[Asrg] where the message originated (was: DKIM role?) (SM)

[Ian Eiloart]

--On 21 January 2009 08:12:09 -0800 Dave CROCKER <dhc at dcrocker.net> wrote:

>
>
> Ian Eiloart wrote:
>>> It's not interesting who publishes SPF records.
>>>
>>> It's interesting who reads them.
>>
>> That depends on whether you're deciding whether to publish them, or
>> whether to read them.
>
>
> For the purpose of assessing Internet utility, consumption is a better
> indicator than publication.

Like I said, you need both. But, I'm not interested in "Internet utility", 
whatever that might be. I'm interested in utility to me. I presume that 
what's useful to me might also be useful to some others.

> I understood this thread to be discussing utility.

There are many kinds.

> Publication without consumption has zero utility.  So does consumption
> without publication, of course, but if there is a community wishing to
> consume, there will almost certainly come to be a community doing the
> publication.

Except that there's no obvious way to advertise the fact that you're 
consuming. We know that there's widespread publication. We don't know 
whether there's widespread consumption - unless someone has stats gathered 
from DNS servers. Even if we had stats from DNS servers, it's hard to know 
what anyone is doing with the information. Widespread selective 
whitelisting would be hard to detect, for example.

>
> In contrast we've had ample demonstration that that publication does not
> have any automatic implication about consumption.
>
> d/



-- 
Ian Eiloart
IT Services, University of Sussex
x3148