Govt. Data and Patents
Jennifer Bell
A question: One potential argument governments may use against open
data such as promoted here is that private companies would then be
able to patent ways of presenting and using the data. That type of
patent would then prevent the government from providing the service
itself as a public good.
Is this a valid concern? If so, is there a stock refutation for it?
Jennifer Bell
http://visiblegovernment.ca
Aaron Swartz
> data such as promoted here is that private companies would then be
> able to patent ways of presenting and using the data. That type of
> patent would then prevent the government from providing the service
> itself as a public good.
>
> Is this a valid concern? If so, is there a stock refutation for it?
That seems like a weird concern. First, the patent could be filed even
if the data wasn't open. Second, I have trouble imagining anything
that specific that anyone would patent. Third, why would you not open
things up at all because if you did someone my prevent all of it from
being open? It's like refusing to serve your guests dinner because
you're not sure if you have wine.
Jennifer Bell
viable market for for-profit opportunities.
This point specifically came up as a comment from an MP candidate in
our 'I Believe in Open' campaign (http://ibelieveinopen.ca). It's the
sort of boogey-man argument that could do wonders as a scare tactic.
I'd rather be able to defuse it than ignore it.
Jennifer
On Dec 8, 11:48 am, "Aaron Swartz" <m...@aaronsw.com> wrote:
> > A question: One potential argument governments may use against open
> > data such as promoted here is that private companies would then be
> > able to patent ways of presenting and using the data. That type of
> > patent would then prevent the government from providing the service
> > itself as a public good.
>
> > Is this a valid concern? If so, is there a stock refutation for it?
>
> That seems like a weird concern. First, the
David Orban
>Yes, but open govt. data lowers the barriers to entry, creating more
>viable market for for-profit opportunities.
It appears that in a capitalist society that would be seen as exactly part of the objectives of government: enabling unfettered economic enterprise.
Once taxpayers covered the cost of generating the data, it is under these assumptions not up to the government to decide what are good uses of that data, and what are bad uses. As long as private enterprise works within the boundaries of the then existing laws, the marketplace decides what uses should be viable, because profitable.
Of course, the caveat is in the "then existing laws", as these might be insufficient, or incorrectly framed, or inadequately enforced or supervised. We have seen plenty of all possible combinations of these.
D
David Orban
skype, twitter, linkedin, sl, etc: davidorban
Brian Behlendorf
A patent on "ways of presenting and using the data" wouldn't read upon a
system that simply provided the data to end-users to do with as they
liked. It might limit what those end-users can do with it, but not
eliminate all uses. Can the MP (or anyone) name a particular patent that
illustrates this example? Or is that just a justification for inaction?
Brian
Josh Tauberer
files and XSLT for the display of roll call vote records of the United
States House of Representatives." This would cover what the House does
now, as well as what I do on GovTrack, and it's possible (but I'm not
sure) that I even did it before the House.
Is this patentable? Should I try? Can patents be enforced against the
government?
I think maybe we'd need some IP lawyers to weigh in.
--
- Josh Tauberer
- GovTrack.us
"Yields falsehood when preceded by its quotation! Yields
falsehood when preceded by its quotation!" Achilles to
Tortoise (in "Godel, Escher, Bach" by Douglas Hofstadter)
Greg Elin
Jennifer, Let's step back a second and imagine that maybe this MP was using the term "Patent" as a catch all phrase for Intellectual Property. The concern is a commercial company could devise a transformation on the data, or copyright a compilation that performs an obvious improvement but may not rationally justify protection and thereby end up with a monopoly (such as WestLaw did with page numbers on court decisions.)
I might suggest the possible replies are as follows:
1. No reply. Better to spend your spirit looking for champions than arguing with fools.
2. No reply until a couple examples offered, "Until the MP could offer a couple of historic examples she wishes to avert, hiding our assets in closets lest phantoms run off with a few seems inefficient."
(I think Aaron and Brian are suggesting these.)
3. Remove the problem by embracing it as David suggested. We welcome the market's creativity in making government information more useful; if someone can create a presentation so non-obvious and beneficial such it deserves a patent, then indeed it deserves a patent.
My experience with is that openness is so efficient and superior to proprietary solutions, that is not worth trying to convince skeptics. Focus on innovators and early adopters.
