Microsoft Patent Too Close to IPv6, Lawyers Say (eWeek) [LWN.net]

eWeek looks at a dispute over a Microsoft-held patent. ""We are aware that the patent should not have issued in view of the prior art available to the patent office but not cited by Microsoft in its application," Moglen said. The patent in question, USP 6101499, filed in 1998 and issued in 2000, concerns automatic generation of IP addresses to facilitate simple network connections. The technology described therein bears "more than a passing similarity" to IPv6, one of the backbones of the Internet, according to Frank Bernstein, a lawyer with Kenyon & Kenyon, a San Jose, Calif., firm."

(Log in to post comments)

Microsoft Patent Too Close to IPv6, Lawyers Say (eWeek)

Posted Mar 31, 2005 9:39 UTC (Thu) by tialaramex (subscriber, #21167) [Link]

This is a patent for IPv4 (yes version 4 not 6) link local auto-addressing, a fact that was brought to the attention of the IETF back in 2000/2001 when the ZeroConf RFC for v4 LLA was put on standards track. Microsoft's official declaration states that they are prepared to license this patent to any implementor on royalty free, RAND terms in exchange for like terms on relevant patents from that implementor. If you believe in defensive patents (which I'm not sure I do), this is one of those.

IPv4 LLA is already found in your Windows PC, your OS X machine and in certain circumstances in your Linux desktop machine.

This patent doesn't even begin to resemble IPv6 auto-addressing, which in any case was invented long before this patent was written, let alone submitted. In fact all the example addresses are even IPv4 dotted-quad addresses! Most of the techniques described are needed only because of the restriction of v4 addressing at the time, such as the lack of a link-local scope marker, and the shortage of address bits.

Unfortunately instead of asking anyone level-headed and technically minded, the journalists behind this article simply quoted whatever they were told by lawyers (if in fact they were told anything at all, there's no press release about these wild accusations on PUBPAT's web site). Eben Moglen is probably a great lawyer but he wouldn't know a "scope bit" from a "dotted quad" so far as I can tell. I've asked the editors to either retract the article or at least get a fact-checker to call the relevant people (shouldn't a balanced article already have a quote, or "no comment" from the IETF and Microsoft?). They've ignored me, which I guess is a reminder that my money spent on LWN was a good investment, at least editors here know their jobs.

Microsoft Patent Too Close to IPv6, Lawyers Say (eWeek)

Posted Apr 1, 2005 14:34 UTC (Fri) by farnz (guest, #17727) [Link]

I've just been and read RFC1971 and draft-ietf-zeroconf-ipv4-linklocal-17, and the only significant difference between the two is that RFC1971 takes advantage of the extra address space, while IPv4 link-local generates random addresses. As a result, RFC1971 IPv6 addresses only work if there are no EUI-64 conflicts, while IPv6 link local addresses can find a new address upon a collision.

Both use conflict detection via the standard mechanisms for determining if an address is in use, so that's not even a difference. In short, about the only innovative step I can see between the two mechanisms is generating addresses using a PRNG seeded from a local identifier, as against using the PRNG directly; if this is innovation, we're all doomed.