For Approval: Microsoft Permissive License

[Michael R. Bernstein]

On Thu, 2007-08-16 at 22:30 -0700, Bill Hilf wrote:

> Chris:
> [snip]
> The other license submission question you raise is about additional
> licenses in an era when the OSI is trying to reduce license
> proliferation.  There are already several hundred community projects
> that use these licenses, including over 150 Microsoft projects; while
> these are two more licenses they represent a reasonably large set of
> existing code, the authors and users of which would benefit from
> having the licenses assessed as Open Source.
>
> > -----Original Message-----
> > From: Chris DiBona [mailto:cdibona@...]
> >
> > [snip]
> >
> > Finally, why should yet another set of minority, vanity licenses be
> > approved by an OSI that has been attempting to deter copycat
> > licenses and reduce license proliferation?

Bill,

While I am sure that having the MS-PL and MS-CL licenses approved as OSD
compliant would be of enormous benefit to the projects in question and
indeed Microsoft itself, I don't think that was the actual thrust of the
question.

How does approving these licenses benefit the Open Source community as a
whole, especially given that these licenses seem deliberately crafted to
dig a moat separating the authors and code of the projects in question
from any other open source license or code cooties, rather than building
a bridge?

This is a serious question, even if the OSD-compliance of the licenses
does not hinge on it.

- Michael R. Bernstein
  michaelbernstein.com

[Nils Labugt]

lør, 18.08.2007 kl. 01.34 -0700, skrev Michael R. Bernstein:

> As far as I can tell, the MS-PL and MS-CL licenses aren't even
> compatible with each other, in either direction.
>
> When we talk about 'license proliferation' as a problem, it's important
> to note that aside from the annoying confusion that the number of
> licenses per-se engenders, the real problems are side effects such as
> immiscible codebases.
>
> This Highlander[1] requirement deliberately and explicitly cranks the
> imiscibility up to eleven, for no good reason I can see.
>
> In the spirit of license-discuss, where it is not uncommon to suggest
> general improvements to submitted licenses even if they are not strictly
> required for OSD-compliance, I want to know if Microsoft is willing to
> improve these licenses by changing 'you may do so only under this
> license' to 'you must do so under this license' or similar phrasing in
> order to leave the door open for compatible licenses and dual-licensing.

Some of us wants to license our code under a permissive license, and
also wants our code to *remain* under a permissive license. Isn't that a
legitimate wish?

There seems to be an attitude of many around here that one should go to
great lengths to satisfy those who wants strong "copyleft" terms, while
a desire to prevent others from adding new restrictive terms is frown
upon. And this license permits linking to code under different licenses,
doesn't it? Unlike GPLv3.


Nils Labugt

[Donovan Hawkins]

On Sat, 18 Aug 2007, Nils Labugt wrote:

> Some of us wants to license our code under a permissive license, and
> also wants our code to *remain* under a permissive license. Isn't that a
> legitimate wish?

That's an inconsistant wish. What you are saying is that you want code
licensed under the license you choose to remain licensed under the license
you choose. That's hardly permissive. While I appreciate that there is a
mild sense of "compatibility" with other licenses (via linking or keeping
different source files under different licenses), as a developer I would
not wish to have licensing issues dictate the design of my software and
would simply avoid your software unless I'm also using your license. That
makes the license spiritually closer to copyleft licenses like GPL than to
permissive licenses like BSDL (again from the perspective of a developer).

IMO, there is no fundamental difference between MS-PL and GPL except in
who they choose to exclude from their code-sharing club: GPL excludes
closed-source projects, MS-PL excludes GPL projects. MS-PL doesn't
have to explicitly prevent linking or keeping different source files under
different licenses because it acquires this level of exclusion from GPL by
virtue of GPL's own clauses interacting with the MS-PL. This is hardly an
accident.

Either you want people to do what THEY want with your code, or you want
people to do what YOU want with your code. The former is permissive, while
the latter is restrictive (though not necessarily copyleft). I have no
objection to calling MS-PL "open source" since it clearly is. I do have
objection to calling it "permissive" since it clearly is not.

It's unfortunate enough that "free" now means "free to do what the the GPL
says you can do". Let's not also have "permissive" mean "permitted to do
what the MPL/MS-PL says you can do".


> And this license permits linking to code under different licenses,
> doesn't it? Unlike GPLv3.

Being more permissive than GPL is a far cry from actually being
permissive, given how restrictive GPL is.