Greg
Josh, it might fun to file a simple a (defensive) patent and put it through the peer-to-patent process and see what happens.
Jennifer, Let's step back a second and imagine that maybe this MP was using the term "Patent" as a catch all phrase for Intellectual Property. The concern is a commercial company could devise a transformation on the data, or copyright a compilation that performs an obvious improvement but may not rationally justify protection and thereby end up with a monopoly (such as WestLaw did with page numbers on court decisions.)
I might suggest the possible replies are as follows:
1. No reply. Better to spend your spirit looking for champions than arguing with fools.
2. No reply until a couple examples offered, "Until the MP could offer a couple of historic examples she wishes to avert, hiding our assets in closets lest phantoms run off with a few seems inefficient."
(I think Aaron and Brian are suggesting these.)
3. Remove the problem by embracing it as David suggested. We welcome the market's creativity in making government information more useful; if someone can create a presentation so non-obvious and beneficial such it deserves a patent, then indeed it deserves a patent.
My experience with is that openness is so efficient and superior to proprietary solutions, that is not worth trying to convince skeptics. Focus on innovators and early adopters.
Greg
--
Greg Elin
Sunlight Foundation (http://sunlightfoundation.com)
Sunlight Labs (http://sunlightlabs.com)
ge...@sunlightfoundation.com
gr...@fotonotes.net
http://twitter.com/gregelin
skype: fotonotes
aim: wiredbike
cell: 917-304-3488
--
Greg Elin
Sunlight Foundation (http://sunlightfoundation.com)
Sunlight Labs (http://sunlightlabs.com)
ge...@sunlightfoundation.com
gr...@fotonotes.net
http://twitter.com/gregelin
skype: fotonotes
aim: wiredbike
cell: 917-304-3488
Dan Knauss
A private entity must be able to protect their presentation/
repackaging of public data, but this would not extend to the data
itself, in every possible presentation of it.
The likely danger is deals between government and private companies
where the gov. deliberately makes the data hard for the public to
access to increase its resale value to private entities. In turn,.
think of private companies that monitor traffic and route people
around congestion by tracking cell phone GPS. Is that "public data?"
It has obvious value to city planners and public works entities. The
landscape is changing and it will take a while for conventions and law
to declare some ground rules.
Some concerns similar to yours are probably valid, but not as reasons
for making data very hard to access. Government deliberately making
public data hard to access is problematic in itself, ethically and
perhaps legally. Eventually not providing data in certain formats will
be seen as akin to refusing to provide it on paper, and providing it
only on paper will be seen as akin to offering it only on clay
tablets.
Maybe the key idea here is that control over accessibility (data
forms, media types) equals proprietary control if duplication and
manipulation is restricted beyond some threshold level.
I have heard of some municipal bodies giving access to commercial data
providers and then treating this as exclusive, maybe even with
contracts and money changing hands.
DK
Thomas Lord
attorney but I have studied patent law as
pertains to software more than most people
so I have a somewhat informed opinion:
The example was given (by Josh):
[a] use of XML files and XSLT for
the display of roll call vote records
of the USHR
could such a use, if sufficiently novel,
be subject to a patent? No, mainly, but
also half-yes:
The public record of roll call votes is
an intangible. There is no transformation
of tangible inputs to outputs here. The
public record is, in essence, a mathematical
construct and the display program is simply
an algorithm. Such a program wold fail the
"State Street" test.
State Street's software / business method patent
held up because it read on a process for
transforming brokerage orders. A brokerage order
is, in essence, tangible. It can be reduced to
a definitive sheet of paper, if need be, and with
two such pieces of paper before a judge a judge
can make a pronouncement that one of those pieces
of paper <i>is</i> the order, the other is merely
a record of the order. Trading offices physically
transform orders to other tangible instruments.
They found a novel transformation that happened to,
to be practical, require an automated control system.
The description of that control system reads on
software algorithms but it is not the algorithms
themselves which are patented but rather a particular
way of running a trading office.
There is no analogy to public records -- there's
no one slip of paper, so to speak, that the display
algorithm is transforming. The display program
takes records in (intangibles) and gives output out
(intangibles). The test of "GOTTSCHALK v. BENSON"
applies instead (the famous BCD conversion case).
Two half-yes ways in which patents could come up:
1) Not any algorithm but the graphic design of the
output could conceivably be subject to a design
patent (just as could the floral pattern on your
wallpaper).
