[PP-DIALOG] MSBlock Comments -- First draft of updated IEEE-SA Patent Policy FAQs
Below are comments for the
IEEE Patent Policy FAQs. I will insert into the unwieldy table shortly.
The line headings refer to FAQ numbers. Red highlights direct to more key
items. These comments are intended to improve the FAQs and do not suggest
abandonment of my concerns with the Policy itself.
******
16/31. The EPC definition here could
be improved to be "what the rep believes to be essential or would
be essential if the spec is finally approved" -- as in other SDOs.
The "potential" ["may become"] and other uncertain
characterizations could lead to an obligation to disclose non-EPCs.
Some people are
asking how to reduce overdisclosure -- this definition promotes overdisclosure.
The concept of a "potential claim" is also potentially
confusing, especially in that patent application claims may be amended
to "become" essential.
The text should change 'potential claims"
and "may become" to "claims believed by the participant
to be essential if the specification was approved in its current
draft form". While legacy, the change is timely.
29/31. Once a party, including a third
party, submits an LOA for one EPC, IT IS COMMITTED TO FILING FURTHER LOAS
FOR OTHER EPCS. This follows the Policy and is a sleeper likely to
surprise third parties and members alike. Put differently, if a party does
not submit an LOA, it is fine. But
once it submits one LOA, it is obliged to file an LOA on any future potential
EPC or is in breach of the Policy. Interesting counterincentive.
While this is not a new FAQ, it is counterproductive. I appreciate the
desire to address a problem from a previous case, but this text isn't the
answer.
After "Additionally, a Submitted of a LOA..." change "is
required" to "is encouraged."
43. This FAQ is hard to follow with
its reference to two alternatives and "incremental value" which
has meaning to economists way beyond what seems to be suggested in the
FAQ. In any event, what does the non-selected alternative have to
do with the value and royalty of the selected alternative? The last
sentence of the FAQ is confusing that "Any incremental value imputed
to the selected option because of its inclusion in the standard is excluded."
I suggest deleting the last four sentences and insert what
I think was intended:
" A value benchmark
of a selected patented invention is the value it would have if there
were no standardization effects. That is, reasonable royalty does not include
value arising from the cost or inability of implementers, who have sizable
sunk in costs, to change away from the selected patented invention."
Otherwise, explain how the royalty is
affected by the second alternative.
44. "smallest saleable compliant
implementation" should be informed by recent cases, such as CSIRO
v Cisco and VirnetX v Cisco et al.
46. There seems to be some inconsistency
here. This factor asserts that EPC value should be measured in the
context of the EPC relative to the entire standard. That
is, "do not value the EPC in isolation of other EPCs in the standard."
But another Policy factor states that EPC value is confined
to the SSCI and should NOT look at the overall value of the product or
standard. So we look at the chip
that embodies the EPC to determine SSCI and Reasonable Rate, BUT then
confine the Reasonable Rate based on all the EPCs in the 802.11 standard--
which can include patents on the waterfront.
There is also no mention here of elasticity
of price, other than that royalties can be more than price could withstand.
There is also no discussion of what implementers are actually licensing/paying
and that royalties are not sought for most EPCs. In the example,
there may be just one or a few EPC holders seeking royalties. Under
the FAQ, the EPC holder may be entitled to 2% royalty for value of its
EPC, but is relegated to 1% of that because of 100 disclosed patents which
may or may not be essential. Court case law has refused to follow the notion
that all EPCs (e.g. the 100 disclosed) are valid and essential and
will evoke a royalty. The fraction approach also fails to recognize that
different patents may have different value.
47. The FAQ does not clarify the
uncertain "implicit threat of Protective Order" included
in the Policy. This condition could attach to and possibly
exclude many helpful licensing negotiations, to the detriment of EPC holders,
implementers, and courts. Consider
the recent CSIRO v Cisco decision in which the implementer
presented evidence of a license agreement offered "prior to the inclusion
of the 069 patent in any 802.11 standard." The alleged infringer [implementer]
argued that "the [agreement] is an actual, real-world license, negotiated
at arms-length, not clouded by litigation, and occurring prior to the hypothetical
negotiation...[T]hese factors isolate the value of the patented technology."
