License Proliferation Dissatisfaction

[Russ Nelson]

     I can't get no satisfaction
     I can't get no satisfaction
     'Cause I try and I try and I try and I try
     I can't get no, I can't get no

Like the Rolling Stones, we can't get no satisfaction with open source
licenses.  We don't have enough, then we have too many, then we don't
have enough.

The large number of open source licenses has been a problem for a long
time.  No need to explain the problem here.  We've been addressing the
problem long before anybody else saw it as a problem, trying to
encourage people to reuse existing licenses.  Most recently we
established a committee to categorize the licenses, which I
co-chaired.  The board has approved the committee's report
(http://opensource.org/lpc) and endorses the categorization
(http://opensource.org/licenses/category).

We knew that this process was going to displease some parties.  When
you set out to recommend some, the others will not be happy.  This is
the case with Larry Rosen, whose licenses, the AFL and OSL, did not
come up as top-ranked licenses.  This is unfortunate, because they are
legally robust licenses.

Larry claims that the license proliferation committee's process was
political, not legal.  We agree!  We took policy issues into account
as well as legal issues.  That is because open source licenses do not
operate in a legal vacuum. They have practical implications and our
intention was to provide practical advice. We did not consider it
practical to recommend licences that are rarely used, however sound
their legal drafting may be.

Unlike Larry, the License Proliferation Committee was not free to
focus only on "legal merit" as a criteria for categorizing open source
licenses.  We respected the communities that have grown up around
these licenses.  We looked at the number of people and projects using
a license, and gave them credit for popularity in spite of the
relative legal merit of the license.

If he can get the numbers up on his license, then we will certainly
consider moving his license(s)s to the "Strong Communities" category.

But given that we expected the process to create unhappiness, the
existance of Larry's unhappiness is not going to persuade us to
change our mind about the results of the committee's deliberation.

--
--my blog is at    http://blog.russnelson.com   | You can do any damn thing
Crynwr sells support for free software  | PGPok | you want, as long as you
521 Pleasant Valley Rd. | [phone]       | don't expect somebody else
Potsdam, NY 13676-3213  |     Sheepdog          | to pick up the pieces.

[David K. Gasaway]

On 22 Apr 2007 at 19:58, Russ Nelson wrote:

> The board has approved the committee's report
> (http://opensource.org/lpc) and endorses the categorization
> (http://opensource.org/licenses/category).

While I respect the committee's efforts and conclusions, please
reconsider the word "redundant".  It is poorly descriptive and even
inflammatory.  "Licenses with similar goals to more popular licenses"
or even "Less popular licenses" would probably go over more smoothly.  
Surely the committee can find language that is appropriately dissuasive
without being offensive.

--
-:-:- David K. Gasaway
-:-:- Email: dave@...
-:-:- Web  : dave.gasaway.org

[Chris DiBona]

I think you might be a little too sensitive. Ranking anything will
make proponents of those licenses ranked lower feel worse. Also,
redundancy and popularity are very different things. There are a lot
of redundant aspects in the licenses in the top tier.

Chris

On 4/22/07, David K. Gasaway <dave@...> wrote:

> On 22 Apr 2007 at 19:58, Russ Nelson wrote:
>
> > The board has approved the committee's report
> > (http://opensource.org/lpc) and endorses the categorization
> > (http://opensource.org/licenses/category).
>
> While I respect the committee's efforts and conclusions, please
> reconsider the word "redundant".  It is poorly descriptive and even
> inflammatory.  "Licenses with similar goals to more popular licenses"
> or even "Less popular licenses" would probably go over more smoothly.
> Surely the committee can find language that is appropriately dissuasive
> without being offensive.
>
> --
> -:-:- David K. Gasaway
> -:-:- Email: dave@...
> -:-:- Web  : dave.gasaway.org
>
>
>

--
Open Source Programs Manager, Google Inc.
Google's Open Source program can be found at http://code.google.com
Personal Weblog: http://dibona.com

[Lawrence Rosen]

Russ Nelson wrote:
> We respected the communities that have grown up around
> these licenses.  We looked at the number of people and projects using
> a license, and gave them credit for popularity in spite of the
> relative legal merit of the license.

