Re: [Extra] Is this a plausible IMAP extension ?

["Bron Gondwana" <[e-mail]>]

On Thu, Feb 28, 2019, at 04:35, Ned Freed wrote:
> > We do something like this already at FastMail, and here's the thing.... the
> > user should be allowed to add or clear this header to change the interpretation
> > of the email in their mailbox. Once they have the message - they should be able
> > to update the BIMI status.
> 
> This seems entirely reasonable to me, but now we're talking about updating the
> message content. This pretty much pushes it out of the header space as 
> far as I'm concerned.

I corrected that almost immediately - but, yay immutable emails. I meant a keyword that they could update, not a header. Sorry.

So the header gets added at delivery time, the same way an Authentication-Results header gets added, saying that DKIM / SPF passed or didn't pass. Fine.

A keyword gets added to the message on delivery if that header is present. Fine.

The keyword has a restricted ACL such that the user can't modify it afterwards. Not OK. Dislike. Unnecessary and likely to cause many issues. The "users could be phished to change the keyword" is an implausible argument in any mass market client that I've seen, and the side effects of trying to make read-only keywords are significant.

> > This is particularly important when you're talking about importing and
> > exporting email between systems. Having the flag to avoid phishing, sure. But
> > restricting MUAs from setting that flag - that's bogus.
> 
> Exactly.
> 
> > I guess I'm going to the BIMI session.
> 
> Sounds like a good idea.

Fun times :)

I've been tracking BIMI for a couple of years anyway, so it's not like I was ever NOT going to go.

Bron.

--
 Bron Gondwana, CEO, FastMail Pty Ltd
 [e-mail]