The limitations in the IEEE Policy "comparables" factor,
left vague and unlimited by the FAQ, would exclude this agreement
[in that there was certainly a threat of injunction in CSIRO] and
do effectively emasciate the most valuable information for determining
reasonableness. [It is
noted that the agreement was not relied on by the CSIRO court, however
for reasons unrelated to the timing or potential injunction aspect.] Expressing,
in the FAQ, one trivial "example" of "implicit threat"
is of no value. If the FAQs are ashamed of the facts, then perhaps the
Policy needs reconsideration. IEEE should consider the anticompetitive
effects of this factor, and the problems it raises with the negotiating
parties.
48. Contrary to prior explanations by
the Policy drafting team of the term "should" as permissive,
this FAQ makes the factors sound like "must" requirements. Moreover
the syntax raises questions about the permissive nature of the factors.
A sentence should be added that "The
Policy factors are permissive, not required."
56. This FAQ states that an EPC
holder can never seek Prohibitive Order. This is contrary to the Policy
itself that states when an Order can be sought.
57. Where does the Policy discuss "burdens
of proof" ? Why not state that IEEE Policy does not change any
laws of jurisdictions? Why leave rules unchanged and not include
laws? This is not a
FAQ but a Policy issue that needs vetting.
61. The
question is whether an implementer is immunized from injunction for any
conduct it engages in. The FAQ provides no useful information in answer
to the question.
83. As
for the statement "If any Accepted LOAs on a standard have selected
Reciprocal Licensing, then an Applicant can invoke the terms of Reciprocal
Licensing", that sentence is flawed and should be deleted. Frankly,
I did not interpret the Policy to
requirethe EPC licensor
to take a license back from the licensee under Reciprocal Licensing. As
I interpreted it, the implementer has the option of requesting a license
and being offered RAND terms. That is, to prevent the EPC licensor
itself from being blocked, the licensor can request
a license back under RAND terms. Is
it intended by the FAQ that, by the EPC holder checking the "reciprocity"
box, the licensee can invoke or force Reciprocal Licensing on the EPC
holder? Suppose
the EPC holder has already paid for a license under licensee's EPCs or
merely does not wish to take a license under the licensee's EPCs (if there
are any) -- after all the licensee has the right to refuse to request or
take a license from the EPC holder. Is the purpose of reciprocity
to place the EPC holder and implementer on equal footing, or is it
to turn the crank on the EPC holder? Is this provision suggesting that
an EPC holder (implementer) can force another to take a license under its
EPCs? This is,
of course, a wide departure from existing standards bodies, such as ETSI
in which "The above undertaking may be made subject to the condition
that those who seek licences agree to reciprocate." The
word "may" in ETSI text allows the EPC holder to seek reciprocity
but does not force him/her to accept it or have it imposed by the licensee.
Please correct me if my interpretation is wrong, especially if my proposed
change reflects the FAQ meaning.
Replace the sentence:
"If an EPC holder has checked the reciprocity box, such licensor
may request a license back from the Applicant in accordance with the definition
of Reciprocal Licensing."
To the extent that the Policy
may be interpreted as stated in the unrevised FAQ text, the Policy
should be revised.
84. This
FAQ cannot be deciphered without the DaVinci Code. It is appreciated that
blanket licensing must conform to the patent transfer provision, but the
reciprocity, RAND v RF, BUYER v SELLER LOA, etc deserve more open discussion
than this FAQ provides. Any
such provision should be considered for the Policy, with a FAQ that explains
it. This is not some
informal guidance but a substantive issue that may impact standards participants.
It is appreciated that the FAQ is largely legacy, but it is still
hard to follow.
Marc Sandy Block,
Counsel, Intellectual Property Law
1B117 / North Castle Drive / Armonk, NY 10504
msb@xxxxxxxxxx
TL 251-4295 (outside 914-765-4295); fax 251-4290