Mailing List Archive

hi ted,

On Thu, 17 Apr 2003, Ted Mittelstaedt wrote:

> 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.

i stand corrected.

> 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.

indeed.

> 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"

nah, i think he's just horrified that all the newish linux users dont
know anything about FSF or GNU or its ideals.

> 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 "this license will NOT be superseded by later versions" or the more

yes, but if a copyright holder grants someone a licence for an
indefinite period of time that licence applies indefinitely. the
copyright holder can not automatically supercede all previous licences
unless those previous licences have a clause granting the copyright
holder such rights.

thats why the GPL /does/ refer to "or later version".

least that's my understanding.

> usual "if any portion of this license is later invalidated then the
> remaining portions remain in force" which is basically the same
> thing.

I'd have to respectfully disagree. I dont see how lack of such a
clause would infer the copyright holder retains rights to
retroactively modify previously granted licences.

> The lack of such a disclaimer infers the "or later version"
> statement your talking about.

I dont see how. care to explain. The recindment of the advertising
clause does not refer to it being retroactive either:

ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change

> 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.

Yes, i'd agree.

> 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,

Indeed. That BSD, aside from certain minor portions, did not contain
AT&T copyrighted code, iirc. That case did not judge on the BSD
licence though, did it?

> 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.

So?

I detect some kind of strange 'hang-up' on your part re the FSF and,
more so, the GPL.

They're just licences :)

> 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,

Err... how exactly?

Code under the present BSD licence (in the generic sense) /allows/
this. How is it piracy?

The person who writes the code decides the licence. If they want
credit, they can have a 'advertise credit' clause. If they want
anybody and everybody to be able to use and incorporate the code, they
use the current BSD licence, if they want the code to remain open and
for derivative works to remain open they probably should choose for
the GPL.

Its the choice of the person who wrote the code.

If someone wrote code and released it under the (current) BSD licence,
then they have chosen to allow people not to have to credit them.

I dont see how piracy comes into it at all. (unless you're suggesting
the majority of contributors to BSD licenced projects all somehow are
now going "D'Oh! i'd forgotten i gave people a licence to not credit
me!" and are kicking themselves for it?).

Anyway, its by choice.

> 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!!!

Did he? stupid of him alright :)

> 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 if the FreeBSD licence, ie BSD, did not allow Apple to incorporate
that code into their proprietary work do you think it is possible
Apple might have chosen not use it? Possibly to go develop their own
OS instead? Possible that then MacOS X wouldnt be Unix at all?

I think it better that MacOS X is Unix and that Apple doesnt say much
about it, than that MacOS X not be Unix at all. And this is the
beauty of the BSD licence, its makes existing code attractive to those
who wish to build proprietary code around it, and hence allows that
code greater and more diverse use than it otherwise might receive.

Its also the downside to the BSD licence, but hey.

> But there are PLENTY of commercial companies that DO credit BSD and
> the bad apples should not spoil the barrel.

But the licence allows them to do this.

> This is a question of morals, not of legalese.

Oh come on. morals are intangible. Law and the licence's and contracts
built on civil law are mans attempt to try codify what is and is not
permissible in society and in dealings between members of that society
in general and in specific cases.

In essence, your argument's is that there are moral obligations which
it does not enumerate which people sometimes ignore. Hence the logical
conclusion to your arguments is that really the BSD licence should be
fixed, to codify these moral obligations into the licence.

I think most BSD coders would disagree with such a conclusion. Hence i
suspect your argument about there being moral obligations to the BSD
licence is flawed.

> The GPL insistence that the advertising clause is burdensome is
> plainly morally wrong.

That is not the GPL's insistence. The problem was that this run-time
advertising clause was incompatible with the GPL.

> Since when is it a burden to credit the originator?

It isnt really, just the BSD licence was not compatible with the GPL.

> Even Microsoft retained the BSD copyright notice when they
> recompiled the Berkeley utilities under Windows.

Indeed. That was the licence they got to use the code under.

> Licenses don't exist in a vacuum, as anyone who has had one tested
> in a court finds out to their dismay.

Absolutely.

> There's a large body of writings and actions that has bearing on a
> license.

/Huge/ amount have a look at the licence-discuss list sometime on, i
think, opendev.org (or somesuch). It varies by jurisdiction too. :(

> You cannot simply take GPL at it's face value - if you do, you
> actually find a rather unpresumptious license.

Well, unpresumptious but deliberately vague in definition.

> 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.

Ah..

> 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.

Why is that hypocrisy?

"I want your lawyers time!"

"Ok, but to do the job well, we need the copyright"

Seems fair enough to me. The other reason for this is that the
copyright holder can issue code under different licences or release
later code under a different licence. The FSF does not want to spend
time and money on code that the author(s)/copyright holders could
concievably later on change to another (eg closed/proprietary
licence).

But hey, if you dont want to assign copyright to the FSF, you can
always get your own lawyers. There's no FSF 'heavy mob' that come
around and threaten to break your legs - free choice.

> 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.

Yes, but with the GPL it doesnt really matter who holds the copyright.
(presuming no one intends to change the licence).

> Another example is FSF's insistence that mere linking into GPL code
> makes any code modules so linked automatically fall under GPL.

That is not their insistence though, is it?

> Once again, a claim that is pure FUD and has never been tested in
> court, and is clearly silly.

