[cabfpub] Ballot process ordering

[Ryan Sleevi]

On Thu, Nov 3, 2016 at 9:14 AM, Gervase Markham <gerv at mozilla.org> wrote:

> Well, as someone who was part of the group drafting the original IPR
> policy, I can tell you that the intent was certainly that "Final
> Guideline" == "new document", and "Final Maintenance Guideline" ==
> "patch to existing document".
>

The problem is what you meant isn't what members signed ;) This is similar
to the appeal to authority - we have a contactual agreement in place on the
IPR policy, so it's a combination of matters - both intent and actuality -
that affect how we behave. This is similar to if you and I personally
entered in a contract where I agreed to pay you "one hundred dollars" a
month for 5 years, and then 2 years in, you tell me you meant to say "one
hundred *Canadian* dollars". You may have meant that, but the lack of
specificity - and precedent, are to some extent relevant.

That said, I readily admit we're far into the weeds here on the nuance, and
it's important to remember the context, which is what do we name things
along each step of the way, and is that supported. I'm entirely in favor of
clarity, and am trying to highlight the 'bugs' here that prevent that
clarity.

This is particularly significant if we allow multiple ballots to proceed
concurrently, because it's important to know what the IPR disclosure is
covering.

Using our old IPR policy as an example, simply because more people are
familiar with the issues, imagine we were discussing 3.2.2.4, and we had 3
methods
1) Encumbered
2) (believed to be unencumbered)
3) Any other method

Now imagine we have two Ballots - Ballot A and Ballot B.
A) Introduces Method 4, believed to be unencumbered
B) Removes Methods 2 & 3

Some of the problem we have with Position 1 is that, in such a scenario,
it's unclear what the Review Period encompasses, and how it works with
respect to Balloting.

Consider the following interpretations:
Ballot A's review period only considers the text "(1,2,3)+4". As such, if 4
is encumbered, it's not an Essential Claim (recall, we're using old IPR
policy for sake of discussion here, I'll explain why more below), so no
disclosure is required for 4.
Ballot B's review period only considers the text "(1,2,3)-(2,3)", and as
such, even if 4 is unencumbered, 1 is still an Essential Claim, so
disclosure is required.
Ballot A and Ballot B's review period considers the union of these ballots
"(1,2,3)+4-(2,3) - aka (1,4)" - and any disclosures must operate on that
text.

I used our old IPR policy simply because the 'ala carte' Essential Claim
issue was something most people understand by now. However, even under our
current IPR policy, this issue exists if we have multiple concurrent
modifications to the text, because the Essential Claimness of one section
can be affected by changes in another (e.g. imagine a change to Section X
saying "This is how you do Y", and a concurrent change to Section A saying
"You MUST set B in accordance with Section X")

Position 1 introduces more IPR ambiguity, because it leaves it unclear how
to handle this, for which the only logical conclusion is that there can
only be a single ballot at a time, and each ballot MUST take either 37 (30
+ 7 day voting) or 44 (30 + 7 day review + 7 day voting) before another can
be scheduled. And I would argue for the 44 on principle, because we
shouldn't be overlapping ballot review with IPR review, as we'd need to
restart IPR review any time a ballot modification was made during ballot
review.

I believe Position 2 avoids this, and matches our precedent, because both
Ballots can be voted on, and the Review Period can encompass the integrated
text of both ballots (if they succeed) or the individual text (if only one
succeeds). The activities of the Forum would only slow down if and when an
Exclusion Notice is put forward, because until the PAG is convened and
resolved, the guidelines (which can still be modified by ballot) are not
formally Approved.

This is in line with our past precedent, and given the significant risk and
ambiguity involved with the Review Period of Position 1, I'm surprised to
see people advocate that it brings clarity, especially given the context in
which this Forum has historically operated, and in which proponents of
Position 1 are presently trying to operate (by asking Patent Counsels to
review 3 significant changes simultaneously). There's nothing to further
prevent someone from putting forward 5, 10, or 50 of these - if they can
find co-endorsers - and completely execute a Denial of Service on companies
patent counsel, or force members with large IP holdings to incur
significant costs or risk, because Members cannot be certain that the
Ballot will fail. It's a system rife for abuse, and I believe the launching
of three concurrent review periods is, itself, an example of abuse, and
fundamentally why we disagree with Position 1.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cabforum.org/pipermail/public/attachments/20161103/be5f3244/attachment-0001.html>