Re: Why is Nvidia given GPL'd code to use in non-free drivers?

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[walt]

Henning P. Schmiedehausen wrote:

> "Dimitrie O. Paun"  writes:
>
>
> >  software. For how much money MS has, what have they innovated?
>
> Face it. Microsoft Software is what made the breakthrough to really
> put a powerful machine in every home and allow the 2-5% of the owner
> base which are Linux users to get really cheap commodity hardware.

I think you are correct in this.  The dinosaurs had their day --
we all got our DNA from them -- and now they are gone.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Adrian Bunk]

On Sun, Jan 05, 2003 at 01:46:18PM -0800, walt wrote:

>...  The dinosaurs had their day --
> we all got our DNA from them -- and now they are gone.

(completely OT)

This is wrong. We didn't got our DNA from the dinosaurs. The split 
in the evolution between the ancestors of the dinosaurs and our 
ancestors happened many million years before the great time of the 
dinosaurs.

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed

Re: Why is Nvidia given GPL'd code to use in closed source drivers?

[Bill Huey]

On Thu, Jan 02, 2003 at 12:31:13AM +0000, Paul Jakma wrote:
> The NVidia shim makes use of several kernel subsystems, the PCI
> device layer, the VM, the module system (well really, the kernel
> makes of use of the functions the module provides :) ), IRQ
> subsystem, the VFS, etc.. These systems are rather large bodies of
> code - without which the NVidia kernel driver could not work.

Well, no, look at the "nm" dump of the object file. It's got a lot of
proprietary code that came from what looks like commerical libraries
that they don't own. Back when they wrote the original drive, the GPL
equivalents of DRM, AGP, etc... sucked so they had to write their own
stuff just to get anything basic working.

> driver is not a derivative work, and hence it seems to me the NVidia 
> driver is technically in material breach of GPL.

Their portability layer wraps the low level calls into their own
terminology and portability API. It's fairly outside of the linux kernel
itself, internally the object file looks very Win32ish.

Obviously a GPL rewrite of this would entail a lot of replicated effort
and would also depend on things that are incomplete, non-existent and
don't have a lot direct interest from the GPL community. 3D isn't a hot
commodity in Linux, FreeBSD unlike with dedicated SGI machines (although
faded).

It's a very practical solution to do it this way.

> So I am not quite sure on what basis one could argue the NVidia 
> 
> You seem to be basing your opinion on:
> 
>  "the nvidia driver uses only the standard interfaces to hook into
>  the Linux kernel"
> 
> How are the standard interfaces not covered by the GPL? 

All I saw where kernel header files include in the sources, nothing
more. They have to support multipule architecture and OSes so keeping
this stuff outside of the driver is a good thing. The GPL-ly stuff is
publically available as source files.
 
> I know Linus' has often posted to l-k that he doesnt care about
> binary only modules as long as they stick to the exported interfaces.  
> However, are all the kernel developers agreed on this? And if so, can
> this exception be formalised and put into the COPYING file? If not, 
> then is NVidia not in breach of the kernel's licence?

I'd rather have the experts do it at NVidia, than a half completed open
source implementation that isn't terribly optimized.

Matrix multiplies, T&L, etc... communication between user and kernel
space that provides this to the OpenGL libraries are all exotic. I'm glad
that nobody has to deal with this stuff directly and that a vendor is
willing to provide support for it.

bill

Re: Why is Nvidia given GPL'd code to use in closed source drivers?

[Paul Jakma]

On Wed, 1 Jan 2003, Bill Huey wrote:

> On Thu, Jan 02, 2003 at 12:31:13AM +0000, Paul Jakma wrote:

> > subsystem, the VFS, etc.. These systems are rather large bodies of
> > code - without which the NVidia kernel driver could not work.
> 
> Well, no, look at the "nm" dump of the object file. It's got a lot of
> proprietary code 

indeed. that doesnt change the fact that this large body of NVidia
specific code still must make use of large parts of linux code
(through function calls).

> It's a very practical solution to do it this way.

yes, but the legalities of it are rather grey.

> > How are the standard interfaces not covered by the GPL? 
> 
> All I saw where kernel header files include in the sources, nothing
> more. 

indeed, and if that were the only issue it would be clear there is no 
issue. however, it must make use of linux code at runtime through 
function calls - as linux makes use of the NVidia proprietary code by 
calling the functions it provides.

> I'd rather have the experts do it at NVidia, than a half completed
> open source implementation that isn't terribly optimized.

I run systems that use many GPL and fully open drivers that are quite
well optimised. Some of these drivers were written by the vendor's
"experts" and are distributed seperately - still GPL though.  
Sometimes one has a choice between drivers written by the vendor and
drivers written by (non-expert???) "community" authors, and often one
finds the vendor driver is the one that isn't terribly optimised.

> Matrix multiplies, T&L, etc...

none of this stuff is done in kernel (least it shouldnt be). Its done 
in user-space libraries.

The XFree licence allows binary only modules, indeed XFree 4 was 
designed to make distribution of (possibly binary) modules as easy as 
possible.

There isnt that much magic the NVidia kernel modules ought to be 
doing really.

> communication between user and kernel space that provides this to
> the OpenGL libraries are all exotic. I'm glad that nobody has to
> deal with this stuff directly and that a vendor is willing to
> provide support for it.

aha.. yes, all that complicated hardware stuff - you dont really want 
those linux kernel amatuers writing that.

> bill

regards,
-- 
Paul Jakma	paul@clubi.ie	paul@jakma.org	Key ID: 64A2FF6A
	warning: do not ever send email to spam@dishone.st
Fortune:
The system will be down for 10 days for preventive maintenance.

Re: Why is Nvidia given GPL'd code to use in closed source drivers?

[Bill Huey]

On Thu, Jan 02, 2003 at 02:57:48AM +0000, Paul Jakma wrote:
> yes, but the legalities of it are rather grey.

It didn't seem that bad to me, it was all pretty abstracted outside of
their code. The glue layer to their object file is GPLed and therefore
public so that should be fine from what I can see.

> indeed, and if that were the only issue it would be clear there is no 
> issue. however, it must make use of linux code at runtime through 
> function calls - as linux makes use of the NVidia proprietary code by 
> calling the functions it provides.

Like what ? PCI IO poking functions ? Things that do mmap() trickery ?
That's pretty freaking basic. There wasn't anything terribly invasive
about the driver and source that I saw.