Indeed it is, but i dont think the FSF has made that claim.

> 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.

Absolutely.

But again, all that matters legally is the licence and the legal
context in which it stands. And it is by that which 3rd parties will
use the code. If a 3rd party is not part of the community then how
will they ever know about the moral obligations? they wont. (hence why
the licence needs to specify what is and is not permissible).

> 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,

Indeed.

> 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,

Course not, because the GPL is designed to keep code GPL compatible.
Thats the way things work.

NB: there are plenty of cases btw of improvements to GPLed code which
originally was derived from BSD licenced code being dual-licenced by
the copyright holder so as to allow it to be incorporated back into
the original BSD code. (various linux drivers eg).

> 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.

Indeed. That is its job. Whats wrong with that?

> 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.

I pity you that you can get so worked up about an inanimate legal
text.

The GPL exists to provide a certain protection to code.

The BSD licence exists to provide a certain protection (very little)
to code.

People who write code get to /choose/ what licence they release it
under, be it GPL, BSD or proprietary or even all 3 of them.

> And the GPL community whines about Microsoft's Embrace and Extend
> policy! They should look in a mirror!

Sorry, now you're ranting. :)

> Ted Mittelstaedt tedm@toybox.placo.com

regards,
--
Paul Jakma Sys Admin Alphyra
paulj@alphyra.ie
Warning: /never/ send email to spam@dishone.st or trap@dishone.st

>-----Original Message-----
>From: Paul Jakma [mailto:paulj@alphyra.ie]
>
>nah, i think he's just horrified that all the newish linux users dont
>know anything about FSF or GNU or its ideals.
>

Heh. :-) Although, I think he's just as horrified that _everyone_
don't know about the FSF.

>> 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
>
>yes, but if a copyright holder grants someone a licence for an
>indefinite period of time that licence applies indefinitely. the
>copyright holder can not automatically supercede all previous licences
>unless those previous licences have a clause granting the copyright
>holder such rights.
>

Not true in this case, let me explain.

Let's assume you, Paul, files a lawsuit against UCB alleging
an illegal retroactive modification of the BSD license.

You go before the judge and say "I am a BSD software user and
the original license granted by UCB has an advertising clause,
UCB just told me that this clause is revoked, but the original
grant of license doesen't permit them to do this. Thus they
are in breech of contract and I'm suing for damages"

Let's assume that you are right and the judge does indeed
agree with you, that UCB is in breech of contract.

Then he is going to ask what damages you suffer.

In this case, because the retroactive modification is less
restrictive, it is impossible for you to have suffered any
damages.

Thus you are awarded a judgement of $1, and the judge then
proceeds to fine you $2000 for bringing a frivolous lawsuit
into the court and wasting their time! :)

In summary, your never going to see such a lawsuit because
nobody is going to risk the ire of a court over this. Thus,
in this case, the copyright holder CAN make an
"advertising clause" retroactive revocation simply because
they say they can do so, and nobody is ever going to challenge
them on it - and that is exactly what happened here.

It's like paying income taxes. Technically, the Income Tax is
unconstitutional, thus illegal in theory. But it's "legal" merely
because the government says it is, and nobody has ever successfully
won against the government over it.

There is, unfortunately, a lot of law out there like this -
stuff that's technically legal, but your going to go to jail for
doing it anyway, if they catch you doing it, and stuff that's
technically illegal, but in reality nobody is ever going to be
prosecuted on it, and if some cop is ever dumb enough to try it,
(like the case a while back somewhere in Alabama where a cop saw
through an open window some guy's wife giving him a blow job and
hauled them both in on sodomy charges) the case is thrown out
of court along with strong language by the judge telling the cops
don't ever try this again.

Anyway, besides all that, software copyright also carries with
it this messy legal concept of "implied rights"

Implied rights are what makes it illegal for you to buy a book
then go make copies of it and give them out. Notice a book does
not have anything in it explicitly prohibiting you from doing
this, yet it's illegal because the implied rights of a book
purchaser don't include this ability.

Now, as books have been published for hundreds of years, what
is defined as implied rights with a book purchase are pretty much
etched in stone these days.

But implied rights associated with software purchases is still
very, very much being argued over. That is why software carries
shrink wrap licenses, because the software houses don't feel
at the current time that what makes up implied rights with
a software purchase is still not clearly defined, so they spell
it out. And even then most judges would reject a shrinkwrap license
that significantly deviated from industry norms, and the software
industry is fighting off efforts to have the entire shrinkwrap
licenses tested in a high court anyway as they know it's shaky.

>thats why the GPL /does/ refer to "or later version".
>

The FSF spent a lot of money in legal fees and the lawyer that
wrote the license had to put a lot of boilerplate in to justify
the fees he was charging. ;-)

>least that's my understanding.
>

Law don't exist in a vacuum. Like I said earlier, everything
around the law has to be considered as well.

>
>I dont see how. care to explain. The recindment of the advertising
>clause does not refer to it being retroactive either:
>
>ftp://ftp.cs.berkeley.edu/pub/4bsd/README.Impt.License.Change
>

Reread the following from this letter:

"Effective immediately, licensees and distributors are no longer required to
include the acknowledgement within advertising materials. Accordingly, the
foregoing paragraph of those BSD Unix files containing it is hereby deleted"

This is a retroactive recindment because what they are talking about
is defined in the letter here:

"certain of the Berkeley Software Distribution ("BSD") sourcecode files "

They don't explicitly say the word retroactive, but what they are talking
about is existing source files, so it is retroactive.