The committee process was blatantly political and whatever attempts were
made by the committee to determine popularity were clearly as
methodologically unsound as was its legal analysis.

What is the popularity or size of the community around CDDL? Or was it
merely necessary to keep the representatives of Sun happy on your committee?
Long before CDDL became a license, as of 1/7/05 when I wrote a paper for
distribution on the OSDL website on the topic of license proliferation,
OSL/AFL were respectively #11 and #14 on the Sourceforge.net site. Treated
as a single license (which in principle they are except for one provision),
it would have been #8, ahead of the CPL (which magically remains "popular"
on your list to keep IBM happy). I haven't run the statistics since; did
your committee?

I'm not trying to disparage CDDL or CPL here. I'm just punching holes in
Russ' nonsense.

And regardless of statistics, when did OSI start to run fashion shows
instead of open source education? Remember, I wrote the bylaws and know what
OSI's mission was supposed to be. "Ranking for popularity" was never a part
of it.

/Larry

[Andrew C. Oliver-2]

I think this categorization is helpful.  With that, if License
proliferation is bad and the board wishes to stop it -- why not stop
it?  Meaning *close* the license approval process and declare "Open
Source means software licensed in its entirety under these  licenses"
and declare the OSD to be a statement of values and that there are no
other open source licenses.  Then state that NO licenses can be approved
unless the board is petitioned and at its discretion reopens the license
approval process.  Finally state that the board is not well disposed to
do so unless there is a change in law or legal climate (such as say the
GPL being proved invalid in court).  Make the statement that new
revisions of existing licenses (such as the GPLv3) will still be
considered provided they are "substantially similar" to their
predecessors in terms and spirit.

Russ Nelson wrote:

>      I can't get no satisfaction
>      I can't get no satisfaction
>      'Cause I try and I try and I try and I try
>      I can't get no, I can't get no
>
> Like the Rolling Stones, we can't get no satisfaction with open source
> licenses.  We don't have enough, then we have too many, then we don't
> have enough.
>
> The large number of open source licenses has been a problem for a long
> time.  No need to explain the problem here.  We've been addressing the
> problem long before anybody else saw it as a problem, trying to
> encourage people to reuse existing licenses.  Most recently we
> established a committee to categorize the licenses, which I
> co-chaired.  The board has approved the committee's report
> (http://opensource.org/lpc) and endorses the categorization
> (http://opensource.org/licenses/category).
>
> We knew that this process was going to displease some parties.  When
> you set out to recommend some, the others will not be happy.  This is
> the case with Larry Rosen, whose licenses, the AFL and OSL, did not
> come up as top-ranked licenses.  This is unfortunate, because they are
> legally robust licenses.
>
> Larry claims that the license proliferation committee's process was
> political, not legal.  We agree!  We took policy issues into account
> as well as legal issues.  That is because open source licenses do not
> operate in a legal vacuum. They have practical implications and our
> intention was to provide practical advice. We did not consider it
> practical to recommend licences that are rarely used, however sound
> their legal drafting may be.
>
> Unlike Larry, the License Proliferation Committee was not free to
> focus only on "legal merit" as a criteria for categorizing open source
> licenses.  We respected the communities that have grown up around
> these licenses.  We looked at the number of people and projects using
> a license, and gave them credit for popularity in spite of the
> relative legal merit of the license.
>
> If he can get the numbers up on his license, then we will certainly
> consider moving his license(s)s to the "Strong Communities" category.
>
> But given that we expected the process to create unhappiness, the
> existance of Larry's unhappiness is not going to persuade us to
> change our mind about the results of the committee's deliberation.
>
>  

--
Catch us at Linuxfest Northwest (4/28-4/29) http://linuxfestnorthwest.org
Buni Meldware Communication Suite
Email, Calendaring, ease of configuration/administration
http://buni.org

[Matthew Flaschen]

Andrew C. Oliver wrote:

> Make the statement that new
> revisions of existing licenses (such as the GPLv3) will still be
> considered provided they are "substantially similar" to their
> predecessors in terms and spirit.