> Sometimes one has a choice between drivers written by the vendor and
> drivers written by (non-expert???) "community" authors, and often one
> finds the vendor driver is the one that isn't terribly optimised.

But this is computationally critical 3D. I mean, what kind of 3D vendor
would intentionally let something like that slide on x86 platforms ?

> > Matrix multiplies, T&L, etc...
> 
> none of this stuff is done in kernel (least it shouldnt be). Its done 
> in user-space libraries.

That stuff is done in hardware these days, not software. I mean, how would
anybody know what they're using. Why replicate that volume of functionality
when it already works well.

It simply doesn't make sense. I'm sure when decent AGP/DRM support is in
place they can start removing that stuff out of the Linux binary and
then make more of that publically available.

There motivations where to simply protect themselve by not releasing
proprietary code.

> The XFree licence allows binary only modules, indeed XFree 4 was 
> designed to make distribution of (possibly binary) modules as easy as 
> possible.
> 
> There isnt that much magic the NVidia kernel modules ought to be 
> doing really.

Notification of event completion from the (just guessing) who knows what
opcode operations the chip is doing, fast draw context switching, who knows.
These things are starting to look like FPU coprocessors, circa 1990, these
days.

Different hardware has differing needs. If it's pretty freaking exotic, then
let it to those folks handle it and the glue layer to userspace. It's not
like folks in GPL community write entire 3D frameworks for this casually.

High performance 3D is a Linux priority at this time. No real games or
heavy 3D apps that use crazy chips stuff...

> > communication between user and kernel space that provides this to
> > the OpenGL libraries are all exotic. I'm glad that nobody has to
> > deal with this stuff directly and that a vendor is willing to
> > provide support for it.
> 
> aha.. yes, all that complicated hardware stuff - you dont really want 
> those linux kernel amatuers writing that.

Well, having a generic kernel person, regardless of who they are, messing
with 3d chips and interfacing it with their OpenGL libs isn't a light topic.
This crap is heavy. So yes, its a good thing they've done this. What the
hell do you think this is ? an Ethernet driver ?

bill

Re: Why is Nvidia given GPL'd code to use in closed source drivers?

[Mark Mielke]

GPL aside (it could be argued forever...):

I regularly use several kernel modules that provide a GPL component that
interfaces the module to the kernel, and a closed source object file that
is dynamically loaded as a kernel module at run time.

If I did not have these modules, I would not be able to use Linux as my
host operating system.

So... to those (Hell.Surfers especially it seems) who believe that they
are doing good by making a scene... realize that while you may succeed
in improving the integrity of the GPL, you will also succeed in convincing
companies that it may be more expensive than it is worth to provide their
hardware or software for Linux. That may mean that I will be forced to
stop using Linux at work, and possibly forced to stop using Linux at home.

Perhaps I am a minority. Are you willing to bet the future of Linux on it?

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in closed source drivers?

[Richard Stallman]

    I regularly use several kernel modules that provide a GPL component that
    interfaces the module to the kernel, and a closed source object file that
    is dynamically loaded as a kernel module at run time.

    If I did not have these modules, I would not be able to use Linux as my
    host operating system.

Many enthusiasts the "Linux" operating system take the popularity of
the system (or of the kernel, Linux) as the supreme goal; but why
should the popularity of any one operating system or program be so
important?  That isn't what really matters.

We developed the GNU system for the sake of freedom, and freedom is
what really matters.  The GNU/Linux system today is important because
it offers a road to freedom.  But it doesn't guarantee you will arrive
there.  If you use non-free drivers, you go just part way along the
road and never arrive at freedom.  That defeats the purpose.  To
achieve freedom, we need to insist on free drivers (and free
applications).

Erik Andersen <andersen@codepoet.org> wrote:

    If nvidia provided non-functional GPL
    source code with all the proprietary 3rd party bits ripped out, 
    I would expect a hoard of developers would jump at the chance to
    fixup the non-functional mess, clean it up, reimplement all the
    missing proprietary bits.  I'd bet you $20 US we could have a
    functional driver within 2 weeks.

If NVidia cooperates with us this much, we should certainly pick up
the ball from there, and I am sure we will manage to go the rest of
the way.  But don't bet on 2 weeks.  Softare always takes twice as
long as you expect ;-).  If it takes a whole month month to be able to
use NVidia hardware in freedom, I won't complain about the delay.

But we could make do with even less cooperation than that.  If they
just provide the necessary specs to a person who wants to extend the
free drivers that exist, that would be sufficient.  It might take more
than 4 weeks to write the code, but surely not more than a few months.

Re: Why is Nvidia given GPL'd code to use in closed source drivers?

[Mark Mielke]

On Thu, Jan 02, 2003 at 10:32:30PM -0500, Richard Stallman wrote:
> We developed the GNU system for the sake of freedom, and freedom is
> what really matters.  The GNU/Linux system today is important because
> it offers a road to freedom.  But it doesn't guarantee you will arrive
> ...

You don't seem to mind the fact that my freedom to use Linux would be
hampered if you successfully prove that closed source modules for
Linux are illegal.

If open source is so good, companies with closed source products will
change. Have some faith in your own set of ideals and stop trying to
jam it down other peoples' throats.

Here is another factor to consider. Copyrights and patents are all nice
and dandy, but unless you can make a court case that proves that you,
or the 'violated party' is *LOSING MONEY*, the case would likely be
thrown out of court. What money are you losing by nVidia using closed
source modules for their proprietary hardware? nVidia believes it is
protecting its own interests. You believe that they don't deserve this
right, and that any company that doesn't agree with you doesn't deserve
to use Linux.

I can see all the freedom in the air. It is overpowering.

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Richard Stallman]

    You don't seem to mind the fact that my freedom to use Linux would be
    hampered if you successfully prove that [non-free] modules for
    Linux are illegal.

I'm not trying to prove this--as I see it, Linus gave permission for
them, which means they are legal.  I regret his decision to do this,
but I cannot change it.

But let's suppose that that were changed.  It would not affect your
"freedom" to use Linux (and GNU/Linux), only whether it runs on a
certain computer.  It is true that this might mean a practical
sacrifice--you might have to get a different kind of computer, for
instance.  I don't see that as a horrible thing.  We look for
computers that work with free drivers; you can too.

You don't really have freedom now, if you need a non-free module.  In
the long run, your best chance of being able to use a fully free
GNU/Linux system on the hardware you use is if we stand firm together
for the freedom of the system.