>> 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.
>
>Yes, i'd agree.
>
>> 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,
>
>Indeed. That BSD, aside from certain minor portions, did not contain
>AT&T copyrighted code, iirc. That case did not judge on the BSD
>licence though, did it?
>

It actually did, by implication. In short, if the BSD license wasn't
considered by the court to be a "real" license, then AT&T would
not have argued that BSD was infringing. By saying that it is, they
are inadvertantly acknowledging that BSD was a separate
distribution which implied a separate license.

Otherwise AT&T would have merely argued that UCB was in breech of
contract because they were misusing AT&T code, making the implicit
assumption that ALL of BSD code was really AT&T code. Instead they
singled out certain files, labeling them AT&T, and the rest of them
they acknowledged were BSD-licensed.

This kind of thing is why court battles are so great - lots of
things happen that are besides the point, some of those things
eventually eclipse the point itself. Witness the finding of fact
that Microsoft is a monopoly. That one thing is going to do
more damage to them over the long term than the simple loss of the
court case. (remember that while the penalties were a slap on the
hand, they DID lose)

>
>I detect some kind of strange 'hang-up' on your part re the FSF and,
>more so, the GPL.
>

Bing bing bing!!! Your absolutely right!!! :)

>They're just licences :)
>

That is like saying the Magna Carta is "just a piece of paper"

Paul, I can't believe that you don't realize the historical
significance of the GPL and BSD software movements by now. I
think your just baiting me.

What is going on now, today, is precedent setting for software
and the results are going to dictate software for at least the
next century. The licenses are a codification of fundamental
directions open source is taking - discussion of them is
discussion of open source's future.

>> 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,
>
>Err... how exactly?
>
>Code under the present BSD licence (in the generic sense) /allows/
>this. How is it piracy?
>

No it does not, reread this section:

* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright


Quite obviously, keeping the copyright in there is crediting them,
and the license itself says your supposed to do that!! Not doing it
is scum, as I said.

>The person who writes the code decides the licence.

I'm not talking about any other license here than GPL or
BSD. Anyone that doesen't want to use either of those can
do whatever they want, fine with me.

>
>If someone wrote code and released it under the (current) BSD licence,
>then they have chosen to allow people not to have to credit them.
>

No, not at all. They are expecting that the copyright be retained,
and the copyright credits UCB. Quite often, it also makes further
credits at the beginning of the license for the file in question,
that is standard procedure actually.

>
>But if the FreeBSD licence, ie BSD, did not allow Apple to incorporate
>that code into their proprietary work do you think it is possible
>Apple might have chosen not use it? Possibly to go develop their own
>OS instead? Possible that then MacOS X wouldnt be Unix at all?
>

As far as the question of would Apple have ever become a UNIX licensee
if FreeBSD hadn't been available? Well I think the answer to that
is that yes, they would have. Keep in mind that Apple ALREADY
went down the road of developing their own OS, and still chose
UNIX in the end. Clearly, MacOS led to a dead end, and espically
for Apple, admitting this was as damaging to their psyche as
when Bill Gates bought 10% or whatever it was of Apple.

>I think it better that MacOS X is Unix and that Apple doesnt say much
>about it, than that MacOS X not be Unix at all. And this is the
>beauty of the BSD licence, its makes existing code attractive to those
>who wish to build proprietary code around it, and hence allows that
>code greater and more diverse use than it otherwise might receive.
>
>Its also the downside to the BSD licence, but hey.
>

No, this isn't the case either.

The beauty of the BSD license is simply that it truly embodies the
idea of real freedom in software licensing.

And as for the downside, that's an engineers phantom. In reality,
the BSD code really doesen't permit people to build proprietary code
around it, unless they are willing to only use the BSD code as a starting
base.

The Apple case is a perfect example. Apple started out on FreeBSD
3.2, but FreeBSD has moved on to 5.X which is far and away more advanced
than 3.2 was, espically in the area of multiprocessing. But since
Apple made it proprietary, they cannot take advantage of any of
the 5.X features without an expensive porting job, a job that gets
harder and harder as future versions of FreeBSD diverge further
away from the public FreeBSD code. Eventually, the Apple code
really becomes indistinguishable from a proprietary project
developed from scratch, and Apple eventually has no competitive
advantage over it's commercial competitors by using BSD code.

The only way to really take advantage of BSD code is to keep
contributing your mods back to it, so that as people change
the source code, they have to keep your mods updated since your
mods are part of the codebase.

>
>> This is a question of morals, not of legalese.
>
>Oh come on. morals are intangible. Law and the licence's and contracts
>built on civil law are mans attempt to try codify what is and is not
>permissible in society and in dealings between members of that society
>in general and in specific cases.
>

Law is founded on morality, and laws are only upheld by people
that are moral people. Without morals, laws would have no meaning.

>In essence, your argument's is that there are moral obligations which
>it does not enumerate which people sometimes ignore.

Correct.

