Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[John Cowan]

Karan, Cem F CIV USARMY RDECOM ARL (US) scripsit:

> Finally, there is opinion within the US Government that while there is no US
> copyright protection, copyright attaches outside of the US.  Thus, if a
> project is downloaded and used outside of the USA, then any work produced by a
> US Government employee will have foreign copyright protection, and the terms
> of the License should apply to that copyright as well.

Presumably it's the US government that holds the foreign copyright, since
its employees are making works made for hire.

You should probably add back "copyright holder" so that the license can be
applied to works made as a whole or in part by contractors.

The Appendix still says "Apache License".

-- 
John Cowan  http://www.ccil.org/~cowanco...@ccil.org
If I have not seen as far as others, it is because giants were standing
on my shoulders.  --Hal Abelson
___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Gervase Markham]

On 22/07/16 22:01, Karan, Cem F CIV USARMY RDECOM ARL (US) wrote:
> Unfortunately, we cannot directly use the Apache 2 license for all of our
> code.  Most of our researchers work for the US Federal Government and under US
> copyright law any works they produce during the course of their duties do not
> have copyright attached, so we have to rely on contract law as a protection
> mechanism within the USA.

Is there not a large problem here, as you have turned what is currently
a license into a contract, and yet it is still titled "license"?

Does the OSI approve contracts?

And am I about to get various opinionated lawyers telling me that the
license/contract distinction is bogus? :-)

Also, another question for Cem: what do you mean by "a protection
mechanism"? If you are trying to release something under liberal open
source terms like Apache, and it turns out that the USG doesn't hold
copyright in the thing anyway, isn't that "job done"? What needs
"protecting"? Things like misrepresentation are surely already illegal
without you needing to say so in an agreed document.

Gerv
___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Philippe Ombredanne]

On Fri, Jul 22, 2016 at 11:01 PM, Karan, Cem F CIV USARMY RDECOM ARL
(US)  wrote:

> Hi, my name is Cem Karan.  I work for the US Army Research Laboratory (ARL) in
> Adelphi, MD.  I'm in charge of defining the Open Source policy for ARL.  As a
> part of this, we need a license that meets our legal and regulatory needs, but
> is ideally fully interchangeable with everything licensed under the Apache 2
> license as defined at http://www.apache.org/licenses/LICENSE-2.0.txt. We also
> want our license to be fully accepted by OSI as a valid Open Source license.
>
> Unfortunately, we cannot directly use the Apache 2 license for all of our
> code.  Most of our researchers work for the US Federal Government and under US
> copyright law any works they produce during the course of their duties do not
> have copyright attached, so we have to rely on contract law as a protection
> mechanism within the USA.

What about this simpler approach: you could release your software under a choice
of two licenses:
- a public domain dedication (such as CC-0 or your own dedication)
OR
- the Apache-2.0

Could this alleviate the need to create a new license and still
address your needs?

-- 
Cordially
Philippe Ombredanne

+1 650 799 0949 | pombreda...@nexb.com
DejaCode : What's in your code?! at http://www.dejacode.com
nexB Inc. at http://www.nexb.com
___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Lawrence Rosen]

Cem Karan wrote:

> Most of our researchers work for the US Federal Government and under US 
> copyright law any works they produce during the course of their duties do not 
> have copyright attached, so we have to rely on contract law as a protection 
> mechanism within the USA.  

 

I don't understand your worry. Not only can't you assert copyright in the U.S., 
you can't prevent anyone in the U.S. from copying or modifying those works even 
without any license needed from you. You can't even rely on contract law in the 
U.S. to protect against those uses here. I have personally, on occasion, 
considered filing a Freedom of Information Act request for useful government 
code to see if that works to pry free software from government hands. I never 
did that. The U.S. government has almost always proven to be very generous 
without demands.

 

It is true that this public domain result doesn't apply outside the U.S. But if 
you apply a valid open source license to it – such as Apache 2.0 – that should 
be good enough for everyone who doesn't live in the U.S. and irrelevant for us 
here.

 

/Larry

 

-Original Message-
From: Karan, Cem F CIV USARMY RDECOM ARL (US) [mailto:cem.f.karan@mail.mil] 
Sent: Friday, July 22, 2016 2:01 PM
To: license-discuss@opensource.org
Subject: [License-discuss] US Army Research Laboratory Open Source License 
proposal

 

Hi, my name is Cem Karan.  I work for the US Army Research Laboratory (ARL) in 
Adelphi, MD.  I'm in charge of defining the Open Source policy for ARL.  As a 
part of this, we need a license that meets our legal and regulatory needs, but 
is ideally fully interchangeable with everything licensed under the Apache 2 
license as defined at   
http://www.apache.org/licenses/LICENSE-2.0.txt. We also want our license to be 
fully accepted by OSI as a valid Open Source license.