    If open source is so good, companies with closed source products will
    change.

I don't support the open source movement, but I know what they say
about this.  They say that open source usually leads to more powerful
and reliable software.  Nothing assures us that will persuade all
companies to adopt the practice.  You have simplified their position
to a point where they would not recognize it.

You seem to be saying that we should sit back and let these inevitable
forces either convince all companies to make software free--or not.
If we had such a passive attitude, no free system would exist.
GNU/Linux exists because of people who were willing to work to have
freedom.  Freedom does not yet prevail, and we have plenty more work
to do to make that happen.  And after we fully have freedom, we will
still have to work, to make sure we don't lose it.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Larry McVoy]

Long rant but it's a worthwhile read if you want to know why I don't
agree with Richard.  

> You seem to be saying that we should sit back and let these inevitable
> forces either convince all companies to make software free--or not.
> If we had such a passive attitude, no free system would exist.
> GNU/Linux exists because of people who were willing to work to have
> freedom.  Freedom does not yet prevail, and we have plenty more work
> to do to make that happen.  And after we fully have freedom, we will
> still have to work, to make sure we don't lose it.

The problem with your point of view is that you are assuming that somehow
progress will continue to be made once you have that freedom.  Let's just
look at that for a minute and see if that makes sense.

Postulate that all the software in the world is GPLed.  All of it.  That's 
your goal as far as I can make out, but let's not argue about if it is or
is not, it doesn't matter.

Anyone who wants to build on that software can, there is almost perfect
code reuse.  Again, something I think you want, certainly a nice idea.

Because all the software is freely available, this sets an upper bound
on how much any company can charge for it.  If the amount they charge
for gathering it up and making a distribution, for example, is low
enough that other people look at it and think that's too little money
for that much work, then their prices will hold.  On the other hand,
if they are charging twice that, another company can spring up which
grabs the software and sells it for the lower price, a price low enough
that no cheaper company can come in.  Obviously, the first company either
drops their prices or goes out of business.

I think this too is what you want, it seems great, people are paying a
small amount of money to get the software, a much, much smaller amount
than they would be paying if the software were closed.  Case in point:
Microsoft.  Nobody would be paying what they charge if they didn't have
to do so.  But people would certainly pay $20 or $50 for a CD full of
Windows + Office + whatever.  But not much more and they'd pay for one CD
and then install it many many times, so the effective revenue per machine
would be less than one cent in an organization of any size.

So what's wrong with that picture, it seems fantastic?  A world where we
can all share each other's work, the prices are low, anyone can tinker,
anyone can package, sounds great.  

The problem is that the amount of money being generated is very small
compared to what software companies get under the closed source model.
So what?  What's wrong with that?  Well, under that model, none of the
software companies can afford to pay for any development out of the
distribution revenue, if they were charging more than it took to pay for
the people to build the distribution then someone would undercut them,
there is nothing they can do to prevent that.  That same argument works
for the support model or anything else.  It doesn't matter what model
you pick, if you charge more than it costs to do the work plus a very
slim profit margin, that presents an attractive opportunity for someone
else to set up shop.

We can argue about this until the cows come up but it's simple economics.
If there is no barrier to entry and a supplier is charging more it
costs, a cheaper supplier will enter the market and force the price down.
Even the most green MBA understands this and I don't think I need to tell
you that the VC's all understand this.  For the sake of discussion, let's
assume that you agree with that statement (if you don't, don't bother
to argue with me, I'll ignore you, I'm not here to teach basic economics).

So we've established that in an all free world, even though some money
will change hands, it can't be significantly more than what it costs 
to perform whatever service is being provided.  In other words, there 
is no extra money.

Is that a problem?  I think so.  If you look at the history of the free
software movement, it has been a history of imitation rather than one of
innovation (sorry to sound like Bill Gates but he has a point here).
Almost everything done in the free software world is a rewrite of
something that already existed.  The GNU folks have made it clear for
years that what they wanted are free versions of the applications they
consider to be useful.  They don't spend much time talking about anything
innovative, they talk about filling in the gaps where there is no free
version of Word or whatever.  Over and over, RMS says "you would better
to not use $APP but instead dedicate some time to rewriting $APP so we
have a free version".  That's not innovation, that's imitation.

Leaving aside the inevitable argument about whether or not the free
software world is or is not innovative, let's look at what it takes
to produce new things.  The problem is that none of us have a real
crystal ball.  We don't know which ideas will take hold in the market
and which won't.  We can guess and maybe get lucky, but in general the
guesses are wrong much more than they are right.  Look at the history
of startups.  With all the screening that VC's do, all the due diligence,
we still have failures of at least 9/10 and these days more like 99/100.
For every Ebay or Google there are hundreds of startups which started
about the same time as Ebay or Google but are are long and forgotten.

If we look at the entire software development world as one big system,
what history shows us is that the vast majority of the effort is wasted,
only 1% of it succeeds (1%, 10%, pick your number, the vast majority of
it fails).

Let's say it is 10% of it that succeeds.  Under the world RMS is proposing,
not only is there no money to pay for the startup costs of that 10%, there
is no money to pay for the 90% that doesn't succeed.  Not from within the
system, the only way that money can exist is if it is from people outside
of the software world, every penny in the software world is spoken for.

Let's say that I'm wrong, we can come up with enough extra money to pay
for the 10% which is going to succeed.  The problem is that we don't
know in advance wich 10% will succeed, which means that we are really
only funding 1/10th of the startups.  Which slows down innovation to
1/10th the speed.

I don't know about you, but that's not a world I want to live in.  Google
wouldn't exist in Richard's world, I know that for a fact.  Sergey and
Larry would have gone into some other field, they are ambitious people.
Ditto for Ebay, Amazon, and any number of other companies which have
become institutions.

You may or may not still be reading, you may or may not agree, but my
view is that Richard's proposed world is a very bleak sort of place.
There won't be any companies coming out with new ideas to copy, there
won't be the frenetic pace of innovation, it will be a sort of gray
place, like selling washing machines.  I don't know about you, but I
like the world we live in.  Yeah, the fact that Microsoft has the money
really sucks but if that's the price I have to pay to get things like
Google and Ebay and Amazon and Shutterfly and <fill in your favorite
new software company here>, that's OK with me.  You might want think
about whether you would trade all of that for Richard's utopian world.
Maybe you would, I wouldn't and I don't think very many other people
would either.
-- 
---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Dimitrie O. Paun]

On January 3, 2003 04:26 pm, Larry McVoy wrote:
> If we look at the entire software development world as one big system,
> what history shows us is that the vast majority of the effort is wasted,
> only 1% of it succeeds (1%, 10%, pick your number, the vast majority of
> it fails).

