[License-discuss] [Non-DoD Source] Re: patent rights and the OSD

Karan, Cem F CIV USARMY RDECOM ARL (US) cem.f.karan.civ at mail.mil
Tue Mar 28 14:24:18 UTC 2017


Hi Nigel, this is my reply to your earlier email.  I see the point you're making much more clearly now.  I didn't have a good answer for it, so I had to talk with ARL's Chief Legal Counsel's office to put together an answer; I hope that this covers it.

First, let’s consider a case where ARL has created software that it, and only it, has control over. This includes possession of the code and any IP rights that may be, or have been, established and held by the government.  We are assuming for the moment that the software and any IP covering the software does not exist elsewhere. In this case, ARL’s internal processes cover what is released and how. If any patents exist that cover the software a decision can be made by ARL that it is in the best interest of the government to open source release the software and waive any existing patent rights. 

So now let's get to the more interesting cases that you brought up.

Imagine that an ARL employee gets code from NASA. Whatever IP rights the government may be able to assert in the code are under the custody and administration of NASA. ARL, or more specifically ARL’s Laboratory Director, is not in a position to waive the Government’s IP rights in this instance. ARL should only use the code from NASA for the purposes identified by NASA. In practice, any agency supplying code to another agency will probably have instructions and restrictive legends as to what can and can’t be done with the software.  It certainly would be bad form if ARL violated those instructions and restrictive legends by inappropriately releasing the software in an unapproved way. However, NASA would not be able to sue ARL. Both organizations are part of the same Federal government and would just have to agree on how to best handle the situation. 

Now let’s go one step further. ARL employees write some code which may be patentable. They go through ARL’s internal process to release the code, with the decision that it is better to release the code as Open Source rather than pursue patenting. The ARL software is made available for download. Sometime later, NASA discovers that ARL’s software is covered by a patent obtained by NASA. Although NASA might be annoyed with ARL that they are releasing their patented invention free of charge, they have no recourse against ARL.  What’s more, releasing the code as Open Source would arguably exhaust any claims the Government might have of patent infringement. Recognize that the Department of Justice is the Federal government’s enforcement agency. Pursuing a case would not be under the purview of NASA. 

So let's go one step even further.  Assume the same initial set of facts above, except that NASA has entered into an exclusive license agreement. NASA’s licensee finds out that ARL has been distributing code covered by the NASA patent that they have exclusive rights to. Does the licensee have a “breach of contract” action it can bring against the government? Does the licensee, typically having the right of enforcement, have a cause of action against those parties that have downloaded the software from ARL?  If so, do the downloading parties then have any recourse against the Government?  As far as I know, none of this has been litigated in court, so I don't have any answers for you.

That said, this is a fairly unlikely case.  Here is how it breaks down.

A = Probability that some USG agency obtains a patent on a given invention.
B = Probability that said patent is exclusively licensed to some external entity.
C = Probability that another USG agency independently invents the same thing.
D = Probability that second agency decides against patenting (which means that the first agency's patent is not discovered).
E = Probability that second agency creates code that is covered by the first agency’s patented invention.
F = Probability that second agency Open Sources the code.
G = Probability that the licensee of the first agency finds the second agency's code.
H = Probability that the licensee decides to sue someone (factor of how much money the patent is worth vs. how much money they'll get.

Multiply all those probabilities together, and that is the probability that someone gets sued.  Is it worth considering?  Maybe, but then the only reasonable approach for ARL is to refuse to Open Source at all.  So the question becomes, is this risk acceptable?  I don't have any answers regarding that; you'd have to decide it yourself.

Thanks,
Cem Karan

PS, none of this is legal advice, I'm not a lawyer, etc., etc., etc.

