Mike Markowitz wrote: > James Donald wrote: >> Date: Wed, 04 Jul 2007 09:19:41 +1000 >> From: "James A. Donald" <[EMAIL PROTECTED]> >> Subject: Re: [FDE] Introductions >> To: [email protected] >> Message-ID: <[EMAIL PROTECTED]> >> Content-Type: text/plain; charset=UTF-8; format=flowed >> >> I was under the impression that the use of ECC was obstructed by >> multiple extremely broad overlapping patents, each patenting much the >> same thing, > > James: > > This is another myth (possibly perpetuated by Certicom).
Thanks, this is the information I needed. ISC has had ECC (more > precisely, ECDH and ECDSA) in its commercial products for more than 10 years > now. (It's been around so long I can't remember when we first implemented it, > but it pre-dates by several years the announcement of the Certicom toolkit in > 1997. And, unlike the situation for RSA, we have not been sued for it yet!) > > We do NIST/NSA Suite B/ANSI/IEEE/IETF-compliant ECC key generation, key > wrapping, > key agreement, and signatures for CMS, TLS, in our CA, etc. -- freely in char > p, > but only with polynomial bases in char 2. > > Whether Certicom's point compression patent would stand up to a challenge is > questionable -- some feel there's plenty of implicit and explicit prior art -- > but it should be noted that the big boys (Microsoft, et al.) have chosen to > avoid > it... something that's easy to do and still comply with the standards. > > But that and the avoidance of (optimal) normal bases in char 2 are pretty much > the only things a *software* implementation needs to do to steer clear of the > patents. > > There are *no* patents I'm aware of that are required to fully comply with the > above mentioned standards and interoperate with the leading commercial > products! > [Usual caveat: I'm not a lawyer, so my response should be taken /cum granis > salis/.] > >> but that if the NSA licenses you, your ass is covered. > > *Our* NSA license defines a rather specific "field of use;" their license with > other companies may vary. (It was pretty clear when we negotiated our license > that > they did not have a fixed contracting procedure in place and we weren't the > first > company to go through the process.) For more info, you might start here: > http://www.nsa.gov/ia/industry/crypto_suite_b.cfm > > As one might expect, Certicom's spin on the deal is somewhat different (and > it's > not at all clear that NSA knows exactly what it paid $25M for): > http://www.certicom.com/download/aid-501/FAQ-The%20NSA%20ECC%20License%20Agreement.pdf > > > > Final determination might come down in the courts, but I'm not aware of > lawsuits > yet. > >> But for non government uses, NSA is presumably not going to license you, and >> cutting a deal with patent holders is going to be too arduous for any >> organization that does not have a menacing team of lawyers in house. > > NSA gives you a license to build, sell, and use products including the IP in > question -- 26 US and Canadian patents are cited -- in char p with p at least > 256 > bits in length to Federal, state, and local government agencies (and to > foreign > governments under certain circumstances) for the protection of classified data > or national security purposes. (Here I'm summarizing the clearer parts of an > MOU > some of which is vague and subject to various interpretations.) > > However, it's worth keeping in mind that you really only need such a license > for > MQV/ECMQV, and that's easily avoided by substituting (unencumbered) DH/ECDH. > (Note that all relevant "standards" -- including Suite B -- permit the use of > either > scheme for key agreement.) > >> What is the patent and licensing situation with that product? Did NSA >> license you to sell it to anyone? > > For the moment, we only offer our products with ECMQV to DoD agencies wishing > to > use them for "national security" purposes -- you might say we're playing it > safe. > (Not to mention the fact that, outside of DoD, no one has ever asked us for > MQV! You certainly don't need it to do TLS as you'd have no one to talk to if > you did. See link below.) > > Bob Jueneman responded: >> I�m not an attorney, but it may be worth observing that Microsoft includes >> ECC >> support within Vista, and Sun, Red Hat, and others have announced plans to >> support ECC as well. > > See http://dev.experimentalstuff.com:8082/ for the current situation regarding > ECC in TLS. (To the best of my knowledge, aside from Certicom, these companies > do not feel that any ECC license is required.) > >> SPYRUS products are unencumbered with license restrictions, and can be >> exported and used worldwide. > > Ditto for (most) commercial ISC products, subject only to BIS export > restrictions. > http://www.infoseccorp.com/products/contents.htm#Export > > -mjm > > > ------------------------------------------------------------------------ > > _______________________________________________ > FDE mailing list > [email protected] > http://www.xml-dev.com/mailman/listinfo/fde _______________________________________________ FDE mailing list [email protected] http://www.xml-dev.com/mailman/listinfo/fde