>Hence the logical
>conclusion to your arguments is that really the BSD licence should be
>fixed, to codify these moral obligations into the licence.
>

No, absolutely not. This is the GPL approach - to codify every last
little thing - because fundamentally the GPL approach is to not
trust people to do the right thing.

Yes there's moral obligations. But as to what to do if people
break those moral obligations - well there are many options. One
option is to simply ignore them.

The BSD license is all about freedom. One of the hardest things
about freedom is that unless you permit people to do things that
you don't want them to do, your not really giving them freedom.
But just because you permit them to do it, doesen't mean you
have to respect them for it, or agree with what they are doing.

This is why it is so hypocritical for the FSF to insist that the
GPL is "free software" It isn't. GPL code is very restrictive,
that is not freedom.

Legally there's now no claim on a BSD software user to honor the
advertising clause in BSD. However, the BSD advertising clause is
central to one of the tenants of the BSD license - giving credit
back to UCB for BSD. Even without the advertising clause in there
I argue that there's a history of crediting UCB on BSD software that
anyone in the business is aware of, and if they use BSD they should
be following. Also it's just plain good manners to credit those
that help you, and people that won't lift a finger to do so (ie:
add 1 line in their manual crediting BSD) are scum. In short, the
advertising clause is an enumeration of a moral obligation that is
implied by anyone using BSD software, in fact, any software. Hell,
I'm using Microsoft Outlook to compose this message right now,
and there's a credit to Microsoft right in the header of the message.
(X-Mailer) Crediting the people that write the software or
create the products, or perform the services that help you in major
ways to to do things is a moral obligation of anyone for anything.

>
>> The GPL insistence that the advertising clause is burdensome is
>> plainly morally wrong.
>
>That is not the GPL's insistence. The problem was that this run-time
>advertising clause was incompatible with the GPL.
>

And the GPL is incompatible with BSD, yet FreeBSD found a way to
accomodate GPL code without adopting the GPL license on it's stuff.
Why is it so hard for the GPL to extend the
same courtesy? Don't they have enough brains to figure out how to
do this in GPL-land? Are the BSD people smarter than them because
they worked out how to do it?

This incompatability is NOT a technical obstacle, there's ways around
it. It's a political obstacle is what is is - because GPL wants to
take BSD code and apply the GPL license to it, for no other reason
than they don't happen to like any other way of licensing code
than GPL, to the point that they don't even want to admit that
other licenses even exist.

>
>> 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.
>
>Ah..
>
>> 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.
>
>Why is that hypocrisy?
>

That isn't. You left out the part where I detailed what is
hypocritical.

>"I want your lawyers time!"
>
>"Ok, but to do the job well, we need the copyright"
>
>Seems fair enough to me. The other reason for this is that the
>copyright holder can issue code under different licences or release
>later code under a different licence. The FSF does not want to spend
>time and money on code that the author(s)/copyright holders could
>concievably later on change to another (eg closed/proprietary
>licence).
>
>But hey, if you dont want to assign copyright to the FSF, you can
>always get your own lawyers. There's no FSF 'heavy mob' that come
>around and threaten to break your legs - free choice.
>

Even better - if you don't want to assign copyright to FSF - simply
don't use GPL!

A license is only worth anything if the copyright holder is willing
to defend it. Otherwise it's only so much air blowing out of someone's
ass. For copyright holders that don't assign copyright to FSF, there
is little point to using the GPL license since they are going to
defend their copyright on their code themselves. But to hear or read the
average GPL advocate, they won't say this at all.

>
>> Another example is FSF's insistence that mere linking into GPL code
>> makes any code modules so linked automatically fall under GPL.
>
>That is not their insistence though, is it?
>

Yes it is. See:

http://www.fsf.org/licenses/gpl-faq.html#LinkingWithGPL

>> Once again, a claim that is pure FUD and has never been tested in
>> court, and is clearly silly.
>
>Indeed it is, but i dont think the FSF has made that claim.
>

Yes they have. See the GPL FAQ.

>> 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.
>
>Absolutely.
>
>But again, all that matters legally is the licence and the legal
>context in which it stands. And it is by that which 3rd parties will
>use the code. If a 3rd party is not part of the community then how
>will they ever know about the moral obligations? they wont. (hence why
>the licence needs to specify what is and is not permissible).
>

This is just an assuption that all people are stupid, plain and
simple. The idea that a commercial developer is going to
just run across BSD code and merrily include it into their
proprietary stuff without investigating the BSD community and
asking questions of that community is fiction. They are going
to investigate the legal context surrounding the code's license
and the will certainly find out about the moral obligations then.

>> 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,
>
>Indeed.
>
>> 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,
>
>Course not, because the GPL is designed to keep code GPL compatible.
>Thats the way things work.
>
>NB: there are plenty of cases btw of improvements to GPLed code which
>originally was derived from BSD licenced code being dual-licenced by
>the copyright holder so as to allow it to be incorporated back into
>the original BSD code. (various linux drivers eg).
>

That is true and it is something that I overlooked. But there's
a difference between what your talking about and what I'm talking about.

It's one thing to take some minor BSD functions in some BSD code that are
useful for your GPL project and incorporate them into your GPL code,
then underneath the GPL license in your code, insert the BSD copyright
with the notation that your only doing that to credit BSD as the GPL
supersedes it for licensing purposes.

