User:N. J. A. Sloane/Discussion of Licenses

Discussion of OEIS Licenses

Introduction

• I'm starting this blog on Dec 05 2010. The purpose is to record discussions about the two OEIS licenses -- the contributor's license and the end-user's license.

• Before we launched the OEIS (on Nov 11 2010), four of us (David Applegate, Russ Cox, Terry Ilardi (counsel) and I) spent a great deal of time drafting licenses that we felt were appropriate for the OEIS. We ended up creating new licenses, since it seemed to us that none of the existing licenses (the various Creative Commons licenses and others) met our needs. There are two primary documents, both in plain text, and both part of the Terms of Service agreed to when requesting an account on the OEIS. Neither is in PDF.

• The licenses are The OEIS End-User License Agreement, linked to at the bottom of every OEIS page, and The OEIS Contributor's License Agreement, linked to at the top of the OEIS edit page, and at the bottom of the Wiki edit page.

• You can also see them directly at The OEIS End User License Agreement and The OEIS Contributor's License Agreement

• In the past couple of weeks there have been a very large number of emails about the licenses. The following is just a selection.

William Stein's criticism, Nov 25 2010

The following message was posted to the Number Theory Mailing List on Nov 25 2010

From: William Stein <wstein@gmail.com>
Date: Thu, 25 Nov 2010 21:10:33 -0800
Subject: Re: Announcement, Nov 17 2010: New Version of OEIS!
To: Number Theory List <NMBRTHRY@listserv.nodak.edu>,         "N. J. A. Sloane" <njas@research.att.com>,         sage-devel <sage-devel@googlegroups.com>

On Thu, Nov 25, 2010 at 7:27 PM, N. J. A. Sloane <njas@research.att.com> wrote:
> Announcement, Nov 17 2010: New Version of OEIS!
> ===============================================
>
> After 2 years of struggle, there is a new version of the
> On-Line Encyclopedia of Integer Sequences (the OEIS).
> There are two parts: the main OEIS page, oeis.org,
> which has all the sequence data, and a wiki, oeis.org/wiki,
> which has discussion pages.
>
> To contribute a new sequence or edit a sequence, you must first register
> with oeis.org/wiki - click Register, enter your real name in 2 places,
> enter your email address, etc. When you get issued a password,
> go to the main page, oeis.org, and login with the same user name
> and password.

Hi,

I strongly urge anybody considering registering at the OEIS site to
carefully read the new "OEIS conributor's license agreement", which
you are forced to agree to when you register, even if you submit
nothing.   In it, you'll find: ""Contribution" shall mean any past,
present or future original work of authorship, including any
modifications or additions to an existing work, that is intentionally
submitted by You to the Foundation, or has been previously submitted
by You to the Foundation or any predecessor version of the OEIS, for
inclusion in, the OEIS.".      That's not necessarily so bad in
itself.  However...

If you read further you'll also learn at
http://oeis.org/wiki/The_OEIS_End-User_License_Agreement  that there
are now new and very significant restrictions on using OEIS content:
"To make copies, distribute, make Adaptations and make copies of and
distribute the Adaptations, of no more than 5% of the OEIS Content".
This is in sharp contrast to how OEIS was before, where, e.g., there
was a file  http://oeis.org/stripped.gz  that contained the sequences
themselves.  (This matters to me, since we make them available for
Sage, with a nice interface.)  This was an incredibly useful tool,
since it meant that even without internet access (or in a secure
closed network), one could do searches of OEIS, which is in my opinion
a critical research tool that was built partly by the effort of the
community (you).   Distribution of this stripped.gz file now appears
to be illegal.

I think the recent direction OEIS is going in is unfortunate.  It's
the exact opposite of how, e.g., Wikipedia operates.  Anybody can
mirror Wikipedia content, there is a complete 6GB tarball you can
download that contains all article, etc.

So before signing up for an account, and signing away the rights to
all data or sequences you've submitted, please consider the above.

I will not be registering or contributing to OEIS until the license changes.

 -- William

>
> Now when you look at a sequence, you will see "history" and "edit"
> buttons next to the "graph" and "listen" buttons. The "edit" button
> lets you edit the sequence. To submit a new sequence, use the "Submit
> new seq." link at the bottom of any sequence page.
>
> If you don't login, you won't see the edit button. All contributions
> will be reviewed by the editors (if you would like to help
> by becoming an editor, let me know).
>
> That this works is due to the enormous efforts of Russ Cox,
> who has completely rewritten all the OEIS software.
> David Applegate has also been of great help in getting the new
> system working.
>
> Some 18 months ago we discovered that the mediawiki software
> was incapable of handling the kind of queries that the OEIS generates.
> I had almost given up hope that these problems would ever be solved.
> But thanks to Russ, as of Nov 11 2010, oeis.org is finally running.
>
> This means that after 46 years, I no longer have to process every single
> contribution to the database.
>
> Thank you, Russ and Dave: you have saved the OEIS!
>
> In the Spring of 2009 we set up the OEIS Foundation,
> a 501(c)(3) public charity, which owns, runs and collects
> funds to support the OEIS. The web site is oeisf.org.
> We need donations!
>
> Neil Sloane
> President, OEIS Foundation Inc.
>

--
William Stein
Professor of Mathematics
University of Washington
http://wstein.org

A subsequent exchange of messages

From: William Stein <wstein@gmail.com>
Date: Thu, 25 Nov 2010 21:34:05 -0800
Subject: Re: oeis
To: njas@research.att.com

On Thu, Nov 25, 2010 at 9:24 PM, N. J. A. Sloane <njas@research.att.com> wrote:
>
> William, this is just to you
>
> we'll be happy to make special arrangements with sage and others
> who used the two "stripped" files - there was no intention of
> stopping that
>
> what we are most worried about is a certain large entity that wants
> to get access of all of the OEIS content to use in its web server.
>
> However, maybe we should weaken the license agreements
> I'll have to talk 2 or 3 people, but that won't be possible
> until Monday.
>
> Best regards
> Neil

Hi Neil,

Sorry if I caught you off guard with that email, but it seemed like
everywhere I looked on the wiki page at terms and agreements, it got
scarier, and when I tried to sign up for an account and so that
retroactive clause on all content ever submitted, I felt I better
write something quickly.