> -----Original Message-----
> From: License-discuss [mailto:license-discuss-bounces at opensource.org] On Behalf Of Tzeng, Nigel H.
> Sent: Sunday, March 19, 2017 7:43 AM
> To: license-discuss at opensource.org; lrosen at rosenlaw.com
> Subject: Re: [License-discuss] [Non-DoD Source] Re: patent rights and the OSD
> 
> All active links contained in this email were disabled. Please verify the identity of the sender, and confirm the authenticity of all links
> contained within the message prior to copying and pasting the address to a Web browser.
> 
> 
> ________________________________
> 
> 
> 
> Cem,
> 
> Larry's advice (tongue in cheek or not) isn't the issue for FedGov or ARL policy. That I randomly picked an ARL patent that happens to
> illustrate the potential issue is simply lucky.
> 
> The issue is that someone in some other part of the government could have released an open source implementation with a patent grant
> of Dr Chen's patent and deprived him (and ARL) of financial reward for his work.
> 
> I'm sure there are some here who would simply shrug and say who cares...rent seeking is evil anyway...but the FedGov policies regarding
> rewarding innovation is one aspect of attracting and keeping inventors in the government as opposed to going to private industry.
> 
> ARL and the DoD has an even harder time than research universities when it comes to the left hand not knowing what the right hand is
> doing. IMHO explicit patent grants provided by ARL in open source releases should be limited to those generated by ARL itself unless ARL
> intends to insure it isn't giving away a patent created by an inventor in another branch of the federal government.
> 
> I say this over and over but while it is laudable for you to give away something you create to the world it is unethical for you to give away
> something someone else created to the world.
> 
> Broad patent grant language in FOSS licenses/policies risks exactly that for very large organizations and the cost to prevent that means it
> is far easier (and cheaper) to simply not open source.
> 
> I really implore you to revisit the ECL V2 patent language and the reasons they didn't feel that vanilla Apache 2.0 was suitable for them
> when you (and the ARL lawyers and your OTT) are helping evolve the ARL open source policy.
> 
> Nigel
> 
> Obligatory Disclaimer: speaking as an individual, the opinions expressed here are my own and no one else's, etc.
> 
> From: Karan, Cem F CIV USARMY RDECOM ARL (US) <cem.f.karan.civ at mail.mil < Caution-mailto:cem.f.karan.civ at mail.mil > >
> Date: Sunday, Mar 19, 2017, 6:42 AM
> To: lrosen at rosenlaw.com <lrosen at rosenlaw.com < Caution-mailto:lrosen at rosenlaw.com > >, license-discuss at opensource.org <license-
> discuss at opensource.org < Caution-mailto:license-discuss at opensource.org > >
> Subject: Re: [License-discuss] [Non-DoD Source] Re: patent rights and the OSD
> 
> I was out Thursday and Friday, but... see attached Federal Register notice regarding this particular patent.  I know that Larry was (mostly)
> pulling everyone's leg, but before anyone gets themselves into trouble, I figured I should put the warning out there.
> 
> I'm not a lawyer, as far as I know this doesn't count as an official warning, I'm not representing the US Government (or any other
> government) in this matter, etc., etc., etc.  I just personally don't want to see anyone spend time on something and then get burned in the
> end.  Please don't shoot (or flame) the messenger.
> 
> Thanks,
> Cem Karan
> 
> > -----Original Message-----
> > From: License-discuss
> > [Caution-mailto:license-discuss-bounces at opensource.org <
> > Caution-mailto:license-discuss-bounces at opensource.org > ] On Behalf Of
> > Lawrence Rosen
> > Sent: Wednesday, March 08, 2017 3:07 PM
> > To: license-discuss at opensource.org
> > Cc: Lawrence Rosen <lrosen at rosenlaw.com>
> > Subject: Re: [License-discuss] [Non-DoD Source] Re: patent rights and
> > the OSD
> >
> > All active links contained in this email were disabled. Please verify
> > the identity of the sender, and confirm the authenticity of all links contained within the message prior to copying and pasting the
> address to a Web browser.
> >
> >
> > ________________________________
> >
> >
> >
> >
> > Nigel Tzeng wrote:
> >
> > > Using US7460689B1 System and method of detecting, recognizing, and
> > > tracking moving targets as an example it could be useful to have
> > an open source copyright license to any USG developed MTI
> > implementation of US7460689B1 because the libraries and functions used to implement the patent could be reusable.
> >
> >
> >
> > Nigel, I am avoiding reading that patent because I have no time, but since you did read it I can ask you some questions:
> >
> >
> >
> > Is the patent already "reusable" for things other than "detecting,
> > recognizing, and tracking moving targets"? What is the relationship
> > between "libraries and functions used to implement the patent" and all those different uses of the patent claims than are disclosed in
> US7460689B1? Are those copyrighted "libraries and functions" themselves patented claims in US7460689B1?
