[Standards] Fwd: Re: [domainrep] scope question - what apps?
stpeter at stpeter.im
Mon Aug 16 19:07:05 UTC 2010
FYI regarding XEP-0275 (from a new IETF list about "domain reputation")...
-------- Original Message --------
Subject: Re: [domainrep] scope question - what apps?
Date: Fri, 13 Aug 2010 12:35:51 +0200
From: Alessandro Vesely <vesely at tana.it>
To: domainrep at ietf.org
Peter Saint-Andre wrote:
> On 8/12/10 4:35 AM, Alessandro Vesely wrote:
>> Many services authenticate users based on their email addresses.
>> Perhaps, some domains would accept responsibility for what their users
>> do, even for activities not directly related with the ability to send
>> mail through the domain's MSA. It looks an intricate policy issue to
>> me. At any rate, classifying domains according to how they grant new
>> accounts may be an appealing task.
> For email services at a domain, yes. This is why I think the right level
> is the application service, such as email, IM, or VoIP, because the
> dimensions along which to measure a domain vary by application.
Along that line, for HTTP, admin-edited content may be considered
than users' blogs.
There is a positive range when reputation measures a domain's ability at
controlling their users.
> For example, in the XMPP community we've sketched out some factors that
> make sense for us but probably don't make sense for email providers:
What is named "Server Reputation" in that document, may well be called
"Domain Reputation" in this list's context. I think that sketch is a good
example of how reputation can be measured. Of course, some criteria are
application specific. The score is set up so as to suggest that scores
are added (or have to be added by the client) to the generic
As a minor point, I'd prefer logarithmic scales. For example, there is no
substantial difference between seven and six years of online service, while
the difference between one and zero is very relevant.
domainrep mailing list
domainrep at ietf.org
More information about the Standards