Re: [Extra] New Version Notification for draft-bosch-sieve-eai-01.txt

[Ned Freed <[e-mail]>]

> It appears that Ned Freed  <[e-mail]> said:
> >I'm afraid this document has a lot of probems, some of which may not be easy to
> >resolve.
> >
> >The biggest problem by far is that the idea of invoking an extension in order to
> >be able to process an EAI message is intrinsicly nonsensical: A message
> >presented to the Sieve interpreter is either an EAI message or it's not, and
> >there's nothing the Sieve interpreter or Sieve author can do to change that. (It
> >may not even know if a given message was intended to be in EAI format - there's
> >no reliable marker for this.) ...

> Yup.  I think it might be possible to redo this draft as advice on what a sieve
> implementation has to do to handle EAI mail, but your sieve code either handles
> EAI mail, in which case it makes sense to let it process EAI messages, or it
> doesn't, so the behavior with EAI messages is undefined.

I'll be happy to suggest text if there's consensus on it being recast as a client
issue, not an extension.

I do think an extension makes some sense in order to be able to redirect to
an EAI address or initiate new EAI messages.

It might also make sense for this to be split into two documents, one dealing
with compliance, the other with any extension. Or not - I have no real
opinion on the point.

				Ned