I'm really, really glad that you're willing to reconsider the current
license.    We've been having a long discussion about that over here:
http://groups.google.com/group/sage-devel/browse_thread/thread/7e61fde73efb4b65

Note that some of the people who posted in that thread aren't people I
agree with, and definitely don't represent my views.

I want to post a quick followup to my other email, just saying
something like: "I've had some offlist discussions with Neil, and he's
definitely interested in getting feedback from the community about the
OEIS license agreement; nothing is set in stone forever.  However,
this will have to wait until next Monday."

 -- William

A followup message from Tom Boothby

Date: Fri, 26 Nov 2010 12:05:49 -0800
Subject: OEIS Contributor's License
From: Tom Boothby <tomas.boothby@gmail.com>
To: "N. J. A. Sloane" <njas@research.att.com>

Neil,

I'm a Sage developer, a graduate student of mathematics, and a daily
user of OEIS.  Literally* half of the original math I have done has,
at some point, relied on information I've found in OEIS.  To date,
I've only made two contributions, but I recently found a family of
sequences which do not occur in the Encyclopedia, and I am excited to
share them.

However, I'm very concerned by the OEIS Contributor's License --
William Stein reported that he's been in contact with you about the
EULA, and that you wanted feedback about the licensing from the
community.  In short, both the CL and EULA are draconian -- requiring
contributors to utterly sign over their rights for all contributions
past, present, and future is enough to give me serious pause.

If you want a long rationale for my suggestion, I can provide it.
But, my suggestion is brief:

Use a Creative Commons license.  Use it for both the Contributor's
License, and for the EULA.  This is what Wikipedia does, and it works
great.  I'd suggest CC-Share Alike or CC-Non-Commercial.  Both of
these protect the Encyclopedia from being "stolen" or "resold", and
have the benefit of being widely-used and scrutinized -- which is as
good a guarantee of being legally sound as one can get.

Thank you for your time, and for the years of maintaining the Encyclopedia.

  --Tom Boothby

*unlike the rest of my generation, I do not use "literally" to
indicate extreme hyperbole or when I mean "figuratively"

A follow-up message from Doug McNeil

Date: Thu, 02 Dec 2010 23:40:47 -0800
From "Douglas McNeil" <mcneil@hku.hk>

> I have some questions about the current end-user
> license.  It seems like there are lots of things I'm not sure I can
> do, including things I've done in the past and might like to do again
> in the future.
>
> "a single copy": I have multiple digital copies of a (now sadly
> out-of-date!) stripped.gz floating around on various machines, at
> times on computers on three different continents.  Okay or not?
>
> The 5% rule: "To make copies, distribute, make Adaptations and make
> copies of and distribute the Adaptations, of no more than 5% of the
> OEIS Content."  This seems to forbid my sending too-large chunks of
> the OEIS to a friend for processing, although I'd have to guess at
> where the 5% line is.  Maybe stripped.gz is below it?  What's the
> content unit?  How would I find out?
>
> As was recently discussed on sage-devel, this may cause problems for
> OEIS-aware software, and since I'm probably the biggest Sage user
> among the Editors-in-Chief and the one most interested in bulk
> sequence processing requiring large amounts of the OEIS at hand, I'm a
> little concerned.
>
> I guess the idea is to avoid someone taking the OEIS, putting his name
> on it, and publishing it as an independent work to get links (maybe
> with some attribution buried, as many of the Wikipedia-rebranded sites
> do) or as a book to make money.  Or a band of rogue editors breaking
> away and forking their own edition.  I'm no lawyer, but hopefully one
> of them could figure out a tweak that would let good things continue
> and grow?
>
> Doug

Proposal sent to Board of Trustees of OEIS Foundation, Dec 03 2010

From: "N. J. A. Sloane" <njas@research.att.com>
To: BRUGGED@aol.com, david@research.att.com, dsj@research.att.com,         graham@ucsd.edu, mlb@well.com, njas@research.att.com, noe@sspectra.com,         rayjchandler@sbcglobal.net, rkg@cpsc.ucalgary.ca, rsc@swtch.com,         simon.plouffe@gmail.com
Subject: New version of the OEIS licenses

Dear Fellow Trustees of the OEIS:

After further discussions about the trade-offs of various license
options, Dave, Russ and I propose changing both the End-User License Agreement and
the Contributor's License Agreement for the OEIS to the "Creative
Commons Attribution Share-Alike (by-sa)" license.

The license is described here:
http://creativecommons.org/licenses/by-sa/3.0/

This differs from Russ's earlier proposal by removing the
non-commercial restriction, but retains the attribution and
"share alike" clause, which mean that anyone who augments the OEIS
data in some way would have to attribute the OEIS, and make it
available under the same terms.

A few advantages of this are:

* Using a standard, known license helps to reduce debate and the number
  of licenses that ordinary users must know about or understand.

* Allowing replication reduces the risk of the OEIS failing due to
  system or organizational failures.

* It eliminates the need for the OEISF to grant special licenses for
  specific use cases (such as Sage).