 

Unfortunately, we cannot directly use the Apache 2 license for all of our code. 
 Most of our researchers work for the US Federal Government and under US 
copyright law any works they produce during the course of their duties do not 
have copyright attached, so we have to rely on contract law as a protection 
mechanism within the USA.  We considered using NOSA 1.3 ( 
 
https://opensource.org/licenses/NASA-1.3), but apparently that it isn't 
interoperable with other licenses.  As I understand it, some work on the NOSA

2.0 license was done to improve interoperability, but that work is currently 
stalled.

 

For that reason, we've tried to modify the Apache 2.0 license to fit our needs. 
 The license in full is below and can be diffed with license text found at  
 
http://www.apache.org/licenses/LICENSE-2.0.txt with relative ease; the text has 
also been attached to this email in case my mail client scrambles the 
formatting.  The summary of changes are as follows:

 

1) Changed the name and version number.

 

2) Changed the definition of 'Licensor' to use 'project originator' instead of 
'copyright owner'.

 

3) Changed the definition of 'Work' to not use the word 'copyright'.

 

4) Changed the definition of 'Contributor' to use 'author' instead of 
'copyright owner'.

 

5) Removed the word 'copyright' from the Appendix's instructions

 

6) In the notice, removed the word 'Copyright'

 

7) In the notice, changed 'copyright owner' to Licensor.

 

8) Scrambled the URL as we don't have one yet.

 

In general, we want our license to provide the following:

 

- Protection against liability, not only for ourselves, but also for anyone 
that contributes to our projects.

 

- Protection against fraud where derivative work is represented as being the 
original.

 

- Protection against malicious contributors that intend to entangle projects 
with IP headaches (patent trolls and the like).  This must protect not only 
protect the US Government, but also anyone that downloads our code (just like 
the Apache 2 license does).

 

- The ability to deal with contributions where copyright may or may not be 
attached.  If there is copyright attached, the License should treat it as 
similarly as possible to the Apache 2.0 license.  If there is no copyright 
attached, then the contribution is in the public domain, but it may have patent 
or other IP rights that also need to be dealt with.

 

Finally, there is opinion within the US Government that while there is no US 
copyright protection, copyright attaches outside of the US.  Thus, if a project 
is downloaded and used outside of the USA, then any work produced by a US 
Government employee will have foreign copyright protection, and the terms of 
the License should apply to that copyright as well.

 

Comments on this proposed license are much appreciated.

 

Thank you,

Cem Karan

 

"""

 

 U.S. Army Research Laboratory Open Source License (ARL OSL)

   Version 0.

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Maarten Zeinstra]

Is that the correct interpretation of the Berne convention? The convention 
assigns copyright to foreigners of a signatory state with at least as strong 
protection as own nationals. Since US government does not attract copyright I 
am unsure if they can attract copyright in other jurisdictions.

I am of the belief that the US government does not get assigned copyright in 
other jurisdiction as it is not a right holder in the country of origin. 

Or am I misunderstanding the Berne Convention?

Regards,

Maarten Zeinstra
--
Kennisland | www.kl.nl | t +31205756720 | m +31643053919 | @mzeinstra

> On 22 jul. 2016, at 23:29, John Cowan  wrote:
> 
> Karan, Cem F CIV USARMY RDECOM ARL (US) scripsit:
> 
>> Finally, there is opinion within the US Government that while there is no US
>> copyright protection, copyright attaches outside of the US.  Thus, if a
>> project is downloaded and used outside of the USA, then any work produced by 
>> a
>> US Government employee will have foreign copyright protection, and the terms
>> of the License should apply to that copyright as well.
> 
> Presumably it's the US government that holds the foreign copyright, since
> its employees are making works made for hire.
> 
> You should probably add back "copyright holder" so that the license can be
> applied to works made as a whole or in part by contractors.
> 
> The Appendix still says "Apache License".
> 
> -- 
> John Cowan  http://www.ccil.org/~cowanco...@ccil.org
> If I have not seen as far as others, it is because giants were standing
> on my shoulders.  --Hal Abelson
> ___
> License-discuss mailing list
> License-discuss@opensource.org
> https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss
___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Philippe Ombredanne]

On Fri, Jul 22, 2016 at 11:23 PM, Lawrence Rosen  wrote:
>
> It is true that this public domain result doesn't apply outside the U.S. But
> if you apply a valid open source license to it – such as Apache 2.0 – that
> should be good enough for everyone who doesn't live in the U.S. and
> irrelevant for us here.

Larry, are you suggesting that Cem considers using  some statement more
or less like this, rather than a new license?
This U.S. Federal Government work is not copyrighted and dedicated
to the public domain in the USA. Alternatively, the Apache-2.0
license applies
outside of the USA ?

On Sat, Jul 23, 2016 at 9:51 AM, Maarten Zeinstra  wrote:
>
> Is that the correct interpretation of the Berne convention? The convention
> assigns copyright to foreigners of a signatory state with at least as strong
> protection as own nationals. Since US government does not attract copyright
> I am unsure if they can attract copyright in other jurisdictions.

Maarten, are you suggesting then that the lack of copyright for a U.S. Federal
Government work would just then apply elsewhere too and that using an
alternative Apache license would not even be needed?