---------------------------------------------------------------------------
Donovan Hawkins, PhD                 "The study of physics will always be
Software Engineer                     safer than biology, for while the
hawkins@...                   hazards of physics drop off as 1/r^2,
http://www.cephira.com                biological ones grow exponentially."
---------------------------------------------------------------------------

[Michael R. Bernstein]

On Sat, 2007-08-18 at 08:25 -0700, Donovan Hawkins wrote:
> On Sat, 18 Aug 2007, Nils Labugt wrote:
>
> IMO, there is no fundamental difference between MS-PL and GPL except in
> who they choose to exclude from their code-sharing club: GPL excludes
> closed-source projects, MS-PL excludes GPL projects. MS-PL doesn't
> have to explicitly prevent linking or keeping different source files under
> different licenses because it acquires this level of exclusion from GPL by
> virtue of GPL's own clauses interacting with the MS-PL. This is hardly an
> accident.

From my POV, the fundamental difference here is *how* they choose to
exclude. Other pairwise license incompatibilities (including those that
involve the GPL) arise out of the side-effects of incompatible terms (in
the GPL's case, 'no other restrictions'), which at least leaves the door
open for pairwise combinations that are compatible.

In contrast, the MS-PL and MS-CL, by virtue of explicitly excluding
other licenses wholesale (including each other!), seem to have license
incompatibility as a *design goal* rather than a side-effect, and it is
*that* which I am very uncomfortable with. Not to mention eliminating
the possibility of dual licensing.

So.

Deliberately excluding all other licenses: Bug Or Feature?

I say, 'Bug'.

- Michael R. Bernstein
  michaelbernstein.com

[Rick Moen]

Quoting Nils Labugt (elabu@...):

> Some of us wants to license our code under a permissive license, and
> also wants our code to *remain* under a permissive license.

Certainly, it's OK to want.  (You can _want_ noonday midnights, too.)

However, it's inherent in the concept of a permissive licence that it's,
well, permissive:  Third-party recipients are expressly permitted to
create derivative works under any licensing terms whatsoever, from
permissive ones to entirely proprietary ones.  

I respect that tradition -- and use that type of licensing myself on
many occasions, by preference.  I respect its practitioners even more
when they grasp the implications of what they're doing.  ;->

> There seems to be an attitude of many around here that one should go to
> great lengths to satisfy those who wants strong "copyleft" terms, while
> a desire to prevent others from adding new restrictive terms is frown
> upon.

Er, no.  One should go to the exact lengths necessary to comply with any
software licence of any type.  Sometimes, the most practical way to
comply turns out to be to eschew that particular software completely,
and say "In _that_ case, no thanks, we'll either do without, or write or
commission our own, under terms more to our liking."

> And this license permits linking to code under different licenses,
> doesn't it? Unlike GPLv3.

(GPLv3 has no power to regulate "linking", as such -- a misconception
carried forward from FSF and related materials about GPLv2.)

Certainly, if one wants the greatest possible range of software licence
compatibility, and that concern outweighs others, one _should_ use a
permissive licence such as new-BSD or MIT / X11.  However, people have
diverse objectives, in their licensing practices.

OSI's role is merely to certify the licences that meet OSD criteria, and
promote the concept of open source in general.

[Nils Labugt]

lør, 18.08.2007 kl. 10.43 -0700, skrev Michael R. Bernstein:

> From my POV, the fundamental difference here is *how* they choose to
> exclude. Other pairwise license incompatibilities (including those that
> involve the GPL) arise out of the side-effects of incompatible terms (in
> the GPL's case, 'no other restrictions'), which at least leaves the door
> open for pairwise combinations that are compatible.
>
> In contrast, the MS-PL and MS-CL, by virtue of explicitly excluding
> other licenses wholesale (including each other!), seem to have license
> incompatibility as a *design goal* rather than a side-effect, and it is
> *that* which I am very uncomfortable with.

For comparison, first from MS-PL:

"(D) If you distribute any portion of the software in source code form,
you may do so only under this license by including a complete copy of
this license with your distribution."


>From CPLv1:


"When the Program is made available in source code form:

a) it must be made available under this Agreement; and
       
b) a copy of this Agreement must be included with each copy of the
Program."


>From GPLv3:


"5. ...
b) The work must carry prominent notices stating that it is
   released under this License and any conditions added under section
   7. [...]

c) You must license the entire work, as a whole, under this
   License to anyone who comes into possession of a copy.  This
   License will therefore apply, along with any applicable section 7
   additional terms, to the whole of the work, and all its parts,
   regardless of how they are packaged.  This License gives no
   permission to license the work in any other way, but it does not
   invalidate such permission if you have separately received it."

"4. ...and give all recipients a copy of this License along with the
Program."