* It encourages collaboration by assuring contributors that everything
  they contribute is equally available.  

A few disadvantages are:

* It allows the possibility that someone would create a competitor to
  the OEIS based on a full copy of the content.  However, such a
  competitive version would have to be released under the same
  license, so the OEIS could incorporate any augmentations from the
  competitive version.

* It prevents the OEISF from granting a special use not covered by the
  license, since the OEISF won't have any more rights to the content 
  than it is already granting.

Let me know if anyone objects to this!

Best regards

Neil

Immediate responses from Trustees

• The following are some extracts from the Trustees' initial responses

> It looks like the license question will not be an easy one to resolve.
> Susanna and Tony (and possibly Marc and myself) are not completely
> happy with the proposed Creative Commons by-sa license.

For our next round objective in this email thread we should try to answer:

What nightmare scenarios must we strive to prevent at all costs?

What use cases do we want to encourage, or avoid inhibiting?

If we were to adopt the by-sa license, how would it fail these requirements?

What if instead (devil's advocate) the OEIS was unrestricted public domain?

Then next round we'll be prepared to ask "Given this, can the by-sa license
be made to work, or can you suggest another option that better meets the
requirements?"

and

"What are we missing?"

Date: Fri, 03 Dec 2010 17:22:34 -0800
From: Marc LeBrun <mlb@well.com>

Let's not get rushed!  The OEIS is also an institution and
a community, beyond being a publication or "work"--hopefully it should have
the persistence and shelf life of a Wikipedia or a Britannica, so haste is
contraindicated.

We can probably manage most of the immediate flare-ups by engaging the
persons involved, assuring them that these policies are in flux, not set in
stone, and we're interested in hearing their input and working with them.

Beyond that, I'm not sure I understand what "differentiated from copies",
"infringement" or other such terms mean, concretely, in this context.

For example, a mirror site of oeis.org hosted in a Google or Amazon cloud is
in a functional sense a "copy", but of a very different kind than, say, a
database shipped with Maple, Sage or Mma, and yet again different from some
new "Sequence World" web site, or a hardcopy book of all the sequences, or
disco remixes incorporating them rendered as sounds.

It's worth trying to distill and articulate the primary goals of the
OEIS[F].  I'll submit it's something like:

  Maintaining and developing the premier canonical online "reference
resource of record" providing information regarding integer sequences.

As long as the OEISF isn't impeded in that mission, I'm not sure what
legitimate claim, or justifiable need, we have to proscribe free use.

Date: Fri, 03 Dec 2010 15:48:55 -0800
From: Marc LeBrun <mlb@well.com>

...
I agree the apparent advantages of using the simplest possible standard
agreement are compelling, and the potential downsides are manageable.

In general I think that it's in the best interests of the OEIS to ensure
that contributing to, and using, it is as straightforward and unencumbered
as possible.

Even if it means relinquishing some proprietary interest (eg enabling
Wolfram to build some commercial superstructure incorporating the entire
OEIS data) as long as there's some requirement for attribution, and the OEIS
"brand" remains visible, unambiguous and unsullied, then the OEIS will
remain the "resource of record", as consistent with the charter of the
Foundation.  That is, as long as new contributions and updates can always be
incorporated into the "real" OEIS then other usages are fine.

That said, there are some aspects that may require a little clarification in
adapting the CC SA license for the OEIS.

In particular, does "the work" refer to 1. The particular component of an
entry as submitted by an individual (eg just the sequence, just the code,
just a comment, just a specific correction) 2. a whole OEIS entry (eg all
the lines associated with A000695), 3. the OEIS database in its entirety, 4.
"slices" of the OEIS database (eg all the "stripped" data, all the entries
linked to A000695), and so on.  That is, with respect to the CC SA is the
OEIS THE "work", a collection of "works", both, or something with a more
complex definition?  What is a user or contributor agreeing to license?

Each interesting use case scenario should be analyzed so we understand the
consequences--possibly by a legal professional.

For example: suppose Wolfram (just to pick on them) decides to compress the
OEIS, then extract an index of the Mathematica formulas and look them up
based on what the Mma user was doing.  Is that a fair use?  A derivative
work?  A "remix"?  Should Mma be required to sport a "OEIS Inside" badge?
Do we care?

The simplest alternative of all, of course, is just to put every thing
associated with the OEIS in the public domain.

Followup from Russ Cox, Dec 03 2010

Date: Fri, 3 Dec 2010 17:58:50 -0800
Subject: Re: New version of the OEIS licenses
From: Russ Cox <rsc@swtch.com>
To: njas@research.att.com
Cc: BRUGGED@aol.com, david@research.att.com, dsj@research.att.com,         graham@ucsd.edu, mlb@well.com, noe@sspectra.com,         rayjchandler@sbcglobal.net, rkg@cpsc.ucalgary.ca,         simon.plouffe@gmail.com

It's certainly important to check with lawyers, and
equally important to consider the decision carefully
and taking appropriate time.  Let's not rush into anything.

Regarding the possibility of infringing or unauthorized copies,
I have rather a lot to say, and I apologize up front for the length.
The short version is this:

  * CC-BY-SA gives contributors a reason to trust the OEISF:
    If they decide they don't like how we handle the OEIS, they
    can (legally) take their contributions and go elsewhere.

  * That gives the OEISF strong motivation to do what is best
    for the OEIS, by making the OEIS stewardship subject to
    competition.

  * If for whatever unforseeable reason the OEISF fails to be
    a good steward of the OEIS, the CC-BY-SA license allows
    anyone else willing to put in the time, effort, and money to
    be a good steward to give it a try and (if successful) build
    a thriving community of contributors like we have today.