Larry, in general I agree with you, but there are a number of things
that you've touched upon which I feel need a bit of a debate.

1. Immitation vs. Inovation
   The free software world has started from 0 relatively recently.
   It is to be expected that we first fill in the gaps that are
   known to be useful, before we start experimenting. I don't think
   this is a major case against us. As for money that is fueling
   research, this is not quite so. It works well in telecom, it
   has a poor record in drug companies, it has a lousy record in
   software. For how much money MS has, what have they innovated?

2. Research & Money
   It seems to me that history has proven that the open ways of the
   scientific community has generated _order_ of magnitude more
   research than the closed business model. Not to mention mathematics.
   Software is a lot more like mathematics than a construction company.
   What have all this money produced in software that's we didn't do
   for free? You see, it is hard for me to say this since I am a
   'rightist': I believe in the free market, capitalism, etc. But it
   certainly looks to me that software works out better in the open.
   And if that's the case, how do we avoid the bleak world you're painting?
   In all honesty, I don't know. But it might not be as ba as you make it
   to be. It might end up like science today: it doesn't pay to be a
   scientist, but I guess it's fun so people do it. Yeah, you will not have
   as much money to test all sort of silly ideas (the 99% that fails), but
   tell me, what do we get out of that? In a free software world, if there
   is a need for something, it will get done. Maybe not now, but in 6 month,
   or 1 year. Big deal. We don't need a Big Brother to invest large amounts
   of money to convince us that we _need_ his useless program.

3. All free software
   This, you know, we'll never happen. We have free software mostly in the
   world of consumer software, and really, it's only a handful of companies
   (MS mostly) that will lose if this turns free. But the vast majority of
   sofware is developed in house, for a client, and that can stay proprietary.
   We're talking mostly infrastructure (OS, X, Office) that need to be free.
   And we all benefit from it. The software that the banks use can stay
   proprietary and continue being shit as it is now. There's no inovation going
   on there, just mony thrown out the window.

-- 
Dimi.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Henning P. Schmiedehausen]

"Dimitrie O. Paun" <dpaun@rogers.com> writes:

>   software. For how much money MS has, what have they innovated?

Without Microsoft, there wouldn't be 2,4 GHz 32/64 bit microcomputers
with 512 megabytes of main memory, 120 gigabytes of hard disk space
and 1600x1200 pixels 32 bit resolution and 3d real time capability for
< $1000 on sale at your local discount store. Simply because there
wouldn't be a market for this.

Face it. Microsoft Software is, what made the breakthrough to really
put a powerful machine in every home and allow the 2-5% of the owner
base which are Linux users to get really cheap commodity hardware.

Not IBM did this. Not Commodore. Not Apple. Not Linux.

	Regards
		Henning

-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     hps@intermeta.de

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   info@intermeta.de
D-91054 Buckenhof     Fax.: 09131 / 50654-20   

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Eric Ortega]

On Sun, Jan 05, 2003 at 09:24:47PM +0000, Henning P. Schmiedehausen wrote:
> "Dimitrie O. Paun" <dpaun@rogers.com> writes:
> 
> >   software. For how much money MS has, what have they innovated?
> 
> Without Microsoft, there wouldn't be 2,4 GHz 32/64 bit microcomputers
> with 512 megabytes of main memory, ...

Take this half-baked crap elsewhere.  I don't lurk on this list for this
drivel.

Find some statistics or facts disproving that someone else would have
filled this "void" and take your argument to someone who cares.

HAND.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Ian Molton]

On Sun, 5 Jan 2003 21:24:47 +0000 (UTC)
"Henning P. Schmiedehausen" <hps@intermeta.de> wrote:

> Without Microsoft, there wouldn't be 2,4 GHz 32/64 bit microcomputers
> with 512 megabytes of main memory, 120 gigabytes of hard disk space
> and 1600x1200 pixels 32 bit resolution

> Face it. Microsoft Software is, what made the breakthrough to really
> put a powerful machine in every home and allow the 2-5% of the owner
> base which are Linux users to get really cheap commodity hardware.

<rant>

BULLSHIT. it may have happened that way, but if M$ didnt do it it would
STILL have happened.

I was using 1600x1200 desktops on my Acorn RISC PC about a year, perhaps
2 years, before that sort of resolution was available on the majority of
PCs.

Before that I was using an ARM 2 and 3 based A410 which literally blew
away all desktops available at the time (286, 386) in terms of
performance.

Even today, My A410 can play a game of DOOM at nearly 486 speeds, in
truecolour!

This hardware had a british designed processor (ARM) and no INTEL or
such chips in sight. Windows never ran on it.

Yet my 15 year old A410/1 remains one of the most useful machines in my
house today, alongside an AthlonXP1800+ which, 15 years later, *STILL*
doesnt have a DTP solution thats as easy to use as Ovation Pro or
Impression Publisher were on the A410.

Only a week ago I used the A410 to draw a PCB layout for an audio
amplifier, using software which was supplied IN THE MACHINES ROM, along
with the OS.

I shall use the A410 to print the layout at 1200dpi onto transparent
film, too.

Not bad for free software on a 15 year old machine

Oh, and the same 15 year old machine can manage a (doublescanned)
1600x600 screenmode too, using dual ported RAM, at negligible speed
penalty.

So dont give me this crap that only M$ could provide the sub 500ukp PC
with bells and whitles.

All M$ have done is force us all to use turbocharged versions of a
jumped up washinmachine control microprocessor, and repeatedly upgrade
it to keep up.

No thanks.

</rant>

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Henning P. Schmiedehausen]

Ian Molton <spyro@f2s.com> writes:

>On Sun, 5 Jan 2003 21:24:47 +0000 (UTC)
>"Henning P. Schmiedehausen" <hps@intermeta.de> wrote:

>> Without Microsoft, there wouldn't be 2,4 GHz 32/64 bit microcomputers
>> with 512 megabytes of main memory, 120 gigabytes of hard disk space
>> and 1600x1200 pixels 32 bit resolution

>> Face it. Microsoft Software is, what made the breakthrough to really
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> put a powerful machine in every home and allow the 2-5% of the owner
   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> base which are Linux users to get really cheap commodity hardware.