It's another to take an entire BSD program, or an entire BSD sourcefile,
and replace the BSD copyright with the GPL copyright just because you
don't like BSD.

>> 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.
>
>Indeed. That is its job. Whats wrong with that?
>

Because there is no need for this in a distribution such as an
operating system, (ie: Linux) or a large complex program which
uses some smaller utilities in it. And there's also no need for
it with BSD-licensed libraries, except that the FSF decided to
arbitrairly say that if you link in a library, it automatically
falls under GPL - something that is such a bald grab that you
didn't even believe it when I wrote you with it, so I had to
send you back to the GPL FAQ.

>> 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.
>
>I pity you that you can get so worked up about an inanimate legal
>text.

Are you saying your education was so lacking that none of your
teachers ever explained that the pen is mightier than the
sword?

More people have been killed over history as a result of
ideas written down on inanimate text than any other reason.

And in any case, your being a hypocrite since you are obviously
just as worked up as I am, otherwise you wouldn't have written
such a long reply to my mail.

>
>The GPL exists to provide a certain protection to code.
>
>The BSD licence exists to provide a certain protection (very little)
>to code.
>
>People who write code get to /choose/ what licence they release it
>under, be it GPL, BSD or proprietary or even all 3 of them.
>

I was never arguing that they don't have a choice. If someone wants
to use GPL for THEIR stuff, more power to them. I am objecting that
the community of GPL proponents basically never miss an opportunity
to snub BSD and flat out lie about what BSD is all about, when they
really owe their existence to BSD code activities that preceded the
creation of GPL.

You (I assume) inadvertantly did this when you posted your comments about BSD.
I am just glad that I was able to educate you as to what the truth is about
the BSD advertising clause is, so you don't continue to
repeat the wrong GPL propaganda about BSD you had stated in the
Zebra mailing list.

Ted Mittelstaedt tedm@toybox.placo.com
Author of: The FreeBSD Corporate Networker's Guide
Book website: http://www.freebsd-corp-net-guide.com

On Fri, 18 Apr 2003, Ted Mittelstaedt wrote:

> Heh. :-) Although, I think he's just as horrified that _everyone_
> don't know about the FSF.

probably :)

> Not true in this case, let me explain.

well, yes, indeed, it would be fruitless in this particular case. i
obviously agree :)

> You go before the judge and say "I am a BSD software user and the
> original license granted by UCB has an advertising clause, UCB just
> told me that this clause is revoked, but the original grant of
> license doesen't permit them to do this. Thus they are in breech of
> contract and I'm suing for damages"
>
> Let's assume that you are right and the judge does indeed agree with
> you, that UCB is in breech of contract.
>
> Then he is going to ask what damages you suffer.
>
> In this case, because the retroactive modification is less
> restrictive, it is impossible for you to have suffered any
> damages.

absolutely.

> Thus you are awarded a judgement of $1, and the judge then proceeds
> to fine you $2000 for bringing a frivolous lawsuit into the court
> and wasting their time! :)

:)

> In summary, your never going to see such a lawsuit because nobody is
> going to risk the ire of a court over this. Thus, in this case, the
> copyright holder CAN make an "advertising clause" retroactive
> revocation simply because they say they can do so, and nobody is
> ever going to challenge them on it - and that is exactly what
> happened here.

ok. fair enough.

> It's like paying income taxes. Technically, the Income Tax is
> unconstitutional, thus illegal in theory.

according to my source: it was made legal by a constitutional
ammendment at the beginning of the 1900's.

> There is, unfortunately, a lot of law out there like this - stuff
> that's technically legal, but your going to go to jail for doing it
> anyway, if they catch you doing it, and stuff that's technically
> illegal, but in reality nobody is ever going to be prosecuted on it,

indeed.

> Anyway, besides all that, software copyright also carries with it
> this messy legal concept of "implied rights"

?

> Implied rights are what makes it illegal for you to buy a book then
> go make copies of it and give them out.

Well, i dont know about implied rights in the USA (i'm not from
there), but I think the implied rights you speak of in general are
those which most nations standardised and agreed to when they signed
the Berne convention on copyrights, after which i presume they were
all obligated to make sure the law of their land implied or specified
the obligations detailed in the Berne convention.

> Notice a book does not have anything in it explicitly prohibiting
> you from doing this, yet it's illegal because the implied rights of
> a book purchaser don't include this ability.

That would presumably be because the law of your land relating to
copyright law specifies a default copyright (the Berne convention does
anyway iirc).

Same as speed limits, just because there is no speed limit posted does
not mean there is no limit, because the law of your land probably
specifies a default limit. At least here in ireland, we have a
'national speed limit', on a road once you see the sign that means
'end of speed restriction' you are still limited to the national speed
limit of 60mph. (though on a motorway, the speed restriction can
actually /increase/ the limit, ie to 70mph here).

> But implied rights associated with software purchases is still very,
> very much being argued over.

Hmm.. dont know about USA, but over here it'd pretty much come under
default copyright - ie you cant, you could use it though. (if we
ignore the fact that usage of software by definition involves copying
and attribute that to fair use.).