Comparing these excerpts, I don't see why it is a "design goal" in the
case of the MS-PL but only a side-effect in the other two cases.

> Not to mention eliminating
> the possibility of dual licensing.


If the license file and every source file states in no uncertain terms
that the program is dual licensed, then I would be surprised if a judge
lets the language "only under this license" in the MS-PL override that.



Nils Labugt

[John Cowan]

Michael R. Bernstein scripsit:

> I think you're missing the point that both of these licenses are
> incompatible with *any* other license, no matter how permissive. A file
> under MS-PL can't even incorporate BSD-licensed code. That takes
> *special* effort.

I believe you have misread MS-PL 3D.  It says that "this software", which
means "the accompanying software", the original code you got along with
the license, must be distributed under the MS-PL and accompanied with it.

It does *not* say that any derivative work (the creation of which is
explicitly permitted by 2A) must be licensed under the MS-PL.  Indeed,
as long as you follow the conditions and limitations in 3A-3E, you
can apparently license your derivative work under any license.

Of course, your changes to the work must rise to the level of being
a derivative work as opposed to a mere transcription or trivial change.

> As far as I can tell, the MS-PL and MS-CL licenses aren't even
> compatible with each other, in either direction.

I believe this to be the product of the same misreading.  The MS-CL does
create a closed commons, but the MS-PL does not.

--
John Cowan   cowan@...  http://www.ccil.org/~cowan
Most languages are dramatically underdescribed, and at least one is
dramatically overdescribed.  Still other languages are simultaneously
overdescribed and underdescribed.  Welsh pertains to the third category.
        --Alan King

[John Cowan]

Donovan Hawkins scripsit:

> IMO, there is no fundamental difference between MS-PL and GPL except in
> who they choose to exclude from their code-sharing club: GPL excludes
> closed-source projects,

And many open-source projects too.

> MS-PL excludes GPL projects. MS-PL doesn't have to explicitly prevent
> linking or keeping different source files under different licenses
> because it acquires this level of exclusion from GPL by virtue of
> GPL's own clauses interacting with the MS-PL.

Which is to say that the GPL excludes the MS-PL, not vice versa.
In particular, 2B is inconsistent with the GPLv2; I don't know whether
it's incompatible with the GPLv3 or not -- I suspect not.  Anyhow, if
you keep me out of your club, you can't also claim that I'm an elitist
because I can't be found there.

> Either you want people to do what THEY want with your code, or you
> want people to do what YOU want with your code.

The situation is by no means so black and white.  Even the 3-clause
BSD license imposes *some* restrictions on (re)users; in particular,
they can't change the attributions.  If you want "people to do what
THEY want", you need to dedicate your code to the public domain, and
then you have to deal with me and Larry Rosen, who don't believe that
you can actually do so.

> It's unfortunate enough that "free" now means "free to do what the
> the GPL says you can do".

The theory of GPL freedom is that it preserves the freedom of users, not
necessarily the freedom of developers; developers get a lot of freedom to
do what they want with the code, but by no means absolute freedom.

--
What is the sound of Perl?  Is it not the       John Cowan
sound of a [Ww]all that people have stopped     cowan@...
banging their head against?  --Larry            http://www.ccil.org/~cowan

[Chris Travers]

Matthew Flaschen wrote:

>
>
> The phrase "you may do so only under this license by including a
> complete copy of this license with your distribution." is in both
> licenses.  Jon confirmed this is the submission email by writing, "The
> MS-PL license explicitly prohibits relicensing of the original licensed
> code under a different license, regardless of whether the original code
> is redistributed in whole, in part or as part of a different piece of
> software.".  In my interpretation (IANAL), this clause means MS-PL code
> can't be used in a GPL program.
>  

Hmmmm.... You may be right....  It may not be able to be used safely in

a GPLd program in all cases.  IANAL, of course.

AFAICS, the GPL does not mandate relicensing any code taken in from
other contributors under other licenses as long as derivatives of the
GPL itself either as the work as a whole or of derivative components.  
In other words, if I include 3 files from BSDL libraries, I am not
required to try to copyleft those files even if I make changes to them
which are not derivative of GPLd code.  The redistribution requirements
on these licenses (in terms of the patent termination clause) conflict
with the GPL v2, but not the GPL v3.

Therefore I would think one could not use code under these licenses in
GPL v2 projects because one cannot guarantee downstream patent licenses
to the extent required by that license.  One could however use the code
in GPL v3 projects provided that the MS-CL or MS-PL code was never
modified in ways derivative of GPL works that the licensor didn't have
permission to relicense.  In short, I don't see how a GPL v3 application
requiring a library under either of these licenses would be a problem
unless those libraries were modified to depend on GPL v3 code.  
Certainly mixing code would also be bad.

