Merit Network
Can't find what you're looking for? Search the Mail Archives.
  About Merit   Services   Network   Resources & Support   Network Research   News   Events   Home

Discussion Communities: Merit Network Email List Archives

North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: Micorsoft's Sender ID Authentication......?

  • From: william(at)elan.net
  • Date: Tue Jun 07 16:04:53 2005


On Tue, 7 Jun 2005, Fergie (Paul Ferguson) wrote:

Wasn't there a lot of turmoil within the IETF last year
on sender authentication because Microsoft was trying to
push it's own sender ID authetication mechasnisms as a
draft standard?
That time it did not work. But they are still trying to push it as experimental RFC and in the mean time continue to advocate it trying
to make it open standard despite technical and other objections.

Or maybe I'm confused...
SenderID is bad technology. Technically:
1. Its proposal contains recommendation that are directly opposite
to existing RFC2822 standard (which specifically mentions in section
section 3.6 that "using resent fields is a different operation from
forwarding")
2. It tries to change the meaning of message sender from real message
source to intermediate agents, this would have problems with other
email authentication technologies
3. Its using spf1 records, but they are used for information on sources
for different identity (SMTP2821 MAIL FROM) and would not always have
the same data as what would apply for RFC2822 Sender or its
derivatives. There is no consent being asked from those entered
SPF1 records if they meant them used for SID and this basically
means people who wanted to participate in SPF inadvertently are
put in danger (it can also be viewed as attempt to steal somebody
else's record).

For more on technical issues see my recent message to IESG regarding SID when they were evaluating it:
http://www.gossamer-threads.com/lists/spf/discuss/19859

Politically it has problems too:
1. Its patent license is not open-source compatible and so can not
be used in any GNU or BSD licensed program
2. The "senderid" word itself is trademarked and its use also
basically being stolen by Microsoft

---

Since it appears NANOG continues to be used for mail-related discussions
and a some of what goes here is based on not understanding technologies
and issues involved, I'll make a link to a paper that I'm working on available (when its ready) and it will hopefully be good information to understand what's up in email authentication front and what each technology can and can not do.

--
William Leibzon
Elan Networks
william@elan.net




Discussion Communities


About Merit | Services | Network | Resources & Support | Network Research
News | Events | Contact | Site Map | Merit Network Home


Merit Network, Inc.