If they're so similar, why should we approve them?  The answer is that
there may be necessary (or at least very useful) changes, as (I think)
GPLv3 makes.  But then why should we only allow useful changes if
they're made to an existing license?  That really doesn't make sense,
since a new license is new regardless of whether it updates an approved
license.  So then we should allow useful updates to existing licenses,
or new licenses that serve a useful purpose.  I think that basically
brings us the board's current (unstated) stance.  Now, I agree with you
that not many new licenses will be both OSD-compliant and useful, and I
think the board and list understand that.

Matthew Flaschen

[Eugene Wee]

> We did not consider it
> practical to recommend licences that are rarely used, however sound
> their legal drafting may be.

 From what I see, effectively the "recommendation" is just a placement
at the top category, according to whether the license is "popular and
widely used or with strong communities". I think Larry's gripe is that
"redundant with more popular licenses" is a negative recommendation, due
to the negative connotation.

Why not compromise with a "legally sound but less widely used licences"
category? That way users can more clearly decide if they want a license
whose enforcement and interpretation may have a precedent and be
established in a large community versus a license that is legally robust
on its own but without strong community use.

> If he can get the numbers up on his license, then we will certainly
> consider moving his license(s)s to the "Strong Communities" category.

To be fair, that becomes something of a Catch-22 situation: anyone who
wants to cite the OSI website when recommending the AFL has to explain
why they recommend it when the OSI considers it "redundant". This
imposes a barrier to its adoption, which in turn would make entry to the
"Strong Communities" category more difficult.

Regards,
Eugene Wee

[Andrew C. Oliver-2]

Matthew Flaschen wrote:

> Andrew C. Oliver wrote:
> If they're so similar, why should we approve them?  The answer is that
> there may be necessary (or at least very useful) changes, as (I think)
> GPLv3 makes.  But then why should we only allow useful changes if
> they're made to an existing license?  That really doesn't make sense,
> since a new license is new regardless of whether it updates an approved
> license.  So then we should allow useful updates to existing licenses,
> or new licenses that serve a useful purpose.  I think that basically
> brings us the board's current (unstated) stance.  Now, I agree with you
> that not many new licenses will be both OSD-compliant and useful, and I
> think the board and list understand that.
>
> Matthew Flaschen
>  

Because nothing is ever perfect and as a software developer I'm loathe
to say that there won't be a 1.1 or 2.0 or 3.0 of anything...

Because I keep hoping that the FSF will wake up and realize that A>
deodorant is good and B> Legal documents are best if they don't read
like Ted Kaczynski wrote it on one of his less coherent days.

Because I keep hoping that the ASF will realize that while the ASL 2.0
is mostly an improvement in terms it is a regression in comprehension
and make ASL 3.0 a little shorter and a little more in the vernacular of
the PRIMARY audience which is not the Larry Rosens of the world but the
Joe Developers (like me) of the world.  (I can clarify this point if its
truth is not immediately apparent)

Mostly Because having a fork with the FSF where GPL 2.0 works are open
source but GPL 3.0 works are not just 'cause would be silly and a
headache for us mere software developers.

Why not?

-andy

--
Catch us at Linuxfest Northwest (4/28-4/29) http://linuxfestnorthwest.org
Buni Meldware Communication Suite
Email, Calendaring, ease of configuration/administration
http://buni.org

[Matthew Flaschen]

Andrew C. Oliver wrote:

> Because nothing is ever perfect and as a software developer I'm loathe
> to say that there won't be a 1.1 or 2.0 or 3.0 of anything...

I agree with you here.  But sometimes you need to start over.  /I'm/
loathe to say, "Oh, update all you want, but no new programs."  I agree
that we should allow new versions of licenses, but I also think it would
be wrong to block out entirely new licenses.