In short:
1)  Linking GPL v3 code to MS-CL/MS-PL code:  Seems OK
2)  Copying GPL v3 code to MS-CL/MS-PL app or vice versa:  Seems not OK.
3)  Mixing the two in a file:  Seems not OK
4)  Writing MS-CL/MS-PL code that depends on GPL v3 libraries:  Seems
not OK.

Does this help?
Best Wishes,
Chris Travers

[chris.vcf]

---

begin:vcard
fn:Chris Travers
n:Travers;Chris
email;internet:chris@...
tel;work:509-888-0220
tel;cell:509-630-7794
x-mozilla-html:FALSE
version:2.1
end:vcard

---

[Michael R. Bernstein]

On Sat, 2007-08-18 at 21:45 -0400, John Cowan wrote:

> Michael R. Bernstein scripsit:
>
> > I think you're missing the point that both of these licenses are
> > incompatible with *any* other license, no matter how permissive. A file
> > under MS-PL can't even incorporate BSD-licensed code. That takes
> > *special* effort.
>
> I believe you have misread MS-PL 3D.  It says that "this software", which
> means "the accompanying software", the original code you got along with
> the license, must be distributed under the MS-PL and accompanied with it.

Hmm. So this sentence:

        (D) If you distribute any portion of the software in source code
        form, you may do so only under this license by including a
        complete copy of this license with your distribution.

Really should be:

        (D) If you distribute any portion of the software in source code
        form, you must distribute under this license by including a
        complete copy of this license with your distribution.

Or some similar less ambiguous construction.

> It does *not* say that any derivative work (the creation of which is
> explicitly permitted by 2A) must be licensed under the MS-PL.  Indeed,
> as long as you follow the conditions and limitations in 3A-3E, you
> can apparently license your derivative work under any license.

See, I read 3D as an exception or condition to 2A, effectively altering
the meaning of 2A to 'any derivative works except in source form'.

> Of course, your changes to the work must rise to the level of being
> a derivative work as opposed to a mere transcription or trivial change.
>
> > As far as I can tell, the MS-PL and MS-CL licenses aren't even
> > compatible with each other, in either direction.
>
> I believe this to be the product of the same misreading.  The MS-CL does
> create a closed commons, but the MS-PL does not.

John, thank you for your analysis. I'd like some kind of statement from
the folks at Microsoft that your interpretation is in line with their
intent, and that 'only under this license' does not mean to exclude
distribution of derivative works in source form under other licenses.

I'd be even happier if the licenses were modified to make this
unambiguous.

- Michael R. Bernstein
  michaelbernstein.com

[Brian Behlendorf-2]

On Sat, 18 Aug 2007, Rick Moen wrote:
> OSI's role is merely to certify the licences that meet OSD criteria, and
> promote the concept of open source in general.

The OSI board's anti-proliferation efforts appear to take them one step
beyond certification though.  It would seem to be that otherwise compliant
licenses could be rejected if they simply duplicate the terms or purpose
of an existing license.  I don't know yet if there has been an explicit
rejection of a license up for certification, so I don't know if we've yet
established how different a new license needs to be.  I would guess that a
license that copied the Apache license and replaced all instances of
Apache with some other abstract word would be rejected, no matter what the
compatibility matrix looked like.  How about a license that had exactly
the same requirements as Apache, but restated them in a completely
different way?  From there, what's the *smallest* difference in licensing
terms that would be worth adding yet another license?

Seems as though while MS-PL is not copyleft, MS-CL is, and thus that
"smallest difference" better be pretty big to offset the potential cost of
two universes of immiscible code, MS-CL-licensed and GPL-licensed.

  Brian

[John Cowan]

Michael R. Bernstein scripsit:

> Hmm. So this sentence:
>
>         (D) If you distribute any portion of the software in source code
>         form, you may do so only under this license by including a
>         complete copy of this license with your distribution.
>
> Really should be:
>
>         (D) If you distribute any portion of the software in source code
>         form, you must distribute under this license by including a
>         complete copy of this license with your distribution.
>
> Or some similar less ambiguous construction.

I don't see any ambiguity:  "do so" means "distribute", and "may ... only"
means "must".

> See, I read 3D as an exception or condition to 2A, effectively altering
> the meaning of 2A to 'any derivative works except in source form'.

That would be plausible if 3D mentioned derivative works, but it doesn't.
It speaks of the original software (or parts of it) only.