><rant>

>BULLSHIT. it may have happened that way, but if M$ didnt do it it would
>STILL have happened.

You might want to reread the underlined sentence. I have no doubts
that there have been better, faster and easier to use machines long
before the PC platform and Microsoft software.

I actually owned an Acorn BBC a long time ago and it was a sweet
little box. Just like the machine after it (which I didn't buy because
I was already hooked on Amiga by then). But how many boxes did they
sell? Did they achieve the "computer as a commodity" goal?

And I didn't write at all that Microsoft gave you that computer. I said
"they made it possible". Most surely you got your computer from Dell, HP,
IBM or that computer store around the corner.

	Regards
		Henning
-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     hps@intermeta.de

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   info@intermeta.de
D-91054 Buckenhof     Fax.: 09131 / 50654-20   

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Shane R. Stixrud]

Larry,

I know your post was directed at RMS, however I felt inspired to respond :) 
Perhaps there is an alternate outcome to the future you see, bare with me.

On Fri, 3 Jan 2003, Larry McVoy wrote:
> The problem with your point of view is that you are assuming that somehow
> progress will continue to be made once you have that freedom.  Let's just
> look at that for a minute and see if that makes sense.
> 
> Postulate that all the software in the world is GPLed.  All of it.  That's 
> your goal as far as I can make out, but let's not argue about if it is or
> is not, it doesn't matter.
> 
> Anyone who wants to build on that software can, there is almost perfect
> code reuse.  Again, something I think you want, certainly a nice idea.

Add to this:

No Software patents or greatly shorting the length of 
software patents.

All hardware specs are published.

> 
> Because all the software is freely available, this sets an upper bound
> on how much any company can charge for it.  If the amount they charge
> for gathering it up and making a distribution, for example, is low
> enough that other people look at it and think that's too little money
> for that much work, then their prices will hold.  On the other hand,
> if they are charging twice that, another company can spring up which
> grabs the software and sells it for the lower price, a price low enough
> that no cheaper company can come in.  Obviously, the first company either
> drops their prices or goes out of business.

You seem to view software like a chair or a box of legos?  Something that 
is designed, implemented and then sold.  This is the predominant view, 
perhaps software can be thought of as simply: 
Ideas + Skill + Time + Natural Resources.

[Snip]
> If there is no barrier to entry and a supplier is charging more it
> costs, a cheaper supplier will enter the market and force the price down.
> Even the most green MBA understands this and I don't think I need to tell
> you that the VC's all understand this.  For the sake of discussion, let's
> assume that you agree with that statement (if you don't, don't bother
> to argue with me, I'll ignore you, I'm not here to teach basic economics).
> 
> So we've established that in an all free world, even though some money
> will change hands, it can't be significantly more than what it costs 
> to perform whatever service is being provided.  In other words, there 
> is no extra money.
> 

Am I correct in thinking that your two primary goals in relation to 
software development are?:

1) For _individuals_ to make a good/great living. 
2) Software is constantly improved/evolves and that new and innovative 
ideas have fertile ground to develop in.

If so, I believe we are in complete agreement thus far.

> Leaving aside the inevitable argument about whether or not the free
> software world is or is not innovative, let's look at what it takes
> to produce new things.  The problem is that none of us have a real
> crystal ball.  We don't know which ideas will take hold in the market
> and which won't.  We can guess and maybe get lucky, but in general the
> guesses are wrong much more than they are right.  Look at the history
> of startups.  With all the screening that VC's do, all the due diligence,
> we still have failures of at least 9/10 and these days more like 99/100.
> For every Ebay or Google there are hundreds of startups which started
> about the same time as Ebay or Google but are are long and forgotten.
> 
[Snip]

In a world where all software is "Free Software" does it not make more 
sense to view the "costs" of software development as being equal to 
ideas+time+skill+available resources?

You must have the idea/problem/design to begin a software project.

You must have the time and skill available to complete the design or solve 
the problem.

Reusable code decreases the amount of time and perhaps skill required to 
complete software tasks.

I would argue that most of the problems you see in a 100% "Free 
Software" World come from a large Enterprise/corporate mind set, where large 
Sums of VC money is required to fund projects.  I see a 
possible future where software development is seen much like getting your 
roof repaired, or adding a new room on to your house.  It's 
labor+materials.  With the advantage that in software world materials are 
reusable, the tools are plentiful and affordable (well most anyways).  In 
this World it's the software developers who are sought after... not the 
VC money.  

Could this Free Software world support Microsoft or it's ilk? No.  Could 
it feed the families of millions of software developers?  I believe so. 

Shane R. Stixrud

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Larry McVoy]

Another long rant worth reading in my opinion...  Explains how Google
helped the birth of our son.

On Fri, Jan 03, 2003 at 04:55:22PM -0800, Shane R. Stixrud wrote:
> I would argue that most of the problems you see in a 100% "Free 
> Software" World come from a large Enterprise/corporate mind set, where large 
> Sums of VC money is required to fund projects.  I see a 
> possible future where software development is seen much like getting your 
> roof repaired, or adding a new room on to your house.  It's 
> labor+materials.  

And therein lies the problem.  You and RMS agree, in his world it is
like getting your roof repaired or a new washing machine installed.
That's not the world in which we currently live.  Our world is much
more exciting than that.  Every day there is a new thing, it's not
a roof, it's not a washing machine, it's Ebay.  Nothing like it 
existed ever before.  It's those new things which change our lives
for the better that are cool.  In the world that you are describing,
software becomes like a toaster, you just go get one and the only
difference is if it is white or black.  In our current world, there
are new things every day.  We don't understand them at first but in
time we learn their value and, after doing so, would never want to
be without them.

Here is an example.  Google.  I was the 4th person at Google, even though
I was only there for a few months, I got a feel for the people and the
place (very cool people, BTW).  Anyway, I had left Google, a year or so
had passed, and they were out there and useful.  My wife Beth and I were
having our first son, Travis.  As it turned out, he showed up 5 weeks
early.  We were unprepared, Beth's water broke, we didn't realize what
that meant but we went to the hospital to check it out and they wouldn't
let us go home.  They wanted to induce her with drugs (stuff that would
force her to go into contractions and make the baby come out).  You need
to realize that neither of us had a clue.  I had some sensation that this
wasn't right, I got on the phone with my sister who had 4 kids, on the
phone with an instructor who was teaching us about the birth process.
Both of them told us "keep that baby in there as long as you can".
But neither could tell me why, they just "knew" it was right.  I trusted
them but the doctor was screaming at Beth "if that baby doesn't come out
right now it could DIE!  It's going to get an infection, that's what happens
when your water breaks".  Beth is crying, she doesn't know what to think.
I don't trust the doctor, I think he's an asshole, but I have no data to
back up my feeling.

