Frank,
On 2008-04-06 03:29, Frank Ellermann wrote:
Brian E Carpenter wrote:
My underlying concern is that 2822upd should not appear
ridiculous to anyone who looks at a typical mail header
and sees the X-headers.
2822upd specifies only about twenty mail header fields.
The rest is either registered and specified elsewhere,
or to some degree "unknown" (unregistered, unspecified,
proprietary, private use, experimental, spam, and so on).
Sure. But X- is described *nowhere* in current documents;
the one example in provisional registration is marked as
"deprecated" on no basis whatever (there is nothing in
2822 or 2822upd or 3864 to suggest a "deprecated"
registration), and...
John's message reached me with X-Original-To,
A draft about Original-* didn't make it so far, ...
I don't care about the details of unregistered headers.
I simply want to see a statement of reality in 2822upd.
It can be simple or complicated but IMHO it needs to
say
X- headers are in widespread use.
They are not part of this standard.
They must not be relied on to have any particular semantics
or even to be delivered.
OK, I've said all I want to say on this.
Brian
_______________________________________________
IETF mailing list
IETF(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf