Revisiting licensing

[puzio1.excite]

Now that we are ready to roll out the new version of the site software
which will make it possible manage content in new ways, it might be a
good time to reconsider some of our policies, in particular licensing
policy. We had some preliminary discussion on this topic at today's
board meeting and, in the interest of inviting more people to weigh in
and have more time to ponder this topic, would like to carry on the
discussion here in the forum and on the listserver.

The current policy is to license encyclopaedia entries under a copyleft
license (originally GNU FDL, now CC-BY-SA) and to allow books, expositions,
and papers to be licensed under any license which allows us to distribute
them (as listed in the "rights section"). As I see it, it is convenent
to split this question of licensing into two directions.

One direction deals with licenses whose terms are more liberal than the
CC-BY-SA license. At the level of principles there is not much to discuss
here --- since our goal is free math, clearly we accept content that is
made available on more liberal terms. Likewise, there is no problem reusing
such content or combining it with existing content. Rather, the question
is one of of practice. Whithout some systematic way of indicating that an
item is available under more liberal terms, users are unlikely to know and
might, for instance, think that a work in the public domain was still
copyrighted. However, setting up a system which allows for multiple licenses
in a collaborative environment will likely require some work and thought
to do correctly in a way that meshes with the workflow and doesn't confuse
users or make a headache for administrators.

A suggestion I have in this regard is to limit to the CC family of licenses.
Such a limitation will avoid the headache of having to keep track of
numerous, likely contradictory licenses while still allowing a meaningful
pallete of alternatives. Given that we already are using one of their
licenses and this family of licenses is quite popular and comes with computer
support which we could plug into our platform, this seems like an option
worth considering.

The other direction has to do with licenses stricter than the CC-BY-SA license,
specifically the question of whether we should host copies of or display works
released under such terms --- I don't think there's any controversy that we should
be making bibliographic references or links to any mathematical work whatsoever,
including "all rights reserved". As I see it, the question further bifurcates
depending on the two main classes of conditions which stricter licenses impose.

One condition is prohibiting commercial use. The importance of this condition
is somewhat different for scientific than artistic works since they typically
don't generate significant royalties for their authors. Nevertheless, some
authors wish to release mathematical works under such terms out of principle.
Of course, since PM is a non-profit organization and doesn't charge for access,
such a condition would not interfere with our using works released under such
terms, althogh it would prevent us from including them in compilations like
the FEM unless we would negotiate some arrangement with their authors.

A more serious condition is that of not allowing derivative works since such a
condition severly limits the utility of such works on PM --- we would be limited
to displaying and distributing them verbatim and prevented from reusing them
in creative ways. It seems to me that the question of whether to allow works
under "no derivative" terms would be most relevant when it comes to papers because
it would determine whether we host reasearch articles by authors who believe in
open access but might be reluctant to grant blanket permission to produce derived
works because they are concerned that an ill-concieved derived work might affect
their reputations adversely.

What do you all think about these issues? As I mentioned earlier, we had some
discussion at and leading up to the board meeting (maybe Joe and Aaron could
post some of their thoughts on the matter) and would like to hear more thoughts
and suggestions on this topic from the community.

Raymond Puzio

[Aaron Krowne]

I agree that prohibiting commercial use and derived works are the thorniest issues.

Let me point out firstly that producing something like the FEM is not necessarily (and indeed, not even likely) a "commercial work".  That is because, if it is sold under a nonprofit entity, it is lawfully NON-COMMERCIAL.  It doesn't matter that money is being exchanged for hard copies.  So that takes care of that.  

As I mentioned on the call, I think non-commercial and no-derived-works terms should be available, but should (1) be prominently demarcated, (2) should involve some measure of segregation (i.e. works with these terms cannot be in the encyclopedia), and (3) should be discouraged.

We might point out, for example, that the real issue with derived works based on research papers is fraud and misrepresentation, not the possibility of a derived work per se.  And if someone is going to engage in fraud and misrepresentation, they aren't likely to be deterred by additionally having a license restriction.

-Aaron

--
You received this message because you are subscribed to the Google Groups "PlanetMath" group.
To post to this group, send email to plane...@googlegroups.com.
To unsubscribe from this group, send email to planetmath+...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/planetmath?hl=en.

[puzio1.excite]

Here are my thoughts on non-commercial license terms. In short, I think that we should accept
content that comes with such restrictions.

Even if license terms restricting sale (say something stricter that just "no commercial use" that
explicitly prohibits printing multiple copies and offering them for sale irregardless of profit) would
keep us from not including some content in HDM, I do not consider that as sufficient reason for PM
to reject that content because: a. There still would be plenty enough material available for making
a compilation. (And, if there were a few off-limit items that we really would like, someone could
always re-express the relevant facts.) b. Those items would still be available on the site and
could be modified and reused online.

An potentially important class of use which I think we should permit is one where an author
has a work published but arranges with the publisher to make it freely available (and permits
it to be reused) subject to a license condition to make sure that the publisher retains exculsive
printing rights. In fact, I would even go so far as to suggest that we encourage authors to ask
their publishers to allow their preprints or books to be released under non-commercial terms
as well (and consider goinhg to a different publisher should the answer be negative).