Being the geek that I am, I had not one but two laptops with me.  I plug
one in and dial up.  Hit Google and search on "infection premature baby"
or something like that.  Within 30 seconds I'm reading a New England
Journal of Medicine article (one of the best if you don't know) about
infection rates in women who's water breaks early.  It said that there
was basically no difference, less than one percent.  I shove this in the
asshole docter's face and say "what about this?"  He backed down a bit
and the baby stayed in there for another 30 hours or so.  Then Beth was
induced and Travis (http://www.bitmover.com/lm/nikon/1999-APR/21.html)
came out and was a happy healthy baby.

Much later I dug into this and found out the coolness which is the 
human body.  It turns out that your lungs are one of the last things
to develop and if you pull a baby out early there is a very high
chance that the lungs will be all screwed up.  "Screwed up" means that
the baby spends a month or two in neonatal care and you get to visit
him once in a while (you REALLY don't want this, it's bad).  On the
other hand, it also turns out that if the water breaks the system
recognizes that and turns up the clock on the lung development.  That
extra day inside made a huge difference in terms of lung development.
I'm completely convinced, based on what I've read, that that was the
difference between our baby going home with us and staying at the 
hospital for a couple of months.

And we owe it all to Google.  To my dieing day I will be grateful to the
Google team, I shudder to think what it would have been like without them.
And Google exists because Larry and Sergey want to be rich.  It's as
simple as that.  They are extremely talented and dedicated people, I 
have nothing but respect for them.  But it was clear that they were 
shrewdly building something they knew was valuable and they wanted to
turn it into a business.

If they had known that if they built it and anyone could steal their
technology because it was all free software, there is ZERO chance that
they would have done it.  They are bright people, they would have found
some other way to use their talents.  For me, it's a damn good thing
that they live in our current world, not your world or RMS' world.
Those are gray, boring, dull worlds.  No thanks.
-- 
---
Larry McVoy            	 lm at bitmover.com           http://www.bitmover.com/lm 

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Shane R. Stixrud]

On Fri, 3 Jan 2003, Larry McVoy wrote:
> And therein lies the problem.  You and RMS agree, in his world it is
> like getting your roof repaired or a new washing machine installed.

I only agree with RMS to a point.  Something like google or Ebay would 
still be very much possible in a world where most software was 
GPLed/BSD/etc..  Why wouldn't it be?

> And we owe it all to Google.  To my dieing day I will be grateful to the
> Google team, I shudder to think what it would have been like without them.
> And Google exists because Larry and Sergey want to be rich.  It's as
> simple as that.  They are extremely talented and dedicated people, I 
> have nothing but respect for them.  But it was clear that they were 
> shrewdly building something they knew was valuable and they wanted to
> turn it into a business.
[snip]

For every Larry and Sergey, I am sure you know ten Tom, Dick and 
Harry's who slave away doing amazing things at a large software 
development house. Besides, my reading of the GPL suggests you only 
need to publish source code _IF_ you distribute it to a 3rd party.  In 
house software development would not qualify.

> 
> If they had known that if they built it and anyone could steal their
> technology because it was all free software, there is ZERO chance that
> they would have done it.  They are bright people, they would have found
> some other way to use their talents.  For me, it's a damn good thing
> that they live in our current world, not your world or RMS' world.
> Those are gray, boring, dull worlds.  No thanks.
> 

Again I don't see why Google or Ebay wouldn't exist in a Free Software 
world nor do I see why Larry and Sergey wouldn't make a butt load.  

I think it is important to remember that copyright law was not designed to 
make you or anyone else "rich".  Copyright law is concerned with 
promoting societies progress.

Shane

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Matthew D. Pitts]

Shane,

> I think it is important to remember that copyright law was not designed to
> make you or anyone else "rich".  Copyright law is concerned with
> promoting societies progress.
>
> Shane
If you had written this a hundred years ago, it might have been more true.
The fact is, in the US anyway, copyright law has become totally
unrecognisable to what the Founding Fathers meant it to be.

Matthew D. Pitts

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Werner Almesberger]

Shane R. Stixrud wrote:
> I see a possible future where software development is seen much like
> getting your roof repaired, or adding a new room on to your house.
> It's labor+materials.

Exactly. Whenever somebody figures out a way to make some work
process more efficient, invariably a group of people will predict
imminent doom if this model catches on, and they will have plenty
of compelling reasons why the old way is vastly superior.

Well, I certainly feel terribly cheated by the cruel world that
has pretty much obsoleted the career choice "weaver" ;-))

- Werner

-- 
  _________________________________________________________________________
 / Werner Almesberger, Buenos Aires, Argentina         wa@almesberger.net /
/_http://www.almesberger.net/____________________________________________/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Mark Mielke]

On Sat, Jan 04, 2003 at 03:00:07AM -0300, Werner Almesberger wrote:
> Shane R. Stixrud wrote:
> > I see a possible future where software development is seen much like
> > getting your roof repaired, or adding a new room on to your house.
> > It's labor+materials.
> Exactly. Whenever somebody figures out a way to make some work
> process more efficient, invariably a group of people will predict
> imminent doom if this model catches on, and they will have plenty
> of compelling reasons why the old way is vastly superior.

If only it were that simple. Instead, the roof repair man repairs one
house, the fix is propagated to all other houses via the Internet, the
roof repair man puts *himself* out of a job, and the only people that
make money are the Internet service providers. The roof repair man
starves to death and finally ends up on (un)employment insurance.

Continually comparing software development to traditional forms of
labour is misleading and evidence that the issue is *not* being
properly understood. A closer comparison would be to compare the design
of software to the design of a building. But then - architects don't
give their blueprints away for free - that would be suicidal...