Matt Flaschen

[Andrew C. Oliver-2]

Matthew Flaschen wrote:
> Andrew C. Oliver wrote:
> I agree with you here.  But sometimes you need to start over.  /I'm/
> loathe to say, "Oh, update all you want, but no new programs."  I agree
> that we should allow new versions of licenses, but I also think it would
> be wrong to block out entirely new licenses.
>
> Matt Flaschen
>  
I'm not saying that at all.  Just that closing the door and say "these
are it" and make the process "door is open or not".  Meaning NO new
licenses (except revisions) can be considered unless the "list of open
source licenses has been reopened by the board for revision" and it is
an "all or not" to "reopining" for revision (meaning any license can be
submitted for approval if the list is reopened but NONE can be until
that time).  The idea is to formally close the door until there is a
compelling enough reason (presumably legal event) to reopen it....and no
one REALLY wants to open the door because I came up with a new and
interesting reword of the BSD license just 'cause right.  Close the
door, make the bar high to reopen it.

-Andy

--
Catch us at Linuxfest Northwest (4/28-4/29) http://linuxfestnorthwest.org
Buni Meldware Communication Suite
Email, Calendaring, ease of configuration/administration
http://buni.org

[Russ Nelson]

Lawrence Rosen writes:
 > I'm not trying to disparage CDDL or CPL here. I'm just punching holes in
 > Russ' nonsense.

I hear you saying that you're unhappy.  I understand that you're
unhappy.  I'm sorry that you're unhappy.  But your unhappiness isn't
going to change the outcome.  Your choice at this point is whether you
choose to be happy or unhappy.  May I recommend the choice to be happy?

--
--my blog is at    http://blog.russnelson.com   | You can do any damn thing
Crynwr sells support for free software  | PGPok | you want, as long as you
521 Pleasant Valley Rd. | [phone]       | don't expect somebody else
Potsdam, NY 13676-3213  |     Sheepdog          | to pick up the pieces.

[Matthew Flaschen]

Andrew C. Oliver wrote:

> The idea is to formally close the door until there is a
> compelling enough reason (presumably legal event) to reopen it

The problem is, that's not how the law works as far as I can tell.
Trends are more common than single decisive events.  When you choose to
deal with the trend is a matter of opinion.  Anyway, I'm not sure I see
a problem with how things are being done now.  There may be a drain on
the board's time, but licenses aren't getting approved willy-nilly.

Matthew Flaschen

[Lawrence Rosen]

Chris DiBona wrote:
> I think you might be a little too sensitive. Ranking anything will
> make proponents of those licenses ranked lower feel worse. Also,
> redundancy and popularity are very different things. There are a lot
> of redundant aspects in the licenses in the top tier.

Of course you're right about the differences between redundancy and
popularity, and about the redundancy of other licenses in the "top tier".
But the categorization of the licenses on OSI's website is wrong on both
counts. Not only is OSL 3.0 far more popular in terms of actual number of
open source licensors than CDDL (which made the cut because Sun's lawyer was
on the license proliferation committee), but OSL 3.0 is also unique (at
least among "popular" licenses) in containing an ASP provision.

Categorizing certain open source licenses on OSI's website as "redundant" or
"popular" is misleading and incorrect.

This is not about sensitivity or happiness. I'm disappointed that an
organization that I helped for so long to review and approve licenses has so
little apparent legal understanding of them and has such a political view of
what to do about the problem of license proliferation.

/Larry

> -----Original Message-----
> From: Chris DiBona [mailto:cdibona@...]
> Sent: Sunday, April 22, 2007 6:33 PM
> To: dave@...
> Cc: license-discuss@...; license-proliferation-
> discuss@...
> Subject: Re: License Proliferation Dissatisfaction
>
> I think you might be a little too sensitive. Ranking anything will
> make proponents of those licenses ranked lower feel worse. Also,
> redundancy and popularity are very different things. There are a lot
> of redundant aspects in the licenses in the top tier.
>
> Chris
>
> On 4/22/07, David K. Gasaway <dave@...> wrote:
> > On 22 Apr 2007 at 19:58, Russ Nelson wrote:
> >
> > > The board has approved the committee's report
> > > (http://opensource.org/lpc) and endorses the categorization
> > > (http://opensource.org/licenses/category).
> >
> > While I respect the committee's efforts and conclusions, please
> > reconsider the word "redundant".  It is poorly descriptive and even
> > inflammatory.  "Licenses with similar goals to more popular licenses"
> > or even "Less popular licenses" would probably go over more smoothly.
> > Surely the committee can find language that is appropriately dissuasive
> > without being offensive.
> >
> > --
> > -:-:- David K. Gasaway
> > -:-:- Email: dave@...
> > -:-:- Web  : dave.gasaway.org
> >
> >
> >
>
>
> --
> Open Source Programs Manager, Google Inc.
> Google's Open Source program can be found at http://code.google.com
> Personal Weblog: http://dibona.com