  * This is most certainly a scary idea, this loss of absolute
    control.  But it guarantees that the intellectual effort of
    the OEIS can live on no matter what might someday happen
    to its stewards.

The long version follows.

The CC-BY-SA license is the one that Wikipedia uses,
and there are certainly "unauthorized" (but legal) copies
of Wikipedia hanging around.  I think that's inevitable
with putting something interesting on the web, regardless
of the license (witness the recent Cooks Source debacle).
But the key point is that no one pays the "unauthorized"
copies of Wikipedia much attention, and everyone knows
that Wikipedia is the real one.  Further, only the real
Wikipedia has the large collection of contributors constantly
making it better.

Similarly, the OEIS is well known in the mathematical
world, and any "unauthorized" copy is likely to be seen
as such.  What's more, a copy does not have what the
OEIS does: the large collection of contributors and editors
working to make it better every day, and also the good will
of the community.

The CC-BY-SA license definitely gives someone else the
legal right to make a copy of the OEIS and try to start their
own "renegade" community.  They would not be able to call
it the OEIS, of course, because that term is trademarked.
And it's a LOT of work to run an OEIS, so that would be no
small undertaking.  A rational person would only go to the
trouble if there were some important reason, like that the
OEISF had gone away or in some other way abdicated its
responsibility to act in the best interests of the OEIS.

That very right - open source people refer to it as the
"right to fork" - is a crucial reason why people are comfortable
contributing to open source projects.  They know that their
contributions can never be lost, in the sense that if the project
loses its way, one way or another, someone else can pick
it up and carry on.  The fruits of a contributor's effort cannot be
later rescinded.

WikiHow even promotes this as a reason to contribute:

    You have the right to fork.

    You have the Right to Fork. In the event that wikiHow's steward
    Jack Herrick fails to act in a manner consistent with our mission,
    the community can move the software and content to a new server
    run by a different steward or organization. This freedom ensures
    that wikiHow will always be run by a person or organization that
    focuses on our shared mission above competing concerns.


    wikiHow and Wikipedia are among the few online communities
    that enjoy this freedom. In most online communities such as MySpace,
    Facebook, and eHow, the software and content are the proprietary
    property of the controlling corporation. Users in those communities
    are locked into those corporations and have only one right: the
Right to Leave.
    Our software is licensed under the GNU General Public License
    and a copy of it can be obtained here. Our content is Creative Commons
    licensed and can be used anywhere under the terms of that license.

    http://www.wikihow.com/wikiHow:Right-to-Fork

Until now, people were comfortable contributing to the OEIS
mainly because they trusted Neil to do the right thing.
(Open source people often refer to this kind of arrangement
as having a "benevolent dictator".)  But now they're being
asked to trust that the foundation Neil is setting up will do
the right thing in perpetuity.  That's more difficult to take on
faith.  Even if you trust the current trustees, what about the
ones 50 or 100 years from now?  The "right to fork" gives the
project accountability: if the OEISF does a terrible job, the
contributors can vote with their feet and move to some other
copy.  Of course, we the OEISF would all hate to see that
happen, which means we should do a good job as stewards
of the OEIS, so that it won't.  But if 100 years from now the
trustees dropped the ball, someone else could take the
OEIS and curate it instead.

Everything I've said up until now is true of any CC or open
source license, not just the one that Neil proposed in his
mail (CC-BY-SA).  The last part - SA or share-alike - adds an
important twist: anyone who wants to put up their own copy
of the OEIS and make changes has to make those changes
available under the same conditions as the actual OEIS,
meaning that the actual OEIS can at any time incorporate the
good changes, meaning a copy cannot "race ahead" of the
OEIS in any meaningful way.

This "right to fork" is very important to the open source
community but actual forks are very rare.  The most notable
case occurred in the gcc compiler in the 1990s.  The gcc
maintainers showed basically no interest in making gcc work
better on the Intel Pentium processor, but that was very
important to enough gcc contributors that they forked a new
project called egcs that added much better support for the
new Intel chips and did various other cleanups.  But then,
because gcc uses a "share alike"-style license, once the gcc
maintainers realized they had made a mistake and that these
changes were in fact in the best interests of gcc, they took
all the improvements from egcs and reincorporated them into
the gcc project, essentially undoing the fork.  Now I certainly
hope that the OEISF will do such a great job that the OEIS
contributors never feel the need to go to such drastic lengths,
but the fact that they could keeps us honest and them
comfortable contributing their time, and the share-alike clause
means that if contributors do "vote with their feet", the OEIS
can follow them as necessary.

For more, I recommend the discussion of forking in the
book "Producing Open Source Software" (itself open source).
http://producingoss.com/en/social-infrastructure.html#Forkability

Ultimately, the best way we can ensure that the OEIS
lives on in perpetuity is to open it so that anyone can make
a copy: even if the OEIS Foundation fails, someone else can
pick up the ball and run with it.  That same fact gives the OEIS
Foundation Trustees an incentive to do the best job they can,
and the share-alike clause means that the OEIS can incorporate
any changes that others publish in other forums in the worst-case
scenario that someone does set up a copycat (in my opinion
extremely unlikely).

It also happens that this copyability means that programs like
Sage can use the OEIS data to make Sage better and the
OEIS data more accessible to Sage users, a win for both sides.
But ultimately that's a good thing.