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Andre Hedrick]

On Sat, 4 Jan 2003, Mark Mielke wrote:

> On Sat, Jan 04, 2003 at 03:00:07AM -0300, Werner Almesberger wrote:
> > Shane R. Stixrud wrote:
> > > I see a possible future where software development is seen much like
> > > getting your roof repaired, or adding a new room on to your house.
> > > It's labor+materials.
> > Exactly. Whenever somebody figures out a way to make some work
> > process more efficient, invariably a group of people will predict
> > imminent doom if this model catches on, and they will have plenty
> > of compelling reasons why the old way is vastly superior.
> 
> If only it were that simple. Instead, the roof repair man repairs one
> house, the fix is propagated to all other houses via the Internet, the
> roof repair man puts *himself* out of a job, and the only people that
> make money are the Internet service providers. The roof repair man
> starves to death and finally ends up on (un)employment insurance.

You mean like today with everyone in the great dot-bomb failure, because
giving it all way is what happened a few years back, and they are now
gone.

> Continually comparing software development to traditional forms of
> labour is misleading and evidence that the issue is *not* being
> properly understood. A closer comparison would be to compare the design
> of software to the design of a building. But then - architects don't
> give their blueprints away for free - that would be suicidal...

Careful, the arguement will shift to, this is a public building and thus
the blueprints will be placed in a public record, in time.

Please keep up the good work of teaching, I am impressed you have the
ability to not shread people after you have explained it six ways from
sunday and they still do not get it.

Cheers,

Andre Hedrick
LAD Storage Consulting Group

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Werner Almesberger]

Mark Mielke wrote:
> Continually comparing software development to traditional forms of
> labour is misleading and evidence that the issue is *not* being
> properly understood.

The basic issue of how well you can live by performing certain work
is exactly the same. And in both cases, the incorrect assumptions
are to underestimate the potential for changes, and to look for a
transition that is risk-free. (*)

(*) Paranoiac's Categorical Imperative: if there's a risk for
    anybody, there could also be a risk for me, so I must oppose
    it.

> A closer comparison would be to compare the design
> of software to the design of a building. But then - architects don't
> give their blueprints away for free - that would be suicidal...

If we sent our kids to school, they could learn to do our jobs.
That would be suicidal... ;-)

Besides, nobody said that developing free software requires you
to decline being compensated for your work.

- Werner

-- 
  _________________________________________________________________________
 / Werner Almesberger, Buenos Aires, Argentina         wa@almesberger.net /
/_http://www.almesberger.net/____________________________________________/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Shane R. Stixrud]

On Sat, 4 Jan 2003, Mark Mielke wrote:

> If only it were that simple. Instead, the roof repair man repairs one
> house, the fix is propagated to all other houses via the Internet, the
> roof repair man puts *himself* out of a job, and the only people that
> make money are the Internet service providers. The roof repair man
> starves to death and finally ends up on (un)employment insurance.
> 
> Continually comparing software development to traditional forms of
> labour is misleading and evidence that the issue is *not* being
> properly understood. A closer comparison would be to compare the design
> of software to the design of a building. But then - architects don't
> give their blueprints away for free - that would be suicidal...
> 

Your position is based on the flawed assumption that the software world 
will ever have enough functionality or run out of problems to solve.  Even 
in the roofing analogy its not every house has the same roof. Software is 
only going to get more complex not less.

This is my last off topic post, sorry for the wasted bits.

Shane.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Mark Mielke]

On Sat, Jan 04, 2003 at 12:52:26AM -0800, Shane R. Stixrud wrote:
> Your position is based on the flawed assumption that the software world 
> will ever have enough functionality or run out of problems to solve.  Even 
> in the roofing analogy its not every house has the same roof. Software is 
> only going to get more complex not less.

My position is based on the observation that quite a significant
portion of features implemented for open source projects such as Linux
are provided by people that *cannot* directly benefit from the fruits
of their labour. Post a problem, or a bug report, and it is *likely*
that somebody will try to solve your problem, even if they do not have
your hardware, or their paying job does not relate to the open source
project at all.

It is based on good will, glory, thrill, enjoyment, distraction, and many
other benefits that do not put food on the table for your benefactor's
family.

This balance needs to be respected, not insulted. nVidia's driver is
not a patch to the roof of your house. nVidia's driver is software
that is released to allow Linux users the freedom to use nVidia
hardware. nVidia makes a lot more money from MSWIN32 users that it
does on Linux users. The software is protected either because nVidia
does not believe it is legal to distribute the source to their code,
or because they do not believe the benefits are worth the hassle. The
Linux community cannot lose something that it never had in the first
place.

Some people think that every company that needs an enhancement is
willing to pay people to enhance open source products, and that the
result will be returned to the public. The only real way that I can
see this working is if the Linux community as a whole created a
world-wide union that required all companies that wished to enhance
Linux would use to provide the enhancement. The pool of developers in
this union would be organized by knowledge and skill. The union
representantives would negotiate fair trades with companies,
encouraging companies to pool resources for features that are needed
by several companies.

Anything less is a little bit wishy washy. Maybe companies will pay
to have features implemented. Maybe companies will return the features
to the public. Maybe companies won't wait until somebody does it for
free. Maybe people who spend 37.5 hours a week working on Linux will
have food on the table for their family. Maybe a system exists. Maybe
companies know about the system.

Feeding ones family cannot be based on maybe.

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Roman Zippel]

Hi,

Larry McVoy wrote:

Somehow I tried to stay out of this, but I simply can't leave this
uncommented. Another long rant, better don't read this if you want to
keep your limitless optimism.

> The problem with your point of view is that you are assuming that somehow
> progress will continue to be made once you have that freedom.  Let's just
> look at that for a minute and see if that makes sense.
> 
> [...]
> 
> We can argue about this until the cows come up but it's simple economics.
> If there is no barrier to entry and a supplier is charging more it
> costs, a cheaper supplier will enter the market and force the price down.
> Even the most green MBA understands this and I don't think I need to tell
> you that the VC's all understand this.  For the sake of discussion, let's
> assume that you agree with that statement (if you don't, don't bother
> to argue with me, I'll ignore you, I'm not here to teach basic economics).
> 
> So we've established that in an all free world, even though some money
> will change hands, it can't be significantly more than what it costs
> to perform whatever service is being provided.  In other words, there
> is no extra money.
> 
> Is that a problem?  I think so.  If you look at the history of the free
> software movement, it has been a history of imitation rather than one of
> innovation (sorry to sound like Bill Gates but he has a point here).