-- 
Cordially
Philippe Ombredanne

+1 650 799 0949 | pombreda...@nexb.com
DejaCode : What's in your code?! at http://www.dejacode.com
nexB Inc. at http://www.nexb.com
___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Christopher Sean Morrison]

> I have personally, on occasion, considered filing a Freedom of Information 
> Act request for useful government code to see if that works to pry free 
> software from government hands. I never did that. The U.S. government has 
> almost always proven to be very generous without demands.

Please do, this issue needs to be pressed!  Doesn’t hurt to file a FOIA.

As far as I know, the answer as to whether software is considered an “agency 
record” subject to FOIA requests depends on the nature of the specific software 
and (unfortunately) the agency involved.  Last survey, approximately 50% of 
agencies felt software did not constitute a record, that most software is 
merely a tool for generating or processing records.  Some agencies have and do 
deliver despite holding a position that they are not required to do so.  Others 
feel that even as records, they fall under FOIA exemption clause 5 U.S.C. § 
552(c)(2) — namely that software solely pertains to the internal practices of 
an agency.

DoD is split.  Air Force holds that most software is a record (save for 
classified, T&E, covered-by-statute, or otherwise clearly exempt codes — DoD 
Reg 5400.7).  Navy vaguely concurs (SECNAV 5720.42F).  Army disagrees saying 
they do not consider most software to be a record (AR 25-55).

Much longer discussion:
https://www.justice.gov/oip/blog/foia-update-department-justice-report-electronic-record-foia-issues-part-ii

Also relevant:
https://www.justice.gov/sites/default/files/oip/legacy/2014/07/23/procedural-requirements_0.pdf

Several cases have gone to court making determinations swinging both ways 
(e.g., Gilmore v DOE: software not a record; Cleary, Gottlieb, Steen & Hamilton 
v DHHS: software is a record).  As such, “it depends."

Cheers!
Sean

___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

Re: [License-discuss] US Army Research Laboratory Open Source License proposal

[Maarten Zeinstra]

Hi,

Yes I am suggesting that if the country of origin of the work does not assign 
copyright to the work then no copyright is assigned world-wide. My reasoning is 
that there is no entity to assign that copyright to. 

An example in a different field might support my argument.

In the Netherlands we automatically assign (not transfer, which is important 
here) any IP rights of the employee to the employer if works are created within 
the duties of the employee. That means that the employer is the rights holder. 
This rights holder is consequently also recognised as the rights holder in 
other jurisdictions. Who might, given a similar situation in their own 
jurisdiction, normally assign the right to the employee. 

Now if there is no rights holder to begin with (the U.S. waives it rights on 
government produced works as I understand, the Netherlands government does the 
same), then no foreign rights can be assigned as well. Hence the work must be 
in the public domain world wide.

I have more experience with Creative Commons-licenses than with Open Source 
license, but in CC licenses the license exists for the duration of the right. I 
assume all Open Source licenses are basically the same in this regard. In that 
sense it does not matter which license is applied as the license is immediately 
void, since there is no underlying right to license.

Finally, in the past I have advised the dutch government to adopt CC0 to make 
the public domain status of their works clear. They have adopted this since 
~2011 on their main site: https://www.government.nl/copyright 
 (english version). I advise the US army 
does something similar as well.

Regards,

Maarten Zeinstra

--
Kennisland | www.kl.nl  | t +31205756720 | m +31643053919 | 
@mzeinstra

> On 24 Jul 2016, at 08:26, Philippe Ombredanne  wrote:
> 
> On Fri, Jul 22, 2016 at 11:23 PM, Lawrence Rosen  wrote:
>> 
>> It is true that this public domain result doesn't apply outside the U.S. But
>> if you apply a valid open source license to it – such as Apache 2.0 – that
>> should be good enough for everyone who doesn't live in the U.S. and
>> irrelevant for us here.
> 
> Larry, are you suggesting that Cem considers using  some statement more
> or less like this, rather than a new license?
>This U.S. Federal Government work is not copyrighted and dedicated
>to the public domain in the USA. Alternatively, the Apache-2.0
> license applies
>outside of the USA ?
> 
> On Sat, Jul 23, 2016 at 9:51 AM, Maarten Zeinstra  wrote:
>> 
>> Is that the correct interpretation of the Berne convention? The convention
>> assigns copyright to foreigners of a signatory state with at least as strong
>> protection as own nationals. Since US government does not attract copyright
>> I am unsure if they can attract copyright in other jurisdictions.
> 
> Maarten, are you suggesting then that the lack of copyright for a U.S. Federal
> Government work would just then apply elsewhere too and that using an
> alternative Apache license would not even be needed?
> 
> -- 
> Cordially
> Philippe Ombredanne
> 
> +1 650 799 0949 | pombreda...@nexb.com
> DejaCode : What's in your code?! at http://www.dejacode.com
> nexB Inc. at http://www.nexb.com
> ___
> License-discuss mailing list
> License-discuss@opensource.org
> https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss

___
License-discuss mailing list
License-discuss@opensource.org
https://lists.opensource.org/cgi-bin/mailman/listinfo/license-discuss