Again, this concept of forkability is undeniably scary, but it's an
important safety net to make sure that the intellectual effort
lives on even if the curators fail, and it also helps keep the
curators accountable to the contributors, which gives the
contributors a concrete answer to "why should I trust you?".
Although some spammers put copies of Wikipedia articles on
their web site to try to attrack search clicks and ad money,
no one has set up a competing Wikipedia project, because people
are happy with the way Wikipedia the process runs now, so there
is no motivation to put in the time, effort, and money to set up
a second one.  Similarly, as long as the OEISF acts as good
stewards of the OEIS, no one will have an incentive to set up
a competing one.

Susanna made the comment that "You certainly want to be sure
that the OEIS stays on top clearly differentiated from any copies."
That's absolutely correct, but the way to do that is to be the best
one, because of the constant improvements and contributions,
and because of the community of contributors, not because of a
legal document that excludes others from trying.  The competition
that the CC-BY-SA license explicitly allows ensures that 100 years
from now, when none of us are trustees anymore, as long as the
OEIS data is useful to someone, there will be a good curator for it.
Hopefully the OEISF, but if not, someone else as dedicated to it
as we are today.  It's the ultimate safety net.

Russ

I completely agree with Russ.

You can't enforce non copying OEIS - it is a matter of few wget options and trivial stripping of the new format.

I suppose a mailing list/forum which just *discusses* sequence will cause more disadvantages to your foundation than a number of copycat read-only sites.

Georgi Guninski 22:49, 5 December 2010 (UTC)

The discussion becomes more public via the Quomodocumque blog, Dec 04 2010

Date: Sat, 4 Dec 2010 19:15:46 -0800
From: Benoit_Jubin <benoit.jubin@gmail.com>
To: Sequence Fanatics Discussion list <seqfan@list.seqfan.eu>
Subject: [seqfan]  The OEIS mentioned on Quomodocumque

The OEIS is mentioned on the blog Quomodocumque, together with some criticism 
of the new license agreement.  
I'd be interested to know what the OEIS users, 
contributors and editors think about it. Link:
http://quomodocumque.wordpress.com/2010/12/02/licensing-changes-at-the-online-encyclopedia-of-integer-sequences/

Responses from the Sequence Fans Mailing List

The following is a selection from the responses on the Sequence Fans Mailing List.

From: Charles Greathouse <charles.greathouse@case.edu>
Date: Sat, 4 Dec 2010 23:07:10 -0500
To: Sequence Fanatics Discussion list <seqfan@list.seqfan.eu>
Cc: William Stein <wstein@gmail.com>
Subject: [seqfan] Re: The OEIS mentioned on Quomodocumque

Attribution-wise, this is just William Stein's views, familiar to
NMBRTHRY subscribers.

I'm quite sympathetic to his efforts to further 'open' the OEIS.  I
see many times when this would be useful: creating an improvement to
the superseeker, a new browsing tool like Robert Munafo's "Sloandora",
finding and correcting errors in the OEIS, and so forth.  I know that
I have used stripped.gz or cat25 many times for various projects over
the years, and surely others have creative uses I haven't even
imagined.

This is a complex issue which deserves study; I wouldn't want to rush
to a decision like 'no, the OEIS is open enough'.

On the other hand, I think Stein is wrong to conflate issues of access
and legality.  Before, there was no licensing agreement allowing
distribution or use of the OEIS data, but it was easy to download
certain data (e.g., the sequence values in stripped.gz).  Now there
are legal ways to use some data.  Using data as complete as
stripped.gz are on the same shaky legal grounds as it was before the
OEIS End-User License Agreement.  But now there is no access.

So I see these issues before us:
* Determine the legal issues surrounding changes to the OEIS End-User
License Agreement.
* Determine competing needs for openness and closedness, and an
appropriate level of openness based on thise
* Make appropriate data available, or minimally tools to allow as much
freedom to our users as possible.

If for reasons I don't currently appreciate it is impossible to make
the data as open as I would like, we should at least build tools that
allow common and "good" uses to whatever extent we can.  For example,
if a project like Sage wanted to be able to do analysis with the OEIS,
building an API that it could use would be a minimal courtesy if we
could not (for whatever reason) sufficiently open the data.
Similarly, for data validation, building an interface that would allow
searching the OEIS with regular expressions would allow at least a
third of the common cases (better than nothing).

Of course these approaches would cause more work for the OEISF
(probably Russ) and would cause increased server load, rather than
offloading both to outside developers like Stein.  This is another
argument in favor of openness, albeit a small one.

I hope we can avoid hasty conclusions on this complex issue.

Charles Greathouse
Analyst/Programmer
Case Western Reserve University

Date: Sun, 5 Dec 2010 14:30:48 +0000 (UTC)
From: "Joseph S. Myers" <jsm@polyomino.org.uk>
To: Sequence Fanatics Discussion list <seqfan@list.seqfan.eu>

I think the purposes of the OEIS would be best served by

(a) licensing it all under copyleft terms such as CC-by-sa;

(b) making the whole database (internal format of the current approved 
versions of all sequences, and all the A-files and B-files) readily 
available to download;

(c) also making the data from the history of the sequences available to 
download.

-- 
Joseph S. Myers
jsm@polyomino.org.uk

From: Marc LeBrun <mlb@well.com>
To: Sequence Fanatics Discussion list <seqfan@list.seqfan.eu>

As Neil mentioned, the trustees have been actively discussing the various
options the OEIS license.  Everyone's thoughtful comments and suggestions
here and off-list are being considered and are much appreciated.

One option under discussion is the "Creative Commons Attribution-Share
Alike" ("CC by-sa" for short) which is the license used by Wikipedia:

http://creativecommons.org/licenses/by-sa/3.0/

We would be interested if you have any comments or concerns about any issues
that might arise should this license be adopted.  For example do you see any
adverse impact on your own or others' use of the OEIS?  Would you foresee
any potential problems for contributors, or for the OEIS itself?

