Re: [Cfrg] Comb algorithm IPR status

You repeatedly insisted (based on advice given to you?) I "prove" there
were other techniques by which the sort of curves being discussed here. I
explained repeatedly that all these alternative implementations of similar
curves meant that either all of them were infringing, in which case there
couldn't be some nefarious plot to patent troll as everyone would already
be infringing, or they were existence proofs of non-infringing techniques.
Neither of those leaves any room for Microsoft patent trolling, which is
what you insinuated (based on advice given to you?) we were doing and
Robert stated outright.

What is clear is there was never an IP problem and the facts (based on
advice given to you?) support exactly what I said from the start. You
helped manufacture (based on advice given to you?) this issue and
participated in baseless, inappropriate and personal maligning of me and
others from Microsoft. CFRG, IETF and the people who rely on the output of
these processes deserve better. Good day.

On Thu, Mar 5, 2015 at 11:58 PM, Alyssa Rowan <akr@akr.io> wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA512
>
> On 06/03/2015 02:54, Benjamin Black wrote:
> > There are only two possibilities: […] Which is it?
>
> Ah, you misunderstood. Let me clarify my thought process for you:
>
> • mLSB and SABS combs are different comb techniques.
>   - mLSB is (I believe, based on advice given to me) patented.
>   - SABS is (I believe, based on advice given to me) clear.
>
> • The NUMS proposal used/referenced mLSB in ECCLib and the paper.
>   - It didn't mention the mLSB patent, so we didn't know.
>   - You claimed you didn't know either.
>
> • You said ECClib passed MS's legal review for Apache2 licence release.
>   - Apache2 [paraph.] (conditionally) grants for that Work/Contribution.
>     · Anyone but MS releasing ECClib → trouble?
>   - Apache2 is incompatible with, for example, GPLv2 (e.g. Linux!)
>     · Anyone reusing ECClib's mLSB design, but not ECClib? → trouble?
>   - We didn't know about the patent, so that could have been disastrous.
>     · I don't want any trouble. [insert Jackie Chan meme here]
>
> • However, thanks to Mike (& Robert etc), we know about it now.
>   - Curve25519 reference implementation & draft doesn't use combs.
>   - And Ed448-Goldilocks reference implementation uses SABS combs.
>   - So, we no longer have a problem. I consider it a closed issue.
>
> I hope that clears things up and we can move on.
>
> - --
> /akr
> -----BEGIN PGP SIGNATURE-----
>
> iQIcBAEBCgAGBQJU+V4sAAoJEOyEjtkWi2t6f0IP/iu5x3DkeghGqE8GJ7Y9OJIq
> +z/KfP5g80JnShEFfsvfoX9MoNaJA5LlWxsR3NFXxJHcI1qwvsXIUvHuL3GBsWgT
> jvRgJwZSkUpdD1Y50IPeo1ZAliLJ6SU8wlrEvgGEt3ZsYNM7L/SaBLpTB8X+oNl9
> CNp9NyWO2PSTbHrseQl1JWpZj4JbonevVDK2amheq5atzkQnkOd3bcvOA3FQp9zS
> 7jwuTZyFWzyCgI3JNBF2IAEzV9l8wl1O/MbvQlrjcPkGWIQCHVM45ZTUHBN+T+PL
> SBo/MX85LD+eckOV9L+hS8+uF9q6v9M4DTyiuumB2A+wX+DF98/6tz1kPEDUN0Vf
> R5PMIbb05Q4LpwLVo4WAX9iyALuLah4cmaX15MWvQmZF4YqcCaBqzWgIvQyBhfkl
> 2IQQqfsQp/OYC0XyDoWNJjs/sD0JCxccSIkH0hRjPdOg2K6htZVq3cPfoMSh37C/
> SMP6x9pWmkVz8cZKZd4CarfwOws463mID9qG5pVIjioWCCadUI71wOasHN8eg6dp
> X4utBrKuZLpwAfa4u8CHZf3fGUfkY3wvErBtBjkNCCa1cCq5EMqkyaBNwja6iWxp
> aCvudkPpifrjPcJwIDdfAuWsUd8PDzprg41hCK+h+jwdW5QjQDK+GgK15+Zjf/8I
> f9T8DWnJ6A+M/5In2oLq
> =4KTO
> -----END PGP SIGNATURE-----
>
> _______________________________________________
> Cfrg mailing list
> Cfrg@irtf.org
> http://www.irtf.org/mailman/listinfo/cfrg
>