[puzio1.excite]

Whoops ----

> keep us from not including some content in HDM,

in the last post should have been

> keep us from not including some content in FEM,

[puzio1.excite]

> As I mentioned on the call, I think non-commercial and no-derived-works terms should be available, but should
> (1) be prominently demarcated,
> (2) should involve some measure of segregation (i.e. works with these terms cannot be in the encyclopedia),

> (3) should be discouraged.

Along these lines, I would like to add

(3') using more liberal license terms should be encouraged
(4) users should be able to filter views by license and set defaults in their personal preferences

This way, for instance, a manager at a company lab who is worried about the Texaco verdict might require
employees to set their preferences not to show works with a NC license or someone who considers
copyleft as a dangerous virus could choose to block content with SA license terms. This would also
provide an incentive for people to release under the most liberal license terms possible because
that makes their work visible to a larger audience.

Regarding points 2, 3, and 3', I would like to see those implemented in an "emergent, market-like" manner
rather than being imposed from above. For point 2, all I think would be needed is to allow mangers of
collections to impose licensing conditions for inclusion automatically (analogous to filtering preferences for
users) and let the managers' self-interest take care of the rest --- for instance, someone curating an
encyclopaedia-type collection would naturally want to permit production of derivative works and someone
curating a collection whose aim to produce a book and interest a publisher like Dover or Springer in carrying
it would naturally not allow NC terms.

The general principle I would like to advocate here is that we should internalize the externalities associated
with licensing choices through appropriate incentives and disincentives. The reason we don't like restrictive
licensing terms is because they restrict the utility of works to the scientific community and therefore should
strive to make the benefits to the user for posting content on PM proportional to the potential utility of that content.

> We might point out, for example, that the real issue with derived works based on research papers is fraud
> and misrepresentation, not the possibility of a derived work per se. And if someone is going to engage in
> fraud and misrepresentation, they aren't likely to be deterred by additionally having a license restriction.

That is certainly a good point that should go into the New User Guide. As for the latter point, someone might
impose license restrictions not so much as a deterrent but because copyright law offers them extra tools for
enforcement, like sending letters to websites to take down the content. (For the record, I don't approve, but
this sort of thing is common enough.)

I'm kind of skeptical of you or me trying to figure out by ourselves what would be acceptable to a researcher
who is worried about having results stolen or reputations damaged by inappropriate derived works because
we don't share their worries, so it would be best to ask them instead. After all, the way this point originally
came to our attention was when Borwein complained about what happened to him. Therefore, if you all
approve, I am proposing to contact our Advisory Board for their opinions and advice regarding this licensing
question, especially about what measures they would feel are important to protest their personal integrity.

[Joe Corneli]

I agree with the general ideas here. What I like best is that this is
starting to form into a case FOR "free math". I think we need to work
hard to make that argument, because there are a lot of people who
wouldn't even consider NC/ND terms, much less CC-By-SA or more liberal
terms.

> Therefore, if you all
> approve, I am proposing to contact our Advisory Board for their opinions and advice regarding this licensing
> question, especially about what measures they would feel are important to protest their personal integrity.

Personally I'd rather delay that until we actually have programmer
hours to contribute towards implementing the solution. (Or, if it's
already implemented, and all we need to know is how to "tune" it.)

[puzio1.excite]

>> Therefore, if you all
>> approve, I am proposing to contact our Advisory Board for their opinions and advice regarding this licensing
>> question, especially about what measures they would feel are important to protest their personal integrity.
>
> Personally I'd rather delay that until we actually have programmer
> hours to contribute towards implementing the solution. (Or, if it's
> already implemented, and all we need to know is how to "tune" it.)

Would you be comfortable with me simply stating something like the
call I put out to the PM community which is focused on the policy
issues and leaves out the business about software implementation
which we have been discussing here since the meeting?

As for me, I'd rather not delay getting in touch with the advisory
board because it's literally been a year since we last said anything
to them was when we told them that they we decided to appoint
an advisory board and that they were on it! This licensing issue
looks like a good occasion to break the silence and ask them
for some advice. Later on, when we get programmers, we could
ask them about their thoughts on software issues but, for now,
I was thinking of starting the conversation by asking them about the
purely policy end of things.

[Joe Corneli]

On Fri, Jan 25, 2013 at 10:31 PM, puzio1.excite <puz...@excite.com> wrote:
>
>>> Therefore, if you all
>>> approve, I am proposing to contact our Advisory Board for their opinions and advice regarding this licensing
>>> question, especially about what measures they would feel are important to protest their personal integrity.
>>
>> Personally I'd rather delay that until we actually have programmer
>> hours to contribute towards implementing the solution. (Or, if it's
>> already implemented, and all we need to know is how to "tune" it.)
>
> Would you be comfortable with me simply stating something like the
> call I put out to the PM community which is focused on the policy
> issues and leaves out the business about software implementation
> which we have been discussing here since the meeting?

Not really, and exactly for the reason you say below:

> As for me, I'd rather not delay getting in touch with the advisory
> board because it's literally been a year since we last said anything
> to them was when we told them that they we decided to appoint
> an advisory board and that they were on it! This licensing issue
> looks like a good occasion to break the silence and ask them
> for some advice.

