> On Oct 24, 2017, at 2:57 PM, Ivan Ristic <ivan.ris...@gmail.com> wrote:
>
> On Tue, Oct 24, 2017 at 7:47 PM, Viktor Dukhovni <ietf-d...@dukhovni.org>
> wrote:
>
>> An MTA is far more heavy-weight infrastructure component than
>> a website. Sure you can start a Web business on someone else's
>> shared platform, but running email hosting on someone else's
>> virtually hosted MTA is entirely unrealistic.
>>
>> MTAs both send and receive email, they run complex anti-spam
>> and anti-virus filters, they are integrated with mailbox
>> stores, they have IP reputations as sending systems, they
>> do DKIM signing, add Authentication-Results headers, store
>> and forward email, ...
>>
> That's exactly why it's a great candidate for outsourcing. I will let
> someone else do all of that techie stuff, and I'll focus on growing the
> business.
Hosting email is *all that techie stuff*. Email is infrastructure.
There's little else you offer by rebranding someone else's
infrastructure. They can just sell the service directly and cut
out the unnecessary middle-man.
One might try to offer enhanced security, like say protonmail.ch,
but that's not possible on a shared platform... Otherwise, email
is a commodity.
>> Mere SNI will not come remotely close to giving you a virtual
>> MTA. An MTA is NOT a website.
>
> No, SNI will give me freedom to migrate my stuff if I want to.
> If I can't do that, I have no long-term business.
In any case this business model is already dead in the water
given the market dominance of the existing hosting players,
who don't have to pay a third party for infrastructure.
An MTA requires a dedicated IP outbound to avoid reputation
damage from co-hosted miscreant third-parties. Once you have
that, you may as well have a dedicated inbound IP and the need
for virtual hosting goes away.
Let's get back to hearing actual, not straw-man use-cases...
--
Viktor.
_______________________________________________
Uta mailing list
Uta@ietf.org
https://www.ietf.org/mailman/listinfo/uta
