference.
"Community preference" for a standards track document should be the 
community of people who read and implement standards, of course; and I'm 
reluctant to speak to that. I live in the world of operations staff and 
highly educated end users, even the best of whom are uncertain about 
which parts of the message are contained in the header and which in the 
envelope. For many years I've tried -- and failed -- to craft language 
that would allow these people to reliably verbalize "RFC2821 MAIL FROM" 
or "RFC 2822 From." Six months ago, we (at Habeas) started using the 
term "bounce address" and the result has been amazing. No one ever needs 
the term exaplained more than once; and from then on the dialog 
continues with no confusion or ambiguity. For the first time in 15 
years, ordinary people can say what they mean and have it understood.
If the architecture doc does not standardize on the term "bounce 
address," it should at least acknowledge that the term is very useful 
for its clarity in ordinary human conversation.
<csg>


//