ietf-mta-filters
[Top] [All Lists]

Re: Working Group Last Call: draft-ietf-sieve-editheader-00.txt

2005-04-05 05:09:17


On Tue, 5 Apr 2005, Arnt Gulbrandsen wrote:


Alexey Melnikov writes:
Cyrus Daboo wrote:
>  For the purpose of weeding out duplicates, a message modified
>  by addheader or deleteheader MUST be considered the same as
>  the original message.

Hmm, even if the scripts replaces the Message-Id header ;-)?

It seems to me that banning removal/addition of certain fields simplifies life a lot without removing significant benefit: Message-ID, Content-Type, Content-Transfer-Encoding and MIME-Version.

Modifying those four fields adds so many corner cases to the implementations. Do we have to go there? Is it perhaps better to just say "editheader operations MUST NOT modify those four fields"?

This is definetely better. I've not seen those header fields ever being modified by intermediate MTAs (forwarders, mail lists, etc) nor is there any reason why they should by MDAs or their agents. The only thing is that
Content-Transfer-Encoding may in theory be subject to changes if MDA wants
to decode BASE64 into something more readable for MUAs (in theory it should not try though).

--
William Leibzon
Elan Networks
william(_at_)elan(_dot_)net