Mailinglist Archive: opensuse (3232 mails)

[Anders Johansson <andjoh@xxxxxxxxxx>]

On Tuesday 15 August 2006 00:10, Dave Howorth wrote:
> I think Felix had a point. The US post office isn't the author, but
> neither is it the publisher. It's [just] a common carrier. Does the RFC
> recognize the notion of a publisher? Because that's what the list is,
> IMHO.

It's not though. It's a distributor. Hence "distribution list"

I admit that my analogy with the post office was bad, a better analogy would 
perhaps be with the third party subscription services that magazines like 
Time use to pester their subscribers with renewal offers (but also for the 
basic stuff, like keeping track of who has a subscription)

>
> I think most of the debate on this issue misses the point. That is,
> IMHO, that the requirements haven't been clearly analysed and
> consequently the solution isn't especially good!
>
> It's clear that a common carrier (ISP etc) is the wrong entity to
> determine the reply-to address, or anything else. It's equally clear
> that the author needs to be the ultimate determiner of the reply-to
> address.

Yes

>
> The problems come with the in-between cases. Mostly, we'd like the
> 'publisher' to decide the reply-to address. But are they smart enough to
> deal with e.g out-of-office replies? Are they smart enough to deflect
> spam? The fundamental problem, IMHO, is that we need a basic redesign of
> email delivery/routing mechanisms, including lists.
>
> Cheers, Dave