[Forrest J. Cavalier III]

Russ Nelson wrote:
> Lawrence Rosen writes:
>  > I'm not trying to disparage CDDL or CPL here. I'm just punching holes in
>  > Russ' nonsense.
>
> I hear you saying that you're unhappy.  I understand that you're
> unhappy.  I'm sorry that you're unhappy.  But your unhappiness isn't
> going to change the outcome.  Your choice at this point is whether you
> choose to be happy or unhappy.  May I recommend the choice to be happy?
>

What Russ apparently did not hear was Larry asking the committee to reveal
what concrete method was used to determine popularity.

In this reply I hear "La, La, La, I'm not listening."

What concrete method was used by the committee to determine popularity?

[Russ Nelson]

Forrest J. Cavalier III writes:
 > In this reply I hear "La, La, La, I'm not listening."
 >
 > What concrete method was used by the committee to determine popularity?

No, what you're heading is "I don't care if you don't like it.  I
expected you to not like it."  And I by "you" I don't mean that we set
out to screw Larry Rosen.  I mean that *any* procedure, *any* number
of groups, *any* categorization was guaranteed to produce
dissatisfaction.  I feel sorry for Larry, but I'm not going to defend
the decision.  It's reasonable.  It may not be perfect.  It's better
than what we had before.  The best is the enemy of the good.  Deal
with improvement and changes like a grown-up (not that many grown-ups
deal with change well -- but we're *supposed* to).

--
--my blog is at    http://blog.russnelson.com   | You can do any damn thing
Crynwr sells support for free software  | PGPok | you want, as long as you
521 Pleasant Valley Rd. | [phone]       | don't expect somebody else
Potsdam, NY 13676-3213  |     Sheepdog          | to pick up the pieces.

[Chris DiBona]

I agree that cddl shouldn't have made the cut. But I wasn't part of
the committee. We made similar but not exact decisions when we created
the project hosting service here (yay #2!) And the only license we
actually have people clamor for is the epl. But even that is too
unpopular to warrant inclusion.

Chris


On 4/23/07, Forrest J. Cavalier III <mibsoft@...> wrote:

> Russ Nelson wrote:
> > Lawrence Rosen writes:
> >  > I'm not trying to disparage CDDL or CPL here. I'm just punching holes
> in
> >  > Russ' nonsense.
> >
> > I hear you saying that you're unhappy.  I understand that you're
> > unhappy.  I'm sorry that you're unhappy.  But your unhappiness isn't
> > going to change the outcome.  Your choice at this point is whether you
> > choose to be happy or unhappy.  May I recommend the choice to be happy?
> >
>
> What Russ apparently did not hear was Larry asking the committee to reveal
> what concrete method was used to determine popularity.
>
> In this reply I hear "La, La, La, I'm not listening."
>
> What concrete method was used by the committee to determine popularity?
>
>
>
>
>
>
>
>

--
Open Source Programs Manager, Google Inc.
Google's Open Source program can be found at http://code.google.com
Personal Weblog: http://dibona.com

[Stephane Dalmas]

> Lawrence Rosen writes:
>  > I'm not trying to disparage CDDL or CPL here. I'm just punching holes in
>  > Russ' nonsense.
>
> I hear you saying that you're unhappy.  I understand that you're
> unhappy.  I'm sorry that you're unhappy.  But your unhappiness isn't
> going to change the outcome.  Your choice at this point is whether you
> choose to be happy or unhappy.  May I recommend the choice to be happy?
>
>  