> >
> >
> >
> > Without such a patent analysis – that no open source engineer is
> > required to undertake prior to an infringement notice from the patent
> > owner – I'm comfortable recommending that we implement such open
> > source software. The hell with random patents tossed out by Nigel
> > Tzeng to scare me. :-)
> >
> >
> >
> > /Larry
> >
> >
> >
> > "If this had been legal advice it would have been accompanied by a bill."
> >
> >
> >
> >
> >
> > -----Original Message-----
> > From: License-discuss [Caution-Caution-mailto:license-discuss-bounces at opensource.org] On Behalf Of Tzeng, Nigel H.
> > Sent: Wednesday, March 8, 2017 7:18 AM
> > To: license-discuss at opensource.org
> > Subject: Re: [License-discuss] [Non-DoD Source] Re: patent rights and
> > the OSD
> >
> >
> >
> > Cem,
> >
> >
> >
> > To give them a more concrete example (hopefully not a flawed one ☺) I skimmed ARLs patent portfolio.
> >
> >
> >
> > Using US7460689B1 System and method of detecting, recognizing, and
> > tracking moving targets as an example it could be useful to have an
> > open source copyright license to any USG developed MTI implementation
> > of US7460689B1 because the libraries and functions used to implement
> > the patent could be reusable.  While a lot of the basic functions may exist today in GDL or SciPy the underlying image processing, change
> map creation, math, etc may have been reusable outside of the context of US7460689B1 and of greater interest to the software
> community when most of those capabilities were only widely available in systems like Matlab or IDL.
> >
> >
> >
> > Even more so the software components required to interface with Army
> > systems that would have to be built around any operational MTI system.  Getting access to source code from another government
> agency/DoD program isn’t always as painless as one might expect.
> >
> >
> >
> > While these could be broken out and individually open sourced it is an
> > additional burden on the Army/DoD Program and it simply may not get
> > done.  It would be far easier for the program to open source the entire system with the note that US7460689B1 and others apply to
> these 4 subsystems and care must be applied to any reuse of those components.
> >
> >
> >
> > I’m also curious of their opinion on the impact of the policy to give
> > away all software patents to Executive Order 10096 and the intent that
> > individual government researchers have the opportunity to benefit from their inventions.  Is there reason that software inventors should
> have less financial opportunity than hardware inventors?
> >
> >
> >
> > What safeguards exist or will exist for the USG inventor that an
> > unrelated ARL or DoD project will not open source code that
> > unknowingly implements their patent and provide a free patent grant?
> > What happens to the entity that may have contracted for an exclusive patent license agreement for that patent through their
> technology transfer office? Will the office responsible for approving ARL open source releases with patent grants be able to review source
> code applicability to not just ARL patents but those from the NRL or DOE?
> >
> >
> >
> > There is a reason I favor the patent grant language in the Educational Community License v2.0 for GOSS and large research
> organizations.
> >
> >
> >
> > Regards,
> >
> >
> >
> > Nigel
> >
> >
> >
> > ObDis:  Not speaking for the lab.
> >
> >
> >
> > On 3/8/17, 9:36 AM, "License-discuss on behalf of Karan, Cem F CIV
> > USARMY RDECOM ARL (US)" <license-discuss- bounces at opensource.org on
> > behalf of cem.f.karan.civ at mail.mil <
> > Caution-Caution-mailto:license-discuss-
> > bounces at opensource.org%20on%20behalf%20of%20cem.f.karan.civ at mail.mil > > wrote:
> >
> >
> >
> >     I can pass it through ARL's lawyers, as well as pass it to the
> > code.gov
> >
> >     people.
> >
> >
> >
> >     Thanks,
> >
> >     Cem Karan
> >
> >
> >
> >     > -----Original Message-----
> >
> >     > From: License-discuss
> > [Caution-Caution-mailto:license-discuss-bounces at opensource.org <
> > Caution-Caution-mailto:license-discuss-
> > bounces at opensource.org > ] On
> >
> >     > Behalf Of Stephen Kellat
> >
> >     > Sent: Tuesday, March 07, 2017 10:41 PM
> >
> >     > To: license-discuss at opensource.org <
> > Caution-Caution-mailto:license-discuss at opensource.org >
> >
> >     > Subject: Re: [License-discuss] [Non-DoD Source] Re: patent
> > rights and the
> >
> >     > OSD
> >
> >     >
> >
> >     > All active links contained in this email were disabled.  Please
> > verify the
> >
> >     > identity of the sender, and confirm the authenticity of all
> > links
> >
> >     > contained within the message prior to copying and pasting the
> > address to a
> >
> >     > Web browser.
> >
> >    >
> >
> >     >
> >
> >     >
> >
> >     >
> >
> >     > ----
> >
> >     >
> >
> >     >
> >
> >     > On Mar 7, 2017, at 10:08 PM, Tzeng, Nigel H. <Nigel.Tzeng at jhuapl.edu < Caution-Caution-mailto:Nigel.Tzeng at jhuapl.edu > >
> wrote:
> >
> >     > >
> >
> >     > > You know the more I think about this, the disclaimer of patent
> > rights in
> >
> >     > > CC0 is probably best for GOSS because it avoids the attempt
> > for
> >
> >     > a one size fit all patent grant language among different
> > agencies with
> >
> >     > different policies and the complexity under which patent rights
> > are
> >
> >     > awarded to whom under the Bayh-Dole Act and Executive Order 10096.
> >
> >     > >
> >
> >     > > Employees of federal agencies, especially research oriented
> > ones, have
> >
> >     > > some financial interest and rights under 10096.
> >
> >     > >
> >
> >     > > Likewise non-profits and small businesses under Bayh-Dole.
> >
> >     > >
> >
> >     > > IMHO patent grant language in FOSS licenses provide a false
> > sense of
> >
> >     > > security.
> >
> >     > >
> >
> >     > > I would rather the government open source as much as possible
> > regardless
> >
> >     > > of patent rights as long as any known patents are disclosed.
> >
> >     > As seen in Ximpleware v Versata the patents typically only cover
> > a small
> >
> >     > portion of the overall system (VTD-XML). While it is relevant
> > from
> >
> >     > the perspective of being able to use the system as built it is
> > less relevant
> >
> >     > from a code reuse perspective.
> >
> >     > >
> >
> >     > > For large government systems significant software components
> > could often
> >
> >     > > be reused without the specific portions covered under
> >
> >     > patent.
> >
> >     > >
> >
> >     > > So just having a copyright license to the entire project would
> > provide
> >
> >     > > significant value to the community. There is code I wrote 30
> > years
> >
> >     > ago I'd love to get access to again even if I couldn't use the
> > rest of the
> >
> >     > system.
> >
> >     > >
> >
> >     > >
> >
> >     > > _______________________________________________
> >
> >     > > License-discuss mailing list
> >
> >     > > License-discuss at opensource.org <
> > Caution-Caution-mailto:License-discuss at opensource.org >
> >
> >     > >
> > Caution-Caution-Caution-https://lists.opensource.org/cgi-bin/mailman/l
> > istinfo/license-discuss
> >
> >     >
> >
> >     > After a less than fabulous day at work for IRS dealing with my
> > tiny corner
> >
> >     > of tax law as well as my accounts work, I am tempted after
> >
> >     > reading this.  Perhaps this could be used as well as the rest of
> > this thread
> >
> >     > as pre-decisional input to open a tight Inquiry in the Federal
> >
> >     > Register.  That's the first step we can take to move into
> > building a formal
> >
> >     > record for a body of law.  Alternatively getting something
> >
> >     > chartered under the Federal Advisory Committees Act might help
> > move this
> >
> >     > forward.
> >
> >     >
> >
> >     > I think the debate has dragged on a bit for more than a few
> > months.  Moving
> >
> >     > to where desirable federal policy/policies are adopted is
> >
> >     > probably doable.  Could we narrow this down to 3 or fewer
> > courses of action
> >
> >     > that might be explored by ARL counsel in an inquiry notice?
> >
> >     > Even if list participants are the only people that respond to a
> > notice in
> >
> >     > the Federal Register we're still building a useful record for
> > later use
> >
> >     > such as Federal Acquisition Rules changes, for example.
> >
> >     >
> >
> >     > Depending upon what shows up in the President's budget set to
> > drop Monday, I
> >
> >     > either will have a lot of time on my hands coming up or
> >
> >     > an ICTAP certificate plus lots of time on my hands.  I want to
> > see Federal
> >
> >     > OSS policy evolve.  We have laid the groundwork here but need
> >
> >     > to get it in the official record soon.
> >
> >     >
> >
> >     > Stephen Michael Kellat
> >
> >     > GS-0962-07/1
> >
> >     > These views are solely my own and not those of the US
> > Government.  Rank,
> >
> >     > position, grade, and bureau are cited for identification
> >
> >     > purposes only.
> >
> >     >
> >
> >     >
> >
> >     >
> >
> >     > _______________________________________________
> >
> >     > License-discuss mailing list
> >
> >     > License-discuss at opensource.org <
> > Caution-Caution-mailto:License-discuss at opensource.org >
> >
> >     >
> > Caution-Caution-Caution-https://lists.opensource.org/cgi-bin/mailman/l
> > istinfo/license-discuss
> >
> >
> >
> >
> >
> > _______________________________________________
> >
> > License-discuss mailing list
> >
> > License-discuss at opensource.org <
> > Caution-Caution-mailto:License-discuss at opensource.org >
> >
> > Caution-Caution-https://lists.opensource.org/cgi-bin/mailman/listinfo/
> > license-discuss < Caution-Caution-https://lists.opensource.org/cgi-
> > bin/mailman/listinfo/license-discuss >
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 6419 bytes
Desc: not available
URL: <http://lists.opensource.org/pipermail/license-discuss_lists.opensource.org/attachments/20170328/f11e14e6/attachment.p7s>


More information about the License-discuss mailing list