I am replying off list with an expanded discussion here.
>-----Original Message-----
>From: owner-zebra@zebra.org [mailto:owner-zebra@zebra.org]On Behalf Of
>Paul Jakma
>
>> Thus ANY code that carries a BSD license is immediately affected no
>> matter how old or how new it is.
>
>Are you sure? Look at the pimd licence, it does not have any kind of
>'supercedence' clause, it does not refer to any licence outside of
>itself.
>
If the pimd license does not say "Copyright Regents of the University
of California" then it is not BSD licensed because an
inherent property of BSD license is that the copyright ownership is
transferred to UCB. That is why it's called BSD, the B is for
Berkeley.
>
>pimd is:
>
> * University of Southern California/Information Sciences Institute.
>
It's not BSD license then. There's lots of BSD-like licenses out there.
One of the problems is that BSD as a term is in the process of becoming
a generic term indicating a whole group of licenses. The example I
gave with David Greenman retaining copyright on a kernel file of
FreeBSD is a symptom of this - to most people, that is "BSD code" and
even the BSD website itself makes that claim, when technically that
file is not BSD as it's not copyrighted by UCB. Most of the general
public can't get the distinction and as a result is buying into the
morphing of the BSD term into a generic term.
The same thing is happening with GNU/GPL and "Linux" and boy is
it creating screaming in some GPL quarters. That is why RMS insists
on calling Linux "GNU/Linux" because he is horrified that the
general public is beginning to use the generic term "Linux" to
mean "GPL'd source"
>> This is an impossible statement because it is impossible for any BSD
>> license to carry an advertising clause, as the copyright holder has
>> revoked this clause.
>
>well, as long as the previous copyright carried some kind of 'or later
>version of this licence'.
>
UCB owns the copyright on BSD code they can do whatever they
want with it. They don't need to have "or later version"
stated in the license simply because they DON'T have a statement
of the effect "this license will NOT be superseded by later versions"
or the more usual "if any portion of this license is later
invalidated then the remaining portions remain in force" which
is basically the same thing. The lack of such a disclaimer infers the "or
later version" statement your talking about.
Also, even more importantly from a court's POV is that the
bulk of the BSD community and certainly all the BSD community
leaders all agree with the advertising clause is dead for all
BSD licensed code.
Don't forget that UCB did go to court over the BSD license and
fought a long expensive court battle with AT&T over it that
resulted in an IN COURT judgement, not some sneaky "lets settle
this out of court and hide the terms" copout. That is
more I might add than the FSF has ever done.
>
>"widely pirate BSD code"? err... the BSD licence specifically allows
>this kind of thing, eg incorporation of BSD code into derivative
>works. At least if the derivative work is GPL the source is still
>available - what about BSD code being incorporated into completely
>closed proprietary works? Isnt that even worse piracy then? (but isnt
>one of the advantages of the BSD licence that it allows this??).
>
Anyone who copies BSD code and then DOESEN'T credit UCB for it or
otherwise indicate that at least portions of the code originate from
Berkeley is filthy pirating scum, whether they are commercial or
not. Yes, some commercial companies do the worst offender I can
think of is Apple, which used huge chunks of FreeBSD 3.2 and then
stupid ass Jobs was such an idiot that he said at a trade show that
MacOS X was partly built from Linux!!! Apple goes to great pains to
conceal the existence of BSD in MacOS X, although I will also say
that they go to pains to conceal that MacOS X is UNIX!!! I still
have no respect for them, though.
But there are PLENTY of commercial companies that DO credit BSD
and the bad apples should not spoil the barrel.
This is a question of morals, not of legalese.
The GPL insistence that the advertising clause is burdensome is
plainly morally wrong. Since when is it a burden to credit the
originator?
Even Microsoft retained the BSD copyright notice when they recompiled the
Berkeley utilities under Windows.
>> the copyright holder be credited. Advertising clause, indeed. The
>> entire GPL license itself is nothing more than one giant
>> "advertising clause"
>
>but that was a concious decision on the part of the people who wrote
>the BSD licence and on the part of the FSF wrt to the GPL. whats the
>problem? :)
>
Licenses don't exist in a vacuum, as anyone who has had one tested in a
court finds out to their dismay. There's a large body of writings and
actions that has bearing on a license. You cannot simply take GPL at
it's face value - if you do, you actually find a rather unpresumptious
license. Instead you must take GPL not only as the license itself, but
what everyone who is using it has to say about HOW to use it. And that
is where you find the hypocrisy. For example, the FSF pretty much demands
in their writings for GPL-licensed code have copyright transferred to
FSF - they make an explicit threat that they will not fund any legal
defense of infringements on GPL unless they get the copyright. You can
infer from this that they are saying that the GPL license is unenforceable
unless copyright is transferred to FSF - which is the same as saying that the
GPL requires that copyright be transferred to FSF - however the GPL license
itself does not state that.
Another example is FSF's insistence that mere linking into GPL
code makes any code modules so linked automatically fall under GPL. Once
again, a claim that is pure FUD and has never been tested in court, and
is clearly silly. You can possibly make a case that a resultant
binary is GPL, but merely including a BSD source file into a program
then compiling the result now somehow invalidates the BSD copyright
on that module? Come off it!!
The GPL is much more than a mere license, it's the license
plus all the body of supporting documentation and use of the license
that surrounds it. In short, the GPL community. The same such
community surrounds the BSD license.
Now, you ask where the problem is - well the problem is that the
BSD community is not insisting that the GPL community modify it's
license or how it works with it, indeed works like FreeBSD have
created an entire ports structure for the benefit of the GPL and licenses like
it, so that those authors don't have to relicense their stuff under BSD, and
can retain their GPL --- ON THE OTHER HAND the GPL community makes
no such accomodation to BSD, (or other license holders) to accomodate
their preferred licenses, from the GPL's point of view, the GPL
license must subsume every other license out there, and demands that
all other licenses either be GPL or be modified so that it's legal
to strip out the license and replace it with GPL.
In short, the GPL community says DO as I SAY, not as I DO. This is
hypocrisy, and if you cannot see that I pity you.
And the GPL community whines about Microsoft's Embrace and Extend
policy! They should look in a mirror!
Ted Mittelstaedt tedm@toybox.placo.com
Author of: The FreeBSD Corporate Networker's Guide
Book website: http://www.freebsd-corp-net-guide.com
>-----Original Message-----
>From: owner-zebra@zebra.org [mailto:owner-zebra@zebra.org]On Behalf Of
>Paul Jakma
>
>> Thus ANY code that carries a BSD license is immediately affected no
>> matter how old or how new it is.
>
>Are you sure? Look at the pimd licence, it does not have any kind of
>'supercedence' clause, it does not refer to any licence outside of
>itself.
>
If the pimd license does not say "Copyright Regents of the University
of California" then it is not BSD licensed because an
inherent property of BSD license is that the copyright ownership is
transferred to UCB. That is why it's called BSD, the B is for
Berkeley.
>
>pimd is:
>
> * University of Southern California/Information Sciences Institute.
>
It's not BSD license then. There's lots of BSD-like licenses out there.
One of the problems is that BSD as a term is in the process of becoming
a generic term indicating a whole group of licenses. The example I
gave with David Greenman retaining copyright on a kernel file of
FreeBSD is a symptom of this - to most people, that is "BSD code" and
even the BSD website itself makes that claim, when technically that
file is not BSD as it's not copyrighted by UCB. Most of the general
public can't get the distinction and as a result is buying into the
morphing of the BSD term into a generic term.
The same thing is happening with GNU/GPL and "Linux" and boy is
it creating screaming in some GPL quarters. That is why RMS insists
on calling Linux "GNU/Linux" because he is horrified that the
general public is beginning to use the generic term "Linux" to
mean "GPL'd source"
>> This is an impossible statement because it is impossible for any BSD
>> license to carry an advertising clause, as the copyright holder has
>> revoked this clause.
>
>well, as long as the previous copyright carried some kind of 'or later
>version of this licence'.
>
UCB owns the copyright on BSD code they can do whatever they
want with it. They don't need to have "or later version"
stated in the license simply because they DON'T have a statement
of the effect "this license will NOT be superseded by later versions"
or the more usual "if any portion of this license is later
invalidated then the remaining portions remain in force" which
is basically the same thing. The lack of such a disclaimer infers the "or
later version" statement your talking about.
