From: "William Gallafent" <william@gallaf.net>
On Monday 14 August 2006 10:00, jdow wrote:
It does not matter. RFC 2822 is not a standard. It is only a draft standard that cannot reach concensus.
It's actually a "proposed standard", which is less far advanced than "draft standard" (I think). It is, nevertheless, the best available document describing how things work at the moment.
I checked STD-1. It is in the "draft standard" list. It seems to have been around since God was a child so I suspect it just cannot gather the necessary concensus.
Insisting on people obeying ridiculous rules is stupid.
I agree, and I haven't noticed anybody insisting that anyone else follows any ridiculous rules. The observation has simply been made that this list is configured according to the closest thing to a standard which _does_ exist, and which is likely eventually to become a standard. Many people are using mail clients which are deficient with respect to another long-lived _proposed_ standard (RFC2369), and suggest that the list should deviate from RFC2822 (by setting Reply-To despite not being the author of the message) in order to save them one click until their mail client implements appropriate handling of the RFC2369 headers.
That doesn't make it "right", though. {^,-}
I'm sure well-reasoned suggestions for the modification of the meaning of Reply-To (which would have to include changing it from an originator header to a resent or trace header, I imagine) in future RFCs designed to supersede RFC2822 will be welcomed by those drafting those documents.
A legal analogy: If you don't like a legal bill before it becomes an act, then lobby democratically to have it altered before it's made law; if you don't like a law which has already been passed, then lobby to have it rescinded or superseded; in the mean time, you may choose to break it or abide by it, as may everybody else. The owner of the list has chosen to abide by the closest thing to a law for email that exists so far.
The internet is very young, so many areas do not have any "laws" that have yet been passed; they are still only proposed. It's not "illegal" to deviate from a proposed standard. To avoid chaos, though, and to promote interoperability, it's advisable to abide by the proposed "laws" which do exist, and it's helpful to suggest improvements to them which will ensure that when, in the fullness of time, they become standards, they do indeed reflect the needs and desires of those whom they affect, viz: us. So, suggest away.
I suspect this issue is not particularly well suited to being a "standard" which should be obeyed and probably never will be on a wide basis. Meanwhile I need to fiddle this list in procmail, too. Then what others do with headers doesn't matter. Replies will go to the list as Ghod intended. {^_^} Joanne