Getting in touch with them is good, but it's hard for me to see how
licensing is the top priority to focus on. If we want to talk with
them about something, let's talk about:

- the new software (and what it might mean for the next decade of PlanetMath)
- the new functionality it brings (and what it may mean for their involvement)
- the fundraising and membership issues that haven't been resolved yet
- the overall plan
- etc.

As for licensing issues, I think it would be better start a discussion
with some people in the broader mathematical community who have
already made "licensing" a key issue (for themselves). We can bring
our findings back to the advisory board if we still need their advice
after that.

> Later on, when we get programmers, we could
> ask them about their thoughts on software issues but, for now,
> I was thinking of starting the conversation by asking them about the
> purely policy end of things.

Again, I don't think we're quite at the point where we can talk to
them about this strategically, but we're potentially close.

I think what would be very helpful would be to draw up a new strategic
plan (this doesn't have to take more than an hour, and therefor I'm
sure we could do it tomorrow when we talk) -- this can include "inputs
required", including places where we may need strategic advice.

Joe

[puzio1.excite]

>> Would you be comfortable with me simply stating something like the
>> call I put out to the PM community which is focused on the policy
>> issues and leaves out the business about software implementation
>> which we have been discussing here since the meeting?
>
> Not really, and exactly for the reason you say below:
>
>> As for me, I'd rather not delay getting in touch with the advisory
>> board because it's literally been a year since we last said anything
>> to them was when we told them that they we decided to appoint
>> an advisory board and that they were on it! This licensing issue
>> looks like a good occasion to break the silence and ask them
>> for some advice.
>
> Getting in touch with them is good, but it's hard for me to see how
> licensing is the top priority to focus on.

Then I suppose that the decision to go ahead and solicit opinions
for the community was a mistake because I thought that asking the
advisory board to chime in would be an obvious and uncontroversial
part of asking for community input.

> If we want to talk with
> them about something, let's talk about:
>
> - the new software (and what it might mean for the next decade of PlanetMath)
> - the new functionality it brings (and what it may mean for their involvement)
> - the fundraising and membership issues that haven't been resolved yet
> - the overall plan
> - etc.

The overall topics look plausible, now it seems an issue of clarifying
them and boiling them down to specific talking points and questions.

> As for licensing issues, I think it would be better start a discussion
> with some people in the broader mathematical community who have
> already made "licensing" a key issue (for themselves). We can bring
> our findings back to the advisory board if we still need their advice
> after that.

Specifically, who do you have in mind?

> I think what would be very helpful would be to draw up a new strategic
> plan (this doesn't have to take more than an hour, and therefor I'm
> sure we could do it tomorrow when we talk) -- this can include "inputs
> required", including places where we may need strategic advice.

That's fine --- I look forwards to talking tomorrow.

[Joe Corneli]

On Sat, Jan 26, 2013 at 2:06 AM, puzio1.excite <puz...@excite.com> wrote:

> Then I suppose that the decision to go ahead and solicit opinions
> for the community was a mistake because I thought that asking the
> advisory board to chime in would be an obvious and uncontroversial
> part of asking for community input.

My point is that they are rather busy people and we should approach
them in a respectfully formal manner with any requests for help. In
this case, a one-off request for help about licensing concerns would
probably be mystifying.

> Specifically, who do you have in mind?

The people I have in mind are authors and publishers of Creative
Commons licensed textbooks (and any other math content we might like
to use). These are the people whose points of view we especially need
to understand. In addition, we can also assay the available content
and think further about what the pros and cons of importing it might
be. We should at least ask a few different authors (or assigned
copyright holders) whether they are fixed on CC-By-NC-ND or not. If
they are, I still think we need to think long and hard about whether
what is technically un-free content is something we can support.

http://www.openculture.com/free_textbooks
http://mathoverflow.net/questions/99987/where-to-publish-a-math-textbook-in-creative-commons
http://creativecommons.org/tag/math
http://florida.theorangegrove.org/og/access/hierarchy.do?topic=d37c6ed5-3822-84a6-721c-6d9033a88541&page=1&sort=rank&q=math

[puzio1.excite]

> My point is that they are rather busy people and we should approach
> them in a respectfully formal manner with any requests for help. In
> this case, a one-off request for help about licensing concerns would
> probably be mystifying.

In order to not mystify and confuse, what about providing them with a
document in two parts, where the first part summarizes what has been
going on with PlanetMath recently and our plans for the near future whilst
the second part would enumerate the issues and questions we are
grappling with in connection with our plans and ask them for their opinions
as well as for general suggestions and advice?

> The people I have in mind are authors and publishers of Creative
> Commons licensed textbooks (and any other math content we might like
> to use). These are the people whose points of view we especially need
> to understand.

While it would be interesting to know what textbook authors think on
the topic, I would be much more interested in the stance of people
involved in open access research such as journal editors and
contributors. This is because the issue of authors being concerned
about derived works possibly harming their reputation or taking away
credit are primarily issues which affect researchers as opposed
to expositors.