Also, even more importantly from a court's POV is that the
bulk of the BSD community and certainly all the BSD community
leaders all agree with the advertising clause is dead for all
BSD licensed code.
Don't forget that UCB did go to court over the BSD license and
fought a long expensive court battle with AT&T over it that
resulted in an IN COURT judgement, not some sneaky "lets settle
this out of court and hide the terms" copout. That is
more I might add than the FSF has ever done.
>
>"widely pirate BSD code"? err... the BSD licence specifically allows
>this kind of thing, eg incorporation of BSD code into derivative
>works. At least if the derivative work is GPL the source is still
>available - what about BSD code being incorporated into completely
>closed proprietary works? Isnt that even worse piracy then? (but isnt
>one of the advantages of the BSD licence that it allows this??).
>
Anyone who copies BSD code and then DOESEN'T credit UCB for it or
otherwise indicate that at least portions of the code originate from
Berkeley is filthy pirating scum, whether they are commercial or
not. Yes, some commercial companies do the worst offender I can
think of is Apple, which used huge chunks of FreeBSD 3.2 and then
stupid ass Jobs was such an idiot that he said at a trade show that
MacOS X was partly built from Linux!!! Apple goes to great pains to
conceal the existence of BSD in MacOS X, although I will also say
that they go to pains to conceal that MacOS X is UNIX!!! I still
have no respect for them, though.
But there are PLENTY of commercial companies that DO credit BSD
and the bad apples should not spoil the barrel.
This is a question of morals, not of legalese.
The GPL insistence that the advertising clause is burdensome is
plainly morally wrong. Since when is it a burden to credit the
originator?
Even Microsoft retained the BSD copyright notice when they recompiled the
Berkeley utilities under Windows.
>> the copyright holder be credited. Advertising clause, indeed. The
>> entire GPL license itself is nothing more than one giant
>> "advertising clause"
>
>but that was a concious decision on the part of the people who wrote
>the BSD licence and on the part of the FSF wrt to the GPL. whats the
>problem? :)
>
Licenses don't exist in a vacuum, as anyone who has had one tested in a
court finds out to their dismay. There's a large body of writings and
actions that has bearing on a license. You cannot simply take GPL at
it's face value - if you do, you actually find a rather unpresumptious
license. Instead you must take GPL not only as the license itself, but
what everyone who is using it has to say about HOW to use it. And that
is where you find the hypocrisy. For example, the FSF pretty much demands
in their writings for GPL-licensed code have copyright transferred to
FSF - they make an explicit threat that they will not fund any legal
defense of infringements on GPL unless they get the copyright. You can
infer from this that they are saying that the GPL license is unenforceable
unless copyright is transferred to FSF - which is the same as saying that the
GPL requires that copyright be transferred to FSF - however the GPL license
itself does not state that.
Another example is FSF's insistence that mere linking into GPL
code makes any code modules so linked automatically fall under GPL. Once
again, a claim that is pure FUD and has never been tested in court, and
is clearly silly. You can possibly make a case that a resultant
binary is GPL, but merely including a BSD source file into a program
then compiling the result now somehow invalidates the BSD copyright
on that module? Come off it!!
The GPL is much more than a mere license, it's the license
plus all the body of supporting documentation and use of the license
that surrounds it. In short, the GPL community. The same such
community surrounds the BSD license.
Now, you ask where the problem is - well the problem is that the
BSD community is not insisting that the GPL community modify it's
license or how it works with it, indeed works like FreeBSD have
created an entire ports structure for the benefit of the GPL and licenses like
it, so that those authors don't have to relicense their stuff under BSD, and
can retain their GPL --- ON THE OTHER HAND the GPL community makes
no such accomodation to BSD, (or other license holders) to accomodate
their preferred licenses, from the GPL's point of view, the GPL
license must subsume every other license out there, and demands that
all other licenses either be GPL or be modified so that it's legal
to strip out the license and replace it with GPL.
In short, the GPL community says DO as I SAY, not as I DO. This is
hypocrisy, and if you cannot see that I pity you.
And the GPL community whines about Microsoft's Embrace and Extend
policy! They should look in a mirror!
Ted Mittelstaedt tedm@toybox.placo.com
Author of: The FreeBSD Corporate Networker's Guide
Book website: http://www.freebsd-corp-net-guide.com