2) If the transformation of the public record were
part of a control element for a larger transforming
process then the State Street test would apply to that
larger process. For example, if State Street decided
to buy and sell stocks based the outcomes of various votes
then the machine that automated the ordering is the
sort of machine that could be patented (if other criteria
are met such as novelty and non-obviousness). Of course,
State Street would not be prevented from getting the same
patent even if they had to pay for the public record
data or piece it together themselves by diligently watching
CSPAN and taking notes :-)
-------------------------------------------------------------
There is a <i>real</i> concern, though, and perhaps
this is where the original question really comes from:
The public record you and I get is an intangible, as I
said. However, there <i>is</i> (in principle) an
official copy of this record -- a "definitive" copy.
In a rare case, a judge could be confronted with two
different pieces of paper, each recording the outcome
of the same vote but differing in their account, and
the judge would make separate decisions as to (a) which
was the official record, (b) whether or not that record
was accurate. The answer to (a) helps decide if
duties that relate to that official record were performed
properly; (b) helps to decide if the official record needs
to be corrected.
Congressional record keepers implement processes that
transform those tangible, official records.
Now, suppose a third party studies the operations of
congressional record keepers and creates a brilliant
innovation: a new process for managing and transforming
those tangible, official records. As in state street,
an important control element is software. This third
party patents the new invention.
Now the algorithm in that software may not itself be
subject to patent protection but the <i>process</i>
may be and therefore the question arises: can this third
party <i>exclude congressional record keepers from
this new process or demand exhorbitent rent from them?</i>
The answer would appear to be a resounding yes however
this is nothing new: Hollerith already played that game
and look where that got us...
-t
--------
Fun Fake Facts: In 1823 the US Mint redesigned the
the dollar bill, shrinking it's size from 6"x6" to
the size of a Hollerith punch card so that the new bills
would conveniently fit in people's bill-folds along-side
the cards they already carried. While a substantially
large minority of Americans opposed the change the majority
was able to unify by rallying around a new group identity
as "card carrying patriots."
Joseph Lorenzo Hall
something that's been public for more than a year... so I think that
would moot Josh's specific example (the case we teach in cyberlaw in
the U.S. on this is Netscape v. Konrad, 295 F.3d 1315:
http://bulk.resource.org/courts.gov/c/F3/295/295.F3d.1315.01-1455.html
). best, Joe
Joseph Lorenzo Hall
ACCURATE Postdoctoral Research Associate
UC Berkeley School of Information
Princeton Center for Information Technology Policy
http://josephhall.org/
Jennifer Bell
markets, I think it's important to remember that markets are capable
of finding one of many possible steady-state solutions to a problem --
usually the one that benefits the group of people who are able to pay
the most, not the one that benefits the most people. Really, I
think Thomas Lord best hit the essence of the concern about patents on
the head when he said:
"can this third party exclude congressional record keepers from this
restrictive patents in the government data space, I can think of a few
horrible process patents in the commecial sphere -- Amazon's one-click
purchase patent being the most astounding. Display patents, while
they may seem trivial, are also spooky. By coincidence, a friend was
complaining to me recently in a crowded bar about how Nortel has a
patent on what is really the best possible way to display a graph data
structure -- because of this patent, for the next 20 years no one else
can draw a graph in what my friend considers the most efficient way.
I can easily imagine patents on:
- amazing ways to display data
- amazing ways of combining data
- amazing ways of processing the data
Which is well and good, because without a monetary incentive, these
amazing ways will likely never come into existence.... However, the
original concern was (I think) that: if the amazing thing that's been
produced is something that would benefit everyone, but is locked down
by patents and only offered to the richest -- what then? Wouldn't it
have been better for us, as an institution in the public good, to have
developed it ourselves so that we could offer it to everybody?
After reading your posts, I think that the best way to diffuse this
argument (*if* ever comes up again), is to take Greg's 3rd point and
embrace the capitalist view whole-sale -- propose that the government
should be like Cisco and out-source it's R and D costs to the
entrepreneurial / VC community, which is better able to spot
opportunity and manage innovation anyway. In fact, if the government
were to get in the Cisco-style habit of buying out the cream of the
public-data innovations, it would create a thriving market for more
of the same.
Jennifer