Minor point in support of John's earlier statement -- in my experience, the
number of viewers that provide any sort of header filtering is far smaller
than the number that provide no filtering at all. The demand for header
filtering is so great that I have been forced to add support for it in my
MTA code. MTA code is no place to do this sort of thing, but I was willing
to do it, as opposed to any of the UA suppliers, and that automatically
made it my job, not theirs.
I even support header trimming and/or stripping on outgoing SMTP, since so
few of the SMTP setups on PCs and Macs handle headers reasonably.
Headers and header stripping are also a major problem when gatewaying to
environments that do not support the concept of extensible headers at all.
This includes practically every non-RFC822 based mail system, including
those like cc:Mail and MHS, which may look superficially RFC822-ish, but
are not. The usual result is that gateways support various ways of mapping
headers into message bodies or attachments. This is rarerly invertible,
usually a nuisance, and often switched off by sites using the software. Ph
yes, X.400-1988 supports a mechanism that allows headers to be mapped, but
X.400-1984 does not, and most systems in use today are X.400-1984 if they
are X.400 capable at all.
Ned