> That is why software carries shrink wrap licenses, because the
> software houses don't feel at the current time that what makes up
> implied rights with a software purchase is still not clearly
> defined, so they spell it out. And even then most judges would
> reject a shrinkwrap license that significantly deviated from
> industry norms, and the software industry is fighting off efforts to
> have the entire shrinkwrap licenses tested in a high court anyway as
> they know it's shaky.

ok/

> The FSF spent a lot of money in legal fees and the lawyer that wrote
> the license had to put a lot of boilerplate in to justify the fees
> he was charging. ;-)

:)

i think they got it for free. Stallman wrote the original, i /guess/
eblen moglen wrote V2, i dont know.

> Law don't exist in a vacuum. Like I said earlier, everything around
> the law has to be considered as well.

of course.

> Reread the following from this letter:
>
> "Effective immediately, licensees and distributors are no longer
> required to include the acknowledgement within advertising
> materials. Accordingly, the foregoing paragraph of those BSD Unix
> files containing it is hereby deleted"
>
> This is a retroactive recindment because what they are talking about
> is defined in the letter here:
>
> "certain of the Berkeley Software Distribution ("BSD") sourcecode files "
>
> They don't explicitly say the word retroactive, but what they are
> talking about is existing source files, so it is retroactive.

ok, fair enough.

> It actually did, by implication. In short, if the BSD license
> wasn't considered by the court to be a "real" license, then AT&T
> would not have argued that BSD was infringing.

hmmm... i honestly think the argument was that they were infringing
the AT&T copyright by their /actions/.

> Otherwise AT&T would have merely argued that UCB was in breech of
> contract because they were misusing AT&T code, making the implicit
> assumption that ALL of BSD code was really AT&T code. Instead they
> singled out certain files, labeling them AT&T, and the rest of them
> they acknowledged were BSD-licensed.

indeed.

> eventually eclipse the point itself. Witness the finding of fact
> that Microsoft is a monopoly. That one thing is going to do more
> damage to them over the long term than the simple loss of the court
> case.

yes.

> That is like saying the Magna Carta is "just a piece of paper"

it is :)

it is what people could do as a consequence of it that is interesting.

> Paul, I can't believe that you don't realize the historical
> significance of the GPL and BSD software movements by now. I think
> your just baiting me.

i'm not. i do realise the significance of them. But that's another
debate :)

> What is going on now, today, is precedent setting for software and
> the results are going to dictate software for at least the next
> century. The licenses are a codification of fundamental directions
> open source is taking - discussion of them is discussion of open
> source's future.

yes. and BSD licence makes no claim to derivative works.

> >Code under the present BSD licence (in the generic sense) /allows/
> >this. How is it piracy?
> >
>
> No it does not, reread this section:
>
> * Redistribution and use in source and binary forms, with or without
> * modification, are permitted provided that the following conditions
> * are met:
> * 1. Redistributions of source code must retain the above copyright
>
>
> Quite obviously, keeping the copyright in there is crediting them,
> and the license itself says your supposed to do that!! Not doing it
> is scum, as I said.

ah, but now you enter the vagueries of derived work. The difference
here between GPL and BSD licences is that the GPL lays claim to
derived works, the BSD licence does not.

Hence the retention of the copyright for BSD works refers to the work
at hand. If you pull it apart and work on it enough that you in the
eyes of the law it is a seperate derived work, then no longer need you
keep that copyright (it makes no mention of derived works).

> >If someone wrote code and released it under the (current) BSD licence,
> >then they have chosen to allow people not to have to credit them.
> >
>
> No, not at all. They are expecting that the copyright be retained,
> and the copyright credits UCB. Quite often, it also makes further
> credits at the beginning of the license for the file in question,
> that is standard procedure actually.

AIUI, it comes down to derived works. BSD licence makes no
restrictions on deriving from BSD licenced code, and once derived it
is no longer the original work to which the copyright clause applies.

> As far as the question of would Apple have ever become a UNIX
> licensee if FreeBSD hadn't been available? Well I think the answer
> to that is that yes, they would have. Keep in mind that Apple
> ALREADY went down the road of developing their own OS, and still
> chose UNIX in the end.

Not really.

Apple developed their own OS. Jobs left and he and others developed a
Unix based OS. Apple kept developing their own OS. Jobs was brought
back to save apple, they decided to switch their OS efforts to build
on NextSTep (or however you capitalise it).

> Clearly, MacOS led to a dead end, and espically for Apple, admitting
> this was as damaging to their psyche as when Bill Gates bought 10%
> or whatever it was of Apple.

Yes, but that they would have chosen for FreeBSD had it not been BSD
licenced is not at all a given.

> No, this isn't the case either.
>
> The beauty of the BSD license is simply that it truly embodies the
> idea of real freedom in software licensing.

i agree.

> And as for the downside, that's an engineers phantom. In reality,
> the BSD code really doesen't permit people to build proprietary code
> around it, unless they are willing to only use the BSD code as a
> starting base.

indeed.

But it also allows source to /become/ proprietary. If the leading
developers of a project (or a significant subset) decide to go and
form a company and close the code, they can.

Much more difficult with GPL. (given a project with multiple
developers).

> The Apple case is a perfect example. Apple started out on FreeBSD
> 3.2,

Well, actually they started with Mach and a BSD4.3 userland :)

They then switched to FreeBSD kernel and userland.

> but FreeBSD has moved on to 5.X which is far and away more advanced
> than 3.2 was, espically in the area of multiprocessing. But since
> Apple made it proprietary, they cannot take advantage of any of the
> 5.X features without an expensive porting job, a job that gets
> harder and harder as future versions of FreeBSD diverge further away
> from the public FreeBSD code. Eventually, the Apple code really
> becomes indistinguishable from a proprietary project developed from
> scratch, and Apple eventually has no competitive advantage over it's
> commercial competitors by using BSD code.

I'd agree completely.

I would add though, that if FreeBSD had been GPL then any modifcations
Apple made would have /had/ to have been released back to the
community. (otherwise they could not redistribute it).

Note: I'm not making any claims GPL is better, i'm just pointing out
the differences.

> The only way to really take advantage of BSD code is to keep
> contributing your mods back to it, so that as people change the
> source code, they have to keep your mods updated since your mods are
> part of the codebase.

Indeed, and the GPL /obligates/ the distributor to do this, BSD
licence does not.

> >Oh come on. morals are intangible. Law and the licence's and contracts
> >built on civil law are mans attempt to try codify what is and is not
> >permissible in society and in dealings between members of that society
> >in general and in specific cases.
> >
>
> Law is founded on morality,

that is in essence the point i made above.

> and laws are only upheld by people that are moral people. Without
> morals, laws would have no meaning.

incorrect. for as long as society at large remains moral and works
within the framework of the laws set out, then law (and hence the
morals of the society) can be imposed upon those in society who do not
have those morals, and if they stil will not act within the law they
can be penalised.

> >conclusion to your arguments is that really the BSD licence should
> >be fixed, to codify these moral obligations into the licence.
> >
>
> No, absolutely not. This is the GPL approach

yes, indeed.

> - to codify every last little thing - because fundamentally the GPL
> approach is to not trust people to do the right thing.

we wouldnt need many of the laws and institutions that arise from them
if people always did the right thing.

the existence of law and institutions to advance, enforce and judge
members of society in accordance with those laws proves this fact.

> Yes there's moral obligations. But as to what to do if people break
> those moral obligations - well there are many options. One option
> is to simply ignore them.

Then do that.

> The BSD license is all about freedom. One of the hardest things
> about freedom is that unless you permit people to do things that you
> don't want them to do, your not really giving them freedom.

Absolutely.

> But just because you permit them to do it, doesen't mean you have to
> respect them for it, or agree with what they are doing.

fair enough.

> This is why it is so hypocritical for the FSF to insist that the GPL
> is "free software" It isn't. GPL code is very restrictive, that is
> not freedom.

No, it seeks to defend code from those who do not believe in freedom,
at the cost of removing certain freedoms.

NB: can you detail exactly why you have a problem with the GPL? ie
which freedom's does it remove that you have a problem with it?

> Legally there's now no claim on a BSD software user to honor the
> advertising clause in BSD. However, the BSD advertising clause is
> central to one of the tenants of the BSD license - giving credit
> back to UCB for BSD. Even without the advertising clause in there I
> argue that there's a history of crediting UCB on BSD software that
> anyone in the business is aware of, and if they use BSD they should
> be following.

Business people would not have a clue about this type of stuff at all!
:) They get their lawyers to look at the licence and act accordingly.

> Also it's just plain good manners to credit those that help you,

absolutely.

> and people that won't lift a finger to do so (ie: add 1 line in
> their manual crediting BSD) are scum.

not really. they're acting according to the licence. they are not
scum, its that /you/ think they're scum.

> And the GPL is incompatible with BSD, yet FreeBSD found a way to
> accomodate GPL code without adopting the GPL license on it's stuff.

indeed, quite probably they got permission from the author. (indeed
the author may primarily be a FreeBSD developer).

> Why is it so hard for the GPL to extend the same courtesy?

because its a licence. people extend courtesies.

The GPL could possibly have a clause to allow incorporation of GPL
code intio BSD work, but then obviously there'd be a huge hole in the
GPL and it would be useless.

> Don't they have enough brains to figure out how to do this in
> GPL-land? Are the BSD people smarter than them because they worked
> out how to do it?

Ok, this is silly.

You have to acknowledge that the GPL is worded in such a way
as to advance a specific goal.

You must also acknowledge that having some kind "yes, you can use it
in BSD licenced code" must completely invalidate the GPL.

no?

> This incompatability is NOT a technical obstacle, there's ways
> around it. It's a political obstacle is what is is - because GPL
> wants to take BSD code and apply the GPL license to it, for no other
> reason than they don't happen to like any other way of licensing
> code than GPL, to the point that they don't even want to admit that
> other licenses even exist.

i dont that's true now. :)

give me examples of BSD licenced code being "sucked in" to GPL land.
Surely you can recognise that what is not important is whether BSD
code is sucked in to GPL projects, but that the BSD code continues to
be developed? And what if the developers of the BSD code have left it,
no longer maintained or active, and some new developers come along
work on it, improve it and subsume it into a GPL project.

That is surely the lesser of several "evils"? namely: code lies dead,
unmaintained forever, code is subsumed into GPL, code is subsumed into
completely closed proprietary work.

Eg, look at gated. Is that better than code being subsumed into a GPL
project?

Now, if the BSD project remains active, then what's the harm? If the
BSD licenced project remains the driving force behind that code then
who care if a GPL project uses some of it? (and presumably they will
dual-licence any modifications to the original BSD licenced code back,
as it is in their interest).

> Even better - if you don't want to assign copyright to FSF - simply
> don't use GPL!

Why?

> A license is only worth anything if the copyright holder is willing
> to defend it. Otherwise it's only so much air blowing out of
> someone's ass.

Not really, it still carries moral weight, even if ultimately you
could not afford to defend it.

> For copyright holders that don't assign copyright to FSF, there is
> little point to using the GPL license since they are going to defend
> their copyright on their code themselves. But to hear or read the
> average GPL advocate, they won't say this at all.

Why should they? Its fairly obvious that if you had a user who
unfortunately was morally bankrupt enough to disregard your licence
you would need to lawyers to bring the weight of the law to bear.

But that might not happen. And you still need a licence - and if one
feels the GPL is a suitable licence then fine.

> >That is not their insistence though, is it?
> >
>
> Yes it is. See:
>
> http://www.fsf.org/licenses/gpl-faq.html#LinkingWithGPL

Ok, but you have an option - dont link.

> Yes they have. See the GPL FAQ.

They claim that if you /want/ to link against GPL code then your code
must be GPL too.

They do /not/ claim that if you /do/ then your code automatically
/becomes/ GPL licenced.

> This is just an assuption that all people are stupid, plain and
> simple.

not all, but enough of them.

> The idea that a commercial developer is going to just run
> across BSD code and merrily include it into their proprietary stuff
> without investigating the BSD community and asking questions of that
> community is fiction.

But it isnt though.

There are lots of developers who havnt a clue about BSD or GPL or the
morals or history behind them, and they're perfectly willing to take
code and incorporate it into their own if the licence so allows.

> They are going to investigate the legal context surrounding the
> code's license and the will certainly find out about the moral
> obligations then.

Not really, they'll give the licence to a copyright lawyer and ask his
opinion.

> That is true and it is something that I overlooked. But there's a
> difference between what your talking about and what I'm talking
> about.

well, there's a difference between our standpoints. Mine would be:

- if you act according to the terms of the licence then you are moral

yours is:

- if you act according to the terms of the licence, but not according
to the moral precedent of the community that traditionally uses that
licence then you are not moral.

ie i believe:

- the licence clarifies how one may use that code (and how moral
people will).

you dont.

> It's another to take an entire BSD program, or an entire BSD
> sourcefile, and replace the BSD copyright with the GPL copyright
> just because you don't like BSD.

that would not be right. i agree. it'd also be a licence violation
AIUI.

> Because there is no need for this in a distribution such as an
> operating system, (ie: Linux) or a large complex program which uses
> some smaller utilities in it. And there's also no need for it with
> BSD-licensed libraries, except that the FSF decided to arbitrairly
> say that if you link in a library, it automatically falls under GPL

they dont say that.

> - something that is such a bald grab that you didn't even believe it
> when I wrote you with it, so I had to send you back to the GPL FAQ.

it still doesnt say that.

no one is forced to link to a GPL library.

> >I pity you that you can get so worked up about an inanimate legal
> >text.
>
> Are you saying your education was so lacking that none of your
> teachers ever explained that the pen is mightier than the
> sword?

it is indeed.

and my statement is wrong, you are getting worked up about issues
outside of the legal text. you are failling to recognise that the
legal texts are all that matter, imo. you're getting up all worked up
about vagueries surrounding BSD and GPL, worst of all i get the
feeling you have a problem with the entire community of Unix/'free
software' users who agree with the GPL by tarring them all with the
same brush.

> More people have been killed over history as a result of ideas
> written down on inanimate text than any other reason.

absolutely, yes.

> And in any case, your being a hypocrite since you are obviously just
> as worked up as I am, otherwise you wouldn't have written such a
> long reply to my mail.

Well, i get worked up about BSD fanatics who like to diss the GPL at
every opportunity. :)

IMO, we're all part of the same community, that the GPL and BSD are
trying to do the same thing but the GPL tries to do it slightly
differently.

> I was never arguing that they don't have a choice. If someone wants
> to use GPL for THEIR stuff, more power to them. I am objecting that
> the community of GPL proponents basically never miss an opportunity
> to snub BSD and flat out lie about what BSD is all about,

aha. And what is that lie?

> when they really owe their existence to BSD code activities that
> preceded the creation of GPL.

But i dont think any true GPL user/believer would say otherwise.
There's tonnes of BSD licenced (and very BSD licence like licenced
code) that runs on my linux boxes (squid, XFree, BIND, etc..), and i'm
grateful for them (and i'm sure all other true GPL believers are too).

FreeBSD though does not predate the linux kernel. Also, the BSD
community has long been dependent on gcc has it not?

> You (I assume) inadvertantly did this when you posted your comments
> about BSD.

how?

> I am just glad that I was able to educate you as to what the truth
> is about the BSD advertising clause is, so you don't continue to
> repeat the wrong GPL propaganda about BSD you had stated in the
> Zebra mailing list.

It wasnt propaganda though. :)

Ok, so the advertise clause is retroactive. :)

> Ted Mittelstaedt tedm@toybox.placo.com

regards,
--
Paul Jakma Sys Admin Alphyra
paulj@alphyra.ie
Warning: /never/ send email to spam@dishone.st or trap@dishone.st