Meng,
MWW> I agree that a relatively freeform "extra" field type is
MWW> desirable; with such a field type, it will be easier to
MWW> experiment with future extensions before standardization.
The history of x- fields from RFC733/RFC822 shows the flaw in that
approach: If the field is of any use, then it gets popular.
It is extremely rare that folks then migrate from the original header to
a header with a different name.
MWW> To ensure that this field doesn't become the "TXT" of ESMTP,
MWW> perhaps we should define a namespace under which all
MWW> extensions not recognized should be ignored.
In other words, you want to standardize the experimental field, by
creating a new administrative effort.
It would be far simpler to just use the existing administrative effort.
This has the benefit of requiring no future transition for the installed
base.
d/
--
Dave Crocker <mailto:dcrocker(_at_)brandenburg(_dot_)com>
Brandenburg InternetWorking <http://www.brandenburg.com>
Sunnyvale, CA USA <tel:+1.408.246.8253>, <fax:+1.866.358.5301>