Slashdot Mirror
MS Releases License For Sender-ID
NW writes "Microsoft published today a new license and FAQ for Sender-ID anti-spam standard being developed by the IETF's MARID WG (based on SPF). To use the license, a signed agreement with MSFT is required. Compatability with the Open Source Definition, the
Free Software Definition, the Debian Free Software Guidelines, and the GPL/LGPL licenses is already in question."
How long will it be before you have to have a signed agreement with Microsoft to send an email?
Seriously, does Microsoft think that highly of themselves that they can dictate a standard that requires a license from them? Sure they have a majority of the OS and E-mail clients, but I doubt they have a majority of the Mail Servers out there.
In the current climate you could never produce a HTTP/SMTP type protocol because everyone is out to make money and gain power. What Microsoft has done is take a relatively open protocol and slapped a 'Microsoft Property' sticker on it, this will effectively limit its usefulness even if they are not charging a penny.
What is stopping them from letting it catch on and then asking for $1 from each project?
Just to get everyone up to speed:
- SPF (http://spf.pobox.com) is the current email authentication protocol that is dominating the world.
- Microsoft proposed Caller ID which was never accepted by the community.
- Microsoft and SPF advocate Meng Weng Wong brokered a deal and formed Sender ID. Basically, SPF is intact, but some features of Caller ID are preserved as an optional extension.
The part of Caller ID that remains is the PRA or "Purported Responsible Authority". The PRA is deteremined by a complicated algorithm that I personally don't believe would work. The algorithm is intentionally vague in some areas, and the results are ultimately subjective. The intention of the PRA algorithm is to determine who wrote the email based on the email headers. As everyone knows, the email headers are spoofable. But the idea goes, if you can track down the PRA, then you can authenticate this email based on that, rather than just the last hop like SPF does.
The problem from day one has been the patent issue. Microsoft is in the process of patenting the PRA algorithm. This isn't a problem. The problem is that Microsoft refuses to put the patent in the public domain or license the patent such that anyone can use it except those who use patents against Microsoft. Both of those strategies are perfectly reasonable, and are pretty much what IBM does for most of its patents.
Microsoft originally wanted to get a copy of the software and a signature before they would grant a license. Well, that doesn't work for F/OSS. The MARID working group who is investigating various solutions to the email authentication protocol for the IETF has been petitioning Microsoft to revise or clarify their licensing procedure. Well, they finally have, and in so doing they have not made it F/OSS compatible.
Microsoft thinks they can bully us around, but they don't realize they are the small kid on the email block. Their Caller ID failed. Now Sender ID is going to fail because Microsoft refuses to participate.
But that's okay. The PRA algorithm isn't anything we'll need to solve the email authentication problem.
The radical sect of Islam would either see you dead or "reverted" to Islam.
As long as the IETF maintains a global perspective, it can not accept standards encumbered by IP more restrictive than the GPL. It seems obvious -- we've all benefited by open standards on the Internet. But who knows, stranger things have happened.
This could be a good test case. MS may continue to pursue its IP Holy Grail business model, but if the IETF can stand firm and refuse restrictive licensing, they will not be able to force it down the world's throat. On the other hand, if the IETF does accept these kinds of IP restrictions, MS may have a path forward in pursuing its new business model of patents and copyrights for obvious and trivial ideas.
So now nobody will implement this, and Microsoft, through patenting something obvious and trying to license it has scared everyone away from some pretty good ideas that would have been implemented otherwise, with or without Microsoft's help.
This is just the latest chapter in IP stupidity.
This stuff has been discussed for years, if this had been treated like most other W3C standards we'd be in the clear by now waiting for implementations, instead everyone's scared. Does anyone realistically think that there aren't patents that W3C standards already infringe? Finally we actually get rights to something and we're inspecting the teeth, simply because the subject has been raised.
The crazy part of this whole deal is that most software is riddled with potential patent violations, including Microsoft's and including projects like Mozilla, Gimp and Open Office. That's why MS are trying to retain *defensive* rights, because they know it would be dangerous to give this IP away, anyone could stand on their shoulders, and a widget and then sue them (and that has happened already) and Microsoft would have no way of countering. If they adopted a more GPL oriented license with the rights being rescinded in the event of any patent suit against M$ it would be golden. They could just do to the protagonists what IBM has just done to SCO, infact that wording is almost already in the GPL.
I think this situation can be salvaged with another revision of the license. We should not give up on this or go for the second best option on such an improtant proposal.
We're getting to witness what the beginning of the web would have been like had Tim Bernards Lee patented some of his ideas. It ain't going to be pretty.