While all seqfan input about the OEIS is of interest, this message is to
solicit responses from the community specifically regarding the CC by-sa
license, particularly any problems or "unintended consequences".

Please keep in mind that CC by-sa is just one of the possible choices, and
the trustees are still investigating alternatives.  Also any license will
require legal vetting before adoption.  Your patience and support for this
process helps ensure that the outcome is in the best interests of the OEIS.

Thanks!
  --MLB

Date: Sun, 5 Dec 2010 23:05:35 +0100
From: "peter.luschny" <peter.luschny@googlemail.com>

> One option under discussion is the "Creative Commons Attribution-Share
> Alike" ("CC by-sa" for short) which is the license used by Wikipedia:
> http://creativecommons.org/licenses/by-sa/3.0/

I am in favour of 'cc by-nc-sa', same on attribution and share alike
as "cc by-sa" but for non-commercial use.
http://creativecommons.org/licenses/by-nc-sa/3.0/

The reason? I hate to see the Bernoulli and the Euler numbers
squeezed in between advertisements for hamburgers and for 'singles'
looking for dates. Jakob und Leonhard haben das nicht verdient! :)

What I am describing is what I see right now when I look on the
page for 'Bernoulli number' on answers.com; the entry in between
is copied from Wikipedia.

Peter

Date: Sun, 5 Dec 2010 15:35:11 -0800
From: Russ Cox <rsc@swtch.com>
Subject: [seqfan] Re: OEIS license: CC by-sa option?

> I am in favour of 'cc by-nc-sa', same on attribution and share alike
> as "cc by-sa" but for non-commercial use.
> http://creativecommons.org/licenses/by-nc-sa/3.0/

Using CC-BY-NC-SA would mean, if we ever invoked the SA
clause to bring any external changes back into the OEIS, that
it could never be put to any commercial uses, even if the
OEIS Foundation wished to approve such uses.  One important
commercial use is the publication of the next book version,
and there are doubtless others (like the ones explicitly allowed
by the current license).  That is the rationale behind proposing
CC-BY-SA instead of CC-BY-NC-SA.

Russ

Date: Mon, 6 Dec 2010 01:22:31 +0100
From: "peter.luschny" <peter.luschny@googlemail.com>
To: seqfan@list.seqfan.eu
Subject: [seqfan] Re: OEIS license: CC by-sa option?

Maybe I do not understand this. Nevertheless let me cite this statement
from wiki.creativecommons.org/Frequently_Asked_Questions .

| the noncommercial license option is an inventive
| tool designed to allow people to maximize the distribution
| of their works while keeping control of the commercial aspects
| of their copyright. To make one thing clear that is sometimes
| misunderstood: the "noncommercial use" condition applies only
| to others who use your work, not to you (the licensor). So if
| you choose to license your work under a Creative Commons license
| that includes the <93>noncommercial use<94> option, you impose the
| <94>noncommercial<94> condition on the users (licensees). However,
| you, the creator of the work and/or licensor, may at any time
| decide to use it commercially. People who want to copy or adapt
| your work, "primarily for monetary compensation or financial gain"
| must get your separate permission first.

From Russ Cox, Dec 05 2010

> Maybe I do not understand this. Nevertheless let me cite this statement
> from wiki.creativecommons.org/Frequently_Asked_Questions .

The text you quoted is true only for the licensor.  Right now that
is indeed the OEIS Foundation.  But in my text above I said
"if we ever invoked the SA clause to bring any external changes
back into the OEIS".  So for example if someone else set up
their own site with a cleaned up OEIS, the SA clause would give
us the right to incorporate their modifications.  *BUT* it would
only give us the CC license rights to them, meaning that we
would then be bound by it w.r.t. the imported changes or else
have to exclude them from any future commercial use.

Russ

From: "T. D. Noe" <noe@sspectra.com>, Dec 05 2010
Subject: Re: OEIS license: CC by-sa option?

I researched [the various Creative Commons licenses] over a year ago when we 
were first starting to talk about it.  My first reaction to the current crisis is:  
(1) we hashed this out before and (2) can't we just tweak the current 
agreement to give Sage a license?

What is good for Wikipedia is not necessarily good for OEIS.  I'm certain
that we have different goals.  In fact, goals is where this whole
discussion should (and did) begin.  Choosing CC BY-SA (or any other
agreement) is a legal question which requires input by knowledgeable
attorneys.  We have already gone this route.  Hence, I am strongly in favor
of tweaking the current agreement.

Best regards,

Tony

From Russ Cox, Dec 5 2010

It's possible that we can make a minimal change to the current
license to support our goals.  It's also possible that we can't.
I think we'd need legal input for any change, though, so I don't
think that is an argument for or against "tweaking".

I think your point about starting with goals (again) is a very good
one, because we clearly missed some the last time.  I am going
to think about this more and write an email tomorrow with what
I think the relevant goals are.  You're right that "it's good for
Wikipedia" is not an argument in favor, but neither is it an argument
against, necessarily.

Do you happen to have a copy of the goals we listed last time?
I looked at the goals for the OEIS Foundation on the web page,
but they do not really state the goals for the OEIS itself.

A final point: any discussion about licenses is always frustrating.
It often seems that the internet has nothing better to do than
focus on licensing terms.  But the license we choose now applies
to the OEIS for all time, long after we're gone, so it's important
to get it right.  I think it's fine (if more frustrating) for the process
to take a while, so that we do get it right.

Russ

Russ Cox's reply to the Quomodocumque blog, Dec 05 2010

This is quite a bit blown out of proportion.