--
John Cowan  cowan@...  http://www.ccil.org/~cowan
Does anybody want any flotsam? / I've gotsam.
Does anybody want any jetsam? / I can getsam.
        --Ogden Nash, No Doctors Today, Thank You

[Donovan Hawkins]

On Sat, 18 Aug 2007, John Cowan wrote:

> Which is to say that the GPL excludes the MS-PL, not vice versa.
> In particular, 2B is inconsistent with the GPLv2; I don't know whether
> it's incompatible with the GPLv3 or not -- I suspect not.  Anyhow, if
> you keep me out of your club, you can't also claim that I'm an elitist
> because I can't be found there.

That is the mechanism, but since the GPL came first I think it is safe to
say that the MS-PL is intentionally incompatible. The fact that the GPL
makes incompatibility easy to achieve is not the issue. In this case the
club has a rule against wearing hats and you put one on solely to get
yourself kicked out. You both made your choice not to associate with each
other and that's fine, I never said MS-PL couldn't do that. I said it's
not permissive and should not be called permissive.


> Even the 3-clause
> BSD license imposes *some* restrictions on (re)users; in particular,
> they can't change the attributions.  If you want "people to do what
> THEY want", you need to dedicate your code to the public domain, and
> then you have to deal with me and Larry Rosen, who don't believe that
> you can actually do so.

The "restrictions" on the new BSDL are of great importance to lawyers but
of minimal significance to ordinary developers and users. I would argue
that they are nothing more than common-sense courtesy not to cause harm to
the person who freely gave you their software, but sadly not everyone has
common sense or courtesy so we have to be explicit.

I would be willing to wager that few people have ever found themselves
unable to use a piece of (3-clause) BSDL software because of a restriction
in the license that they were unable to comply with. The same is not true
for the GPL nor will it be true for the MS-PL. That is the line I am
saying represents the difference between permissive and non-permissive.

Incidentally, why is public domain not an option? I joined this list
recently and haven't heard your discussions in the past on that subject.


> The theory of GPL freedom is that it preserves the freedom of users, not
> necessarily the freedom of developers; developers get a lot of freedom to
> do what they want with the code, but by no means absolute freedom.

There's a vast difference between GPL freedom and "absolute freedom"...one
that BSDL manages to fit in quite nicely I might add. I'm not attacking
the GPL and I don't necessarily disagree with copyleft, but I'd have
preferred to see "free" reserved for the most free of the open-source
licenses (permissive licenses like BSDL) rather than one of the more
restrictive open-source licenses.

If restrictions like those in MS-PL still count as a permissive license,
could we at least reserve a term for the licenses that BSDL that don't
tell me, the developer, how I'm supposed to license my derivative code?
The good terms are getting used up fast, and it would be a shame if the
category of license that gives the most rights to the community gets
shafted because groups like Microsoft and the FSF want to label their
licenses with terms like "permissive" and "free" in order to sound better.

What word means more free than "free" and more permissive than
"permissive"?

---------------------------------------------------------------------------
Donovan Hawkins, PhD                 "The study of physics will always be
Software Engineer                     safer than biology, for while the
hawkins@...                   hazards of physics drop off as 1/r^2,
http://www.cephira.com                biological ones grow exponentially."
---------------------------------------------------------------------------

[Chris Travers]

>
> How does approving these licenses benefit the Open Source community as a
> whole, especially given that these licenses seem deliberately crafted to
> dig a moat separating the authors and code of the projects in question
> from any other open source license or code cooties, rather than building
> a bridge?
>  
Although I am not Bill nor do I work for Microsoft (nor are they even an
active customer of mine at the moment), I figure I would mention that I
have contributed some works under the MS-PL.  While these works are
technical documentation instead of software, and so may be outside the
bounds of this discussion, I do have some experience with this license
(even if I still choose GPL v2 for *most* work today).

To my knowledge these patent defense clauses were started by IBM in the
IBM Public License Version 1, and were actually initially extremely
controversial.  The argument was that if you *use* the software, I can
incorporate your patents into the software and if you sue me, you lose
your rights to *use* the software because other patent licenses are
revoked.  Some lawyers suggested that this amounted to a license to
steal.  IANAL myself however.

I don't think the question of whether Microsoft incorporated these terms
because they are incompatible with the GPL v2 (which requires that you
*guarantee* patent licenses downstream possibly through an action
separate from the license, and there is no procedure for revoking them
if patent suits are filed) is the issue.  I think the discussion should
be limited to the 4 corners of the license as much as possible aside
possibly from concerns over proliferation of licenses and even there I
think we are better to ask people to reconsider than to reject licenses
otherwise.  The sole exception might be if two licenses were merely
copies of eachother with names changed.  Otherwise, I don't think that
we are qualified to evaluate whatever specific legal concerns may have
lead to different licenses, and whether these are legitimate or not.

