Re: [Idr] IPR related to the flow-spec I-D

[Thomas Narten <narten at us.ibm.com>]

John Leslie <john at jlc.net> writes:

> > From: Ross Callon <rcallon at juniper.net>
> >> From: Thomas Narten <narten at us.ibm.com>
> >> 
> >> IIRC, the December disclosure only listed a single patent application.
> >> The revised one lists one patent (issued) plus 5 patent applications
> >> (no information available AFAICT). So the disclosure has been expanded to
> >> include significantly (?) more possible IPR compared with the original
> >> disclosure.

>    (It is a significant annoyance to those of us without "free" access to
> patent applications not published by the US patent office to be left
> guessing what IPR claims may be being disclosed.)

My understanding is that all patent applications that are public are
findable via the free http://patft.uspto.gov/ web site. However,
anyone applying for a patent may request that it remain private until
it is issued. My assumption is that that is the situation we are in
here, since I have not been able to access any of the applications
either.

> > The additional applications in the more recent Juniper disclosure are
> > patents where the inventors have not been involved in the IDR working
> > group, and are not authors of the flow spec document.

>    I am certainly happy to take Ross's word for this -- but additional
> possible (unknown) IPR claims still leave me wondering...

There is always the possibility (some would say likelyhood) of
additional, unknown patents covering any given IETF document. Not much
we can do about the unknown.

> > Therefore the connection between the applications and the IDR document
> > was not noticed until very recently (there is no requirement that all
> > IETF participants pay attention to every document in every IETF working
> > group just in case they happen to have IPR on it).

>    Indeed -- I hope nobody is asking to add such a requirement.

The IPR WG (when it discussed this in the past) understood this
reality very well. Hence, the disclosure  rules apply to
"contributers" to an effort. If you aren't participating in an effort,
you don't have to worry about making disclosures.

> > The text that you refer to talks about "contributors". It is my
> > impression however that if someone notices IPR that applies to an IETF
> > document then they are supposed to disclose it when the connection is
> > discovered, even if neither the inventors nor the person noticing the
> > connection are contributors to the document in question.

>    That is the way I read it, also.

>    As far as the additional patent applications being listed, I fear
> we have to accept this: the requirement of "contributors" to disclose
> IPR claims they become aware of is unambiguous, while their ability
> to disclose the particulars of those claims is often severely limited
> by company policy.

Right. Like it or not, this is reality. Many companies consider patent
applications "confidential". This is one reason why the IPR rules
don't *require* that patent applications be made available to WGs. 

>    In this particular case, we're faced with a late disclosure (the
> December one, not the revision), and serious reason to believe that
> a "contributor" was aware of IPR claims but trusted Juniper lawyers to
> handle the disclosure. Lawyers habitually put things off as long as
> possible; and that's the behavior we'd like to change.

Sorry, while it may be somewhat true, I don't agree with the
implication. It is the contributor's responsibility to make
disclosures. If they have to go through their own lawyers first (which
I would strongly advise!), they need to keep pushing on them to come
to closure.

>    We have two "sticks" we might apply:

> 1) asking for actual "royalty-free" terms, not Mutually-Assured-
>    Destruction terms, based upon either the late filing or the
>    vagueness of the claims disclosed;

> 2) refusing to proceed until the actual claims in all disclosed patent
>    applications are available to all WG participants.

>    Under the circumstances, I think either of these would be
> reasonable; though I could probably live with an assurance that the
> Juniper lawyers are willing to change their ways and that Juniper
> "contributors" will not in the future tolerate lenghty delays by
> Juniper lawyers.

Without taking a particular stance on how the WG should proceed, I do
think it is worth noting that the WG has a number of options on what
to do. Many may not be particularly attractive, or may not be
particularly realistic.

Some options:

1) do nothing. I.e., forward the document on Standards Track to the
   IESG anyway,

   Actually, this isn't "doing nothing", this needs to be "the WG has
   considered the IPR disclosures, but believes that publishing this
   document on standards track still is the best thing to do, all
   things considered". 

2) publish as info/experimental rather than on standards track. 

3) Table the document, and try to extract more attractive licensing
   terms from the IPR holder first. This approach could take a long
   time, and there are no assurances better terms could be
   obtained. Moreover, could the WG even agree on what would be
   acceptable terms? (The definition of what are "acceptable" terms
   tends to vary among different folk.)

4) Kill the document outright. I.e, to punish the IPR holders for
   failing to disclose. (And note, I'm talking specifically about the
   patent that should have been disclosed many years ago).

Or something else.

Overall, the WG needs to be realistic about what its options are and
then make a decision about how to proceed. The worst thing it could do
is simply not decide and leave the issue of what to do with the
document unresolved indefinitely...

Thomas