For what it's worth... Larry Rosen is not the only one to be "unhappy"
or "surprised" by this work on license proliferation (at least from a
legal point of view and especially from someone on the other side of the
Atlantic Ocean). I think he raised good points that you did not answer.
Moreover, I'm also suprised by the tone of your response that sounds not
very "neutral" to me. I understand that you know Larry probably well but
from the outside it is not the kind of answer one expects from a member
of the OSI board on a public mailing list.

    Stéphane.

[Brian Behlendorf-2]

On Mon, 23 Apr 2007, Russ Nelson wrote:

> Forrest J. Cavalier III writes:
> > In this reply I hear "La, La, La, I'm not listening."
> >
> > What concrete method was used by the committee to determine popularity?
>
> No, what you're heading is "I don't care if you don't like it.  I
> expected you to not like it."  And I by "you" I don't mean that we set
> out to screw Larry Rosen.  I mean that *any* procedure, *any* number
> of groups, *any* categorization was guaranteed to produce
> dissatisfaction.

It's pretty disenheartening to hear an OSI board member claim that they
didn't even try for objectivity in decision making, based on a premise
that objectivity is a total myth.  I certainly think there are approaches
that would have built more confidence in the quality of that list.  I find
myself agreeing with Larry that the phrases "redundant" and "less popular"
are not truthful when describing the list of disrecommended licenses.  Now
that OSI has "solved" the license proliferation problem, are any board
members interested in working on the credibility problem?

> I feel sorry for Larry, but I'm not going to defend the decision.  It's
> reasonable.  It may not be perfect.  It's better than what we had
> before.  The best is the enemy of the good.  Deal with improvement and
> changes like a grown-up (not that many grown-ups deal with change well
> -- but we're *supposed* to).

Isn't the position that new licenses are disrecommended (by definition,
licenses that start out as "less popular"), also a position against
improvement and change?

  Brian

[Forrest J. Cavalier III]

Russ Nelson wrote:

> Forrest J. Cavalier III writes:
>  > In this reply I hear "La, La, La, I'm not listening."
>  >
>  > What concrete method was used by the committee to determine popularity?
>
> No, what you're heading is "I don't care if you don't like it.  I
> expected you to not like it."  And I by "you" I don't mean that we set
> out to screw Larry Rosen.  I mean that *any* procedure, *any* number
> of groups, *any* categorization was guaranteed to produce
> dissatisfaction.  I feel sorry for Larry, but I'm not going to defend
> the decision.  It's reasonable.  It may not be perfect.  It's better
> than what we had before.  The best is the enemy of the good.  Deal
> with improvement and changes like a grown-up (not that many grown-ups
> deal with change well -- but we're *supposed* to).
>

In the 21st century, quality is recognized by examining the process, not the result.

You say Larry didn't like the result. You keep pointing to the result and
saying it is "good and reasonable."  But...

I asked about the process.    What concrete method was used by the committee
to determine popularity?

How is that "not being grown-up?"  Is asking about the process second-guessing
the committee?

I expected a quick URL or something, since it would be in the decision-making
minutes of the committee.  But you are avoiding the question so much, that it
seems like you are hiding something.  Maybe it is a lack of reproducible process
that you are hiding?

[David K. Gasaway]

On 22 Apr 2007 at 18:33, Chris DiBona wrote:

> I think you might be a little too sensitive.

Not at all - I'm merely an interested bystander.  I don't have any
complaint about the categories, per se (I'll leave that to others more
qualified), but the word "redundant" is unnecessarily negative.

> There are a lot
> of redundant aspects in the licenses in the top tier.

And plenty of unique aspects in licenses in the "redundant" tier.  See
the problem?  Again, I'm not asking that the category be removed, but
let's be honest about what it is.

--
-:-:- David K. Gasaway
-:-:- Email: dave@...
-:-:- Web  : dave.gasaway.org