Whether Microsoft intended these to be incompatible with the GPL v2 and
compatible with the Apache License 2.0 would not be an issue except in
some peoples minds.  The fact is, I don't read anything in these
licenses which preclude being included in GPL v3 "corresponding source"
or work as a whole.  Certainly copying and pasting of code would be
problematic, and there are other cases where one might run into
problems, but incompatibility is not complete with either license and is
far less with the GPL v3.
> This is a serious question, even if the OSD-compliance of the licenses
> does not hinge on it.
>  
I think that upholding a consistent standard ultimately benefits OSI.  I
think that we need to do so.  Thus far, I haven't heard any argument
*against* approval which doesn't come down to questioning Microsoft's
intentions.  As I have outlined elsewhere, including this is a dangerous
mistake for any future license submissions because it opens up the
possibility of questioning anyone else's intentions too, seriously
undermines our ability to have a credible presence in the community, and
otherwise undermines our work.


Best Wishes,
Chris Travers

[chris.vcf]

---

begin:vcard
fn:Chris Travers
n:Travers;Chris
email;internet:chris@...
tel;work:509-888-0220
tel;cell:509-630-7794
x-mozilla-html:FALSE
version:2.1
end:vcard

---

[Rick Moen]

Quoting Donovan Hawkins (hawkins@...):

> Incidentally, why is public domain not an option? I joined this list
> recently and haven't heard your discussions in the past on that subject.

Some writings on that subject:
"Public Domain" on http://linuxmafia.com/kb/Licensing_and_Law/

--
Cheers,           find / -user your -name base -print | xargs chown us:us
Rick Moen
rick@...

[David Woolley (E.L)]

Donovan Hawkins wrote:

> Incidentally, why is public domain not an option? I joined this list
> recently and haven't heard your discussions in the past on that subject.

IANAL, but my understanding is:

1) In most countries the only way of getting into the public domain is
by waiting until the copyright expires (death plus 70 years, typically,
for an author who is a natural person); I think there is some
controversy as to whether anyone except the US Federal government can
force something into the public domain, even in the USA;

2) Abandoning copyright doesn't mean that you abandon legal
responsibility; having a licence means that you can attempt to get the
recipient to waive the right to sue you in return for receiving the
permissions given in the licence.

Incidentally the BSD attribution requirements probably come under "moral
rights", which I don't believe one can even assign in the EU.

--
David Woolley
Emails are not formal business letters, whatever businesses may want.
RFC1855 says there should be an address here, but, in a world of spam,
that is no longer good advice, as archive address hiding may not work.

[Philip Hunt-3]

On 8/19/07, Chris Travers <chris@...> wrote:
>
> I don't think the question of whether Microsoft incorporated these terms
> because they are incompatible with the GPL v2 (which requires that you
> *guarantee* patent licenses downstream possibly through an action
> separate from the license, and there is no procedure for revoking them
> if patent suits are filed) is the issue.

I agree, I don't think it's the issue either. However I do think some
similar questions are valid.

Consider someone who is proposing a new open source license. For their
proposal to make sense, they should consider other common open source
licenses. They should do this for at least two reasons: (1) to see if their
proposal adds anything new and is not too similar to an existing license,
and (2) to consider whether they intend for their proposed license to be
compatible with other common licenses.

They should do the 2nd not just for the reason that compatibility is,
other things being equal, a desirable goal, but also so that they can be
clear about what their license intends. For example, a proponent might
intend their license to be GPL-compatible, but it is not; or they might
intend it to be incompatible, but it is.

Finding out what the proponent intends of their license is a necessary
step in checking that the license actually does what people think it does.

(Note in the above I do not want to give the impression that I think
a license is necessarily undesirable because it is incompatible with
another license; a license can be incompatible with others and still be
worthwhile)

So, what licenses should a proponent consider when proposing a new one?
I would suggest at the very least the GPLv2 and the MIT license. Ideally
they should consider all the ones labelled "Licenses that are popular and
widely used or with strong communities" on
<http://www.opensource.org/licenses/category>.

So in the spirit of the above, I would like to ask Jon Rosenberg:

Is the MS-PL intended to be compatible with the MIT license?
Is the MS-PL intended to be compatible with the GPLv2?
Is the MS-PL intended to be compatible with the GPLv3?

> Whether Microsoft intended these to be incompatible with the GPL v2 and
> compatible with the Apache License 2.0 would not be an issue except in
> some peoples minds.

The GPLv2 is by far the most common open source license in use today (it's
used by 65% of all projects on Freshmeat). If a proponent of a license hasn't
asked themself the question "should my license be compatible with the GPL?"
and come to a firm answer, then I would suggest they need to think more about
the issue and come back when they have done so.

--
Philip Hunt, cabalamat@...

[John Cowan]

Donovan Hawkins scripsit:

> That is the mechanism, but since the GPL came first I think it is safe to
> say that the MS-PL is intentionally incompatible.

"The devil himself knoweth not the mind of man."  --Justice Brian, 1477

That is, people are responsible for their words and deeds, but not for
their thoughts.

> You both made your choice not to associate with each other and that's
> fine, I never said MS-PL couldn't do that. I said it's not permissive
> and should not be called permissive.

Well, consider this counterevidence, from the FSF licenses page at
http://www.gnu.org/licenses/license-list.html :

Apache License, Version 1.1

    This is a permissive non-copyleft free software license with a few
    requirements that render it incompatible with the GNU GPL.

Apache License, Version 1.0

    This is a simple, permissive non-copyleft free software license
    with practical problems like those of the original BSD license,
    including incompatibility with the GNU GPL.

Original BSD license

    This is a simple, permissive non-copyleft free software license with a
    serious flaw: the "obnoxious BSD advertising clause". The flaw is not
    fatal; that is, it does not render the software non-free. But it does
    cause practical problems, including incompatibility with the GNU GPL.

Old OpenLDAP License, Version 2.3

    This is a permissive non-copyleft free software license with a few
    requirements (in sections 4 and 5) that render it incompatible with
    the GNU GPL.

XFree86 1.1 License

    This is a simple, permissive non-copyleft free software license,
    incompatible with the GNU GPL because of its requirements that
    apply to all documentation in the distribution that contain
    acknowledgements.

Zope Public License version 1

    This is a simple, fairly permissive non-copyleft free software
    license with practical problems like those of the original BSD
    license, including incompatibility with the GNU GPL.


Note the presence of the Yang Worship Word you are talking about in each
and every one of these entries.  I'd say your complaint is rather too
little, too late.

> I would be willing to wager that few people have ever found themselves
> unable to use a piece of (3-clause) BSDL software because of a restriction
> in the license that they were unable to comply with. The same is not true
> for the GPL nor will it be true for the MS-PL. That is the line I am
> saying represents the difference between permissive and non-permissive.

    When I use a word," Humpty Dumpty said, in a rather scornful tone,
    "it means just what I choose it to mean, neither more nor less."

> Incidentally, why is public domain not an option? I joined this list
> recently and haven't heard your discussions in the past on that subject.

Rick's reasons are mine.

> If restrictions like those in MS-PL still count as a permissive license,
> could we at least reserve a term for the licenses that BSDL that don't
> tell me, the developer, how I'm supposed to license my derivative code?

MS-PL doesn't do that either.  It simply requires that you not impose
a license that tries to waive the patent peace clause.  If the GPLed
code in question is your own, you can always add an additional clause
to the GPL either waiving the conflict or imposing the patent peace as
a further restriction.  Or you can use the GPLv3.

> What word means more free than "free" [...]?

Just another word for nothin' left to lose (or give away).

    A cocky novice once said to Stallman: "I can guess why the editor
    is called Emacs, but why is the justifier called Bolio?"  Stallman
    replied forcefully, "Names are but names.  'Emack & Bolio's' is the
    name of a popular ice cream shop in Boston-town.  Neither of these
    men had anything to do with the software."

    His question answered, yet unanswered, the novice turned to go,
    but Stallman called to him, "Neither Emack nor Bolio had anything
    to do with the ice cream shop, either."

--
John Cowan <cowan@...>             http://www.ccil.org/~cowan
Today an interactive brochure website, tomorrow a global content
management system that leverages collective synergy to drive "outside of
the box" thinking and formulate key objectives into a win-win game plan
with a quality-driven approach that focuses on empowering key players
to drive-up their core competencies and increase expectations with an
all-around initiative to drive up the bottom-line. --Alex Papadimoulis

[John Cowan]

Brian Behlendorf scripsit:

> Seems as though while MS-PL is not copyleft, MS-CL is, and thus that
> "smallest difference" better be pretty big to offset the potential cost of
> two universes of immiscible code, MS-CL-licensed and GPL-licensed.

