Re: [tcpm] Juniper's IPR on draft-ietf-tcpm-tcp-auth-opt-03

[Eric Rescorla <[e-mail]>]

Merging two messages:

[# 1]
> Is there anything in the disclosure language that suggests that the
> license is revocable or otherwise not perpetual, so long as the one
> licensing condition is met? If so, the licensing language says something
> that we didn't mean to say. We can fix that.

Well, that's a pretty big condition, IMO.


> The one condition of the license (i.e., we won't sue you if you don't
> sue us) is copied verbatim from many other IETF IPR disclosures. We ask
> only for the same protection that is granted routinely to other vendors.
> 
> Is there some reason why this protection should not be granted? If so,
> there are lots of problematic IPR disclosures on the IETF web page.


[# 2]
> This kind of equilibrium is exactly what Juniper is hoping to achieve.
> Juniper' like the vast majority of its competitors, makes money from
> sales, not litigation. We want to spend our dollars on engineers, not
> lawyers.
> 
> To some degree, I believe that that equilibrium has already been
> reached. I hold this belief because this licensing condition is so common.
> 
> Searching the IPR disclosure web page, I see one company (aside from
> Cisco and Juniper) using exactly the same language. One other company
> uses nearly identical language to say the same thing. Most other
> disclosures have some condition of "reciprocity".

I'm not sure what to say to this other than "huh?"

IETF isn't granting any protection to anyone. Rather, IETF determines
whether to incorporate specific technologies into their standards.
One consideration when making those decisions is whether those
technologies are encumbered and if so, the conditions under which the
license is granted. The license Juniper proposes is substantially less
free than it could be, and I, for one, would prefer that if the TCPM
WG has the option it select a technology which isn't subject to
licensing terms of this type. 

Again, I can certainly understand why Juniper would like to have
a mutually assured destruction patent situation with other companies,
but why exactly is it in the IETF's interest to assist them in 
that endeavor?

-Ekr