The OEIS never had a license before. It was available as a file on a
web server, copyrighted by AT&T, with no explicit license granting
anyone redistribution rights, though of course people did anyway. The
situation was murky at best.

The OEIS recently transferred ownership from AT&T to the newly
established OEIS Foundation, and more recently transferred editing
mechanisms from mailing changes to Neil to a web-based, collaborative,
wiki-like editing system. These have both been very big transitions
for the OEIS, and neither is yet complete.

As part of transitioning to ownership by the new OEIS Foundation, we
(the trustees) added an explicit license. The fact that it grants
anything at all makes it less restrictive than what was there before,
but we recognize that it is not right yet. William Stein raised some
valid objections to the license. We definitely think that the uses he
raises should be allowed, so the license will have to change. The OEIS
Foundation Board has been discussing alternate licenses for the past
few days.

That said, the OEIS is 45 years old and intends to be a lasting
reference work; it does not move at <E2><80><9C>internet speed<E2><80><9D>. The transitions
I mentioned above have themselves already been going on for over a
year. I am sure that we will end up with the right license, one that
works for the OEIS and works for its users, Sage and William Stein
among them. But it will probably take a few weeks, maybe even into
January due to the upcoming holidays.

Please be patient. All will turn out right.

Further comments, especially in regard to the CC-BY-SA suggestion

Date: Mon, 6 Dec 2010 11:58:08 -0500
From: Alonso Del Arte <alonso.delarte@gmail.com>
To: Sequence Fanatics Discussion list <seqfan@list.seqfan.eu>
Subject: [seqfan] Re: OEIS license: CC by-sa option?

The goal of the license, in my opinion, should be twofold:

1) To provide unhindered access to honest researchers; and
2) To deter charlatans from profiting who contribute neither work nor
creativity.

But to have someone who is not a lawyer write the license could cause the
opposite to happen: researchers feel too restricted, and crooks find the
loophole to take what they want and it's perfectly legal.

Al

From Ron Hardin, Dec 06 2010, rhhardin@mindspring.com

----- Original Message ----
> From: Alonso Del Arte <alonso.delarte@gmail.com>
> Subject: [seqfan] Re: OEIS license: CC by-sa option?
> 
> The goal of the license, in my opinion, should be twofold:
> 
> 1) To provide  unhindered access to honest researchers; and
> 2) To deter charlatans from  profiting who contribute neither work nor
> creativity.

I think that worrying about free-riders is misguided.  The chief thing is load 
on the website.

As to somebody else building a better site based on it, more power to them.  
Competition for better quality and more usefulness is good.

An interesting podcast on intellectual property in general, and why it might not 
empirically be a good idea, links off of
http://www.econtalk.org/archives/_featuring/michele_boldrin/

Chiefly you get monopoly rent-seeking from intellectual property, without any 
increase in invention, is the idea.

Maybe preventing monopoly rent-seeking ought to be the plan.

Added by Marc LeBrun, Dec 07 2010, mlb@well.com

I have been exploring the CC licenses for a few days (including applying them to
one of my own works to get a feel for the process) and conclude the following:

  *  I concur and recommend that the OEIS be offered under the CC by-sa license. 

  *  To the extent I can determine, the spirit of the CC by-sa concept seems 
entirely compatible with the objectives of the OEIS and the OEISF's charter.

  *  Switching to this standard license from the OEIS's current unique licenses 
forestalls controversy, distraction and other overhead for both the OEIS and its 
users and contributors.

  *  As appealing as it might seem, I recommend AGAINST trying to control 
commercial or otherwise "unsavory" use, as this burdens the OEISF with more 
enforcement, yet seems to provide no benefit to the OEIS itself.  Specifically, 
I recommend against adding the "NC" license option.

  *  I think it's necessary to ALSO use a regular copyright notice along with CC 
license, in order to actually establish the ownership of the IP being licensed.

  *  I think it's necessary to ALSO add a "Terms of Use" agreement to the web 
site, as the scope of the license is only the content.  Such terms would address 
abusive loading of the servers and similar issues.  (Wikipedia does this).

  *  These conclusions must of course be verified by legal experts, particularly 
to identify any substantive differences between the protections afforded by CC 
by-sa as compared to the requirements that drove the current licenses.

Russ Cox, Dec 07 2010:

Marc LeBrun and Tony Noe made the excellent suggestion
that laying out the goals for the OEIS would help guide us
in the discussion of licenses.

I believe the goals of the OEIS are as follows:

 * Continue to be the premier source for integer
  sequence and related information.

 * Serve as the coordination point for people
  interested in integer sequences.

 * Maintain the current high standards of accuracy
  and scientific rigor expected of a mathematical
  reference work.

 * Be as accessible and useful to as many people and in
  as many forms, media, and contexts as possible.

(It is the final goal that I believe we did not emphasize
enough the last time.)

The core of the complaints about the current license
center on the redistribution terms.  People want to be able
not just to download a copy of the OEIS for themselves
but to be able to pass it on to other people, include it in
their own programs or works, and generally share it as
much as they have in the past.

The restricted redistribution terms cut into the final goal
of making the OEIS as accessible and useful as possible.

To understand what kinds of redistribution we'd want to
allow, consider these concrete examples:

 * Distribution of the OEIS data as part of Sage to implement
   the sloane_find(2,3,5,8,13,21) function.

 * Distribution of the OEIS data as part of Mathematica to
   implement an imagined OEISLookup[2,3,5,8,13,21] function.

 * Incorporation of OEIS content into MathWorld as long as
   it is properly attributed (links back to oeis.org).