I think (as I thought two years ago) that this is a case where the
anti-proliferation rules should be set aside.  We are dealing with an
organization that has the potential of being a major player in free and
open source software (and if they don't like the GPL, there are plenty
of other FLOSS-producing organizations that don't like it either).
If they can only bring themselves to release such software under their
own particular licenses, so much the worse; but not more the worse than
if they never released any FLOSS software at all.

        That would be the John Cowan I know, yes: always assuming the
        best of people and always having the best of intentions. But he
        still doesn't always think all of the consequences through :)
                --Tsela (Christophe Grandsire) on Groklaw, December 2005

--
John Cowan   http://ccil.org/~cowan    cowan@...
In might the Feanorians / that swore the unforgotten oath
brought war into Arvernien / with burning and with broken troth.
and Elwing from her fastness dim / then cast her in the waters wide,
but like a mew was swiftly borne, / uplifted o'er the roaring tide.
        --the Earendillinwe

[Donovan Hawkins]

On Sun, 19 Aug 2007, John Cowan wrote:

> Donovan Hawkins scripsit:
>
>> You both made your choice not to associate with each other and that's
>> fine, I never said MS-PL couldn't do that. I said it's not permissive
>> and should not be called permissive.
>
> Well, consider this counterevidence, from the FSF licenses page at
> http://www.gnu.org/licenses/license-list.html :

<snipped descriptions for brevity>

> Apache License, Version 1.1
> Apache License, Version 1.0
> Original BSD license
> Old OpenLDAP License, Version 2.3
> XFree86 1.1 License
> Zope Public License version 1
>
> Note the presence of the Yang Worship Word you are talking about in each
> and every one of these entries.  I'd say your complaint is rather too
> little, too late.

Looking through all those licenses, all but two (discussed below) are
incompatible due to a flawed advertising clause like that of original
BSDL. For the other two:

OpenLDAP 2.3 appears to be compatible with GPL v3. Section 4 of OpenLDAP
is specifically allowed by section 7d of GPL, and section 5 of OpenLDAP is
specifically allowed by section 7c of GPL.

XFree86 1.1 appears to be compatible with GPL v3. Section 7b of GPL
specifically allows "requiring preservation of specified reasonable legal
notices or author attributions in that material or in the Appropriate
Legal Notices displayed by works containing it." I don't see anything in
XFree86 1.1 that is obviously unreasonable in that regard.


So the only thing standing between these "permissive" licenses and
compatiblity with GPL v3 is the flawed advertising clause from BSDL,
something that is widely regarded as an impractical mistake. Whether you
call these "non-permissive" or "permissive but flawed" is a minor
point...I would choose "non-permissive" myself.


>> Incidentally, why is public domain not an option? I joined this list
>> recently and haven't heard your discussions in the past on that subject.
>
> Rick's reasons are mine.

Ah ok, I see the problem. (thank you to those who posted on that subject
here)


>> If restrictions like those in MS-PL still count as a permissive license,
>> could we at least reserve a term for the licenses that BSDL that don't
>> tell me, the developer, how I'm supposed to license my derivative code?
>
> MS-PL doesn't do that either.  It simply requires that you not impose
> a license that tries to waive the patent peace clause.  If the GPLed
> code in question is your own, you can always add an additional clause
> to the GPL either waiving the conflict or imposing the patent peace as
> a further restriction.

Yes, that would be "telling me how I'm supposed to license my derivative
work". I can't license my entire derivative work under BSDL or GPL v3 or
another MS license or any other license. Your suggested workarounds don't
change this fact.


> Or you can use the GPLv3.

Why do you believe that MS-PL is compatible with GPL v3? Section 5c says
of modified versions of GPL works:

"You must license the entire work, as a whole, under this License to
anyone who comes into possession of a copy.  This License will therefore
apply, along with any applicable section 7 additional terms, to the whole
of the work, and all its parts, regardless of how they are packaged."

At best I could release my changes under GPL v3 together with other code
under MS-PL, but no further derived works could be created under either
license because 5c cannot be met. If you mean I can waive this requirement
with an additional permissive term, and that all derivative works can do
the same for the code they add, well...that's not GPL v3 that I used
anymore. I've simply created a new license (GPL v3 + that term), and this
new license is not compatible with GPL v3.

This also requires that my changes be separable from the MS-PL code so
that different licenses can be applied to each. This is the more serious
problem which applies to any license you use for the derivative work.

---------------------------------------------------------------------------
Donovan Hawkins, PhD                 "The study of physics will always be
Software Engineer                     safer than biology, for while the
hawkins@...                   hazards of physics drop off as 1/r^2,
http://www.cephira.com                biological ones grow exponentially."
---------------------------------------------------------------------------