 * Inclusion of OEIS content on a free CD-ROM of mathematical
   and scientific resources for school children
   (with proper attribution).

 * Inclusion of OEIS content on a similar CD-ROM for sale.

 * Inclusion of OEIS content in a freely available PDF including
   the best OEIS sequences, nicely formatted (again with proper
   attribution).

 * Inclusion of OEIS content in a hardcopy book based
   on the PDF.

Neil has said that the first three uses would be okay but
was uneasy about the final four, mainly out of concern
that the good reputation of the OEIS not be tarnished with
sub-par publishing.  Unfortunately, it is very difficult to
distinguish, say, distribution of OEIS data with Mathematica
from distribution of OEIS data on a for-pay CD-ROM, and
it is similarly difficult to distinguish the CD-ROM from the
book (notes [1], [2]).  I believe that the accessibility goal
implies that we should allow all of these uses and instead
maintain the good name of the OEIS by controlling the use
of the trade mark rather than the content (note [3]).

I understand that allowing arbitrary copying, redistribution,
and adaptation of the OEIS Content can be a scary idea.
However, it is common for open collaborative data sets
and software and in all the cases I know has led to more
benefits for the project than drawbacks; and further we
can still defend the reputation of the OEIS trade mark
(see again note [3]).

For me the scariest possibility is that we continue to try to
impose restrictive terms on copying, redistribution, and
adaptation of the Content.  I think that, by far, the most likely
reason for someone to set up a competing database would
be to obtain one with more permissive licensing terms.
The recent history of technology in general and the Internet
in particular boils down to "you can be the open version,
or you can be replaced by it," whether in science or software
or standards or many other arenas.


Of course there are many possible ways to implement
this broader allowance of redistribution.  One is to replace
the current EULA and CLA with CC-BY-SA (the "CC-BY-SA option")
as has been discussed, and another is to adapt the EULA
as needed and keep the CLA intact.

The main point in favor of the CC-BY-SA option is that in the
event of a competitor to the OEIS, the SA clause gives us
the legal right to fold any improvements back into the OEIS.
I've written before about how if the OEIS is sufficiently open
and the OEISF acts as a good steward, such "competitors"
are unlikely to arise.

The main point in favor of the adapted EULA and current CLA
is that the current CLA gives the OEISF very broad authority
to relicense the OEIS under a new license as necessary to
adapt to changing technology or copyright laws.  (In contrast
the CC-BY-SA option would fix the license at CC-BY-SA
irrevocably.)

Since one of the charges of the OEISF Board is to "take the
long view" toward perpetuating the OEIS and adapting it to
future media and contexts, I believe the long-term flexibility
of being able to adapt the license to future events weighs
strongly in favor of keeping the current CLA and changing
the EULA to permit redistribution and adaptation of the
OEIS content with proper attribution.


The exact wording would need to be worked out by legal
counsel, but my proposal is to keep everything as it is and
change Section 2 License Grant to allow arbitrary copying,
redistribution, and adaptation of the OEIS Content.
(http://oeis.org/wiki/The_OEIS_End-User_License_Agreement)


Russ


____________________________________________________________
Notes

[1] One possible avenue is to exclude distribution of
derivative works, but even that is fraught with trouble in
an age of executable content: I can distribute any of the
above examples (except the paper book) with a pristine,
unmodified copy of the OEIS data and simply process it
on the fly as necessary.

[2] Another possible avenue is to require explicit agreements
with each proposed use of the OEIS.  I think this is untenable
in the long term: the OEISF should not be in the position of
deciding who can and cannot use the OEIS in their work.

[3] The current EULA contains a restriction that

 "Nothing contained in this Agreement shall be construed
 as conferring any right to use in advertising, publicity,
 or other promotional activities any name, trade name,
 trademark, trade dress, or other designation of the Foundation."

so it would (I believe and we should confirm with legal counsel)
be well within our rights to prohibit someone from publishing
a sub-par reference book containing OEIS content as "The OEIS
Handbook".  They could however sell an equivalent book
titled "A Sequence Lookup Dictionary" with appropriate
attribution inside the book, but they would not be able to
use the name of the OEIS in advertising or publicity.

Date: Tue, 7 Dec 2010 14:59:37 -0800
To: njas@research.att.com
From: "T. D. Noe" <noe@sspectra.com>
Subject: Re: New proposal from Russ for the licenses

There are people that want to keep this topic on the front burner.  I think
we should resist (and to continue the metaphor) and let it simmer.  My
understanding of CC-BY-SA is that it is a one-time decision -- you can't go
back.  That in itself is grounds for going slow.

Best regards,

Tony

Date: Wed, 8 Dec 2010 10:45:46 -0800
To: Marc LeBrun <mlb@well.com>
From: "T. D. Noe" <noe@sspectra.com>
Subject: Re: I've concluded the OEIS should adopt CC by-sa
Cc: Neil Sloane <njas@research.att.com>

Without using Google, how many organizations other than Wikipedia do you
know that use cc-by-sa?  In the Wikipedia article about CC licenses,
http://en.wikipedia.org/wiki/Creative_Commons_licenses, there is a partial
list that is under-whelming.  I think the short list is a red flag.

Here's a link to a foundation that decided not to use cc-by-sa:
http://www.osmfoundation.org/wiki/License/Why_CC_BY-SA_is_Unsuitable

Best regards,

Tony

Resolution, February 2011

After two months of discussion (not shown here), the OEIS Foundation has adopted the Creative Commons Attribution Non-Commercial 3.0 license (CC-BY-NC) as its End-User License Agreement. The Contributor's License Agreement will remain unchanged. Neil Sloane President, OEIS Foundation Inc