You really shouldn't use economics to argue against free software, you
will only lose badly.
You make it sound as if imitation would be a bad thing, but without
imitation there would be no competition and without competition there
would be no incentive to innovate. Your remaining argumentation
basically says, that we should give up a little bit of our freedom to
sell it for money. Will it work? Maybe, the question is just for how
long.
To stay competitive there are basically only two ways, either you
destroy the competition or you offer better conditions than your
competition. It should be more or less obvious, why the first is a bad
idea. The constant force to undercut the competition has driven
innovation, but this is more and more becoming a curse. The innovation
cycles are becoming shorter and shorter, which makes it more and more
difficult to get the investments back. So the need to make money will
soon rather cause stagnation. Why is VC needed in first place and why is
it not possible to grow slowly anymore? Development costs? Not really,
it's mainly because you have to occupy a market as quickly as possible
to still make any money at all. Afterwards you want to protect your
market share and so any revolution will turn conservative.
You can blame whoever you want, but it's unavoidable that it's soon not
possible anymore to finance innovation, all you can do is to slow down
the development, but it's just a matter of time. Already now a lot of
companies are still busy to get their investments back, banks are
sitting on a lot of bad credits and at some point they want their money
back. Are they really interested in new competition? The last years we
lived vastly beyond our possibilities and we have become very dependent
on growth to pay our past debts. This slowly doesn't work anymore,
globalization is over, there will be no significant growth anymore, all
what is left is a cruel fight over market shares.
The problem with your point of view is that you are assuming that the
current system will somehow continue forever and it's the current system
that makes less and less sense. It's very possible that in a few years
you will be thankful to have that freedom that you're now so willing to
give up. It's indeed all basic economics, you just have to put aside the
blind faith in money for a moment.

bye, Roman

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Eric W. Biederman]

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Mark Mielke]

On Fri, Jan 03, 2003 at 03:30:32PM -0500, Richard Stallman wrote:
> > You don't seem to mind the fact that my freedom to use Linux would be
> > hampered if you successfully prove that [non-free] modules for
> > Linux are illegal.
> I'm not trying to prove this--as I see it, Linus gave permission for
> them, which means they are legal.  I regret his decision to do this,
> but I cannot change it.

You can't "regret" a decision that somebody else has made. To "regret"
something means to feel sorry for something. You don't have anything to
be sorry about.

> But let's suppose that that were changed.  It would not affect your
> "freedom" to use Linux (and GNU/Linux), only whether it runs on a
> certain computer.  It is true that this might mean a practical
> sacrifice--you might have to get a different kind of computer, for
> instance.  I don't see that as a horrible thing.  We look for
> computers that work with free drivers; you can too.

You are limited the scope of this discussion to hardware. For an
example of a software module that I regularly use in my every day job,
consider the MVFS module used to allow dynamic file system access to
ClearCase views. The MVFS module comes with the ClearCase distribution
as a small bit of open source glue, and a closed source object file
that implements MVFS.

> You don't really have freedom now, if you need a non-free module.  In
> the long run, your best chance of being able to use a fully free
> GNU/Linux system on the hardware you use is if we stand firm together
> for the freedom of the system.

I have the freedom to use Linux and ClearCase. If closed source modules
were to be disallowed, it would be illegal for me to use this configuration,
and I would be forced to use HP-UX or Solaris, and not Linux.

> You seem to be saying that we should sit back and let these inevitable
> forces either convince all companies to make software free--or not.
> If we had such a passive attitude, no free system would exist.
> GNU/Linux exists because of people who were willing to work to have
> freedom.  Freedom does not yet prevail, and we have plenty more work
> to do to make that happen.  And after we fully have freedom, we will
> still have to work, to make sure we don't lose it.

I'm saying that if you truly have a just cause, you don't need a hammer
or a sickle to force people to see things your way. Intelligent people
will have no choice but to follow your lead.

Visionaries should have faith in their own vision.

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Richard Stallman]

    I have the freedom to use Linux and ClearCase. If closed source modules
    were to be disallowed, it would be illegal for me to use this configuration,
    and I would be forced to use HP-UX or Solaris, and not Linux.

You can't have freedom while using ClearCase, because it is non-free
software.  What we really need is a free replacement for it.  Will
people write one?  Our main influence on whether people do this is by
what we say.  A strong Free Software Movement will inspire more people
to reject non-free software and write free replacements.

Allowing non-free modules (whether they are open-source or not)
weakens the impetus for people to make free extensions to Linux.  The
general attitude Linux developers take towards non-free software also
weakens it.  Your own message, citing this gap in Linux, will tend to
discourage people from working to close the gap.

All else being equal, I'm glad that you use a variant of the GNU
system, but what system you use is not really important except to you.
If you used HP-UX or Solaris, it would be your loss, not our
community's loss.  Spurring the broader development of free software
should be higher priority than keeping you as a user.

    I'm saying that if you truly have a just cause, you don't need a hammer
    or a sickle to force people to see things your way. Intelligent people
    will have no choice but to follow your lead.

Since our views have little in common with Communism, it is remarkable
that our enemies sometimes call us Communists.  Perhaps they do this
because it is easier to attack Communism than confront our real views.

It is the system of non-free software that resembles Stalinism.  For
more about this, see http://www.gnu.org/philosophy/why-free.html.

    Visionaries should have faith in their own vision.

Real visionaries know that just having a vision does not change
society.  Sustained effort is necessary.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Mark Mielke]

On Sat, Jan 04, 2003 at 06:44:38PM -0500, Richard Stallman wrote:
>     I have the freedom to use Linux and ClearCase. If closed source modules
>     were to be disallowed, it would be illegal for me to use this
>     configuration, and I would be forced to use HP-UX or Solaris, and
>     not Linux.
> You can't have freedom while using ClearCase, because it is non-free
> software.  What we really need is a free replacement for it.  Will
> people write one?  Our main influence on whether people do this is by
> what we say.  A strong Free Software Movement will inspire more people
> to reject non-free software and write free replacements.

In the *mean time*, I need a practical response, and not a claim that
'in a more perfect world, such and such'. If binary-only modules are
illegal, then I lose the freedom to legally use Linux + ClearCase MVFS.

You attach the word freedom to everything that suits your own political
agenda, but you refuse to allow me to use the word freedom in the above
context? Do you have a trademark on the word freedom?

> All else being equal, I'm glad that you use a variant of the GNU
> system, but what system you use is not really important except to you.
> If you used HP-UX or Solaris, it would be your loss, not our
> community's loss.  Spurring the broader development of free software
> should be higher priority than keeping you as a user.

The point where you are wrong, is that if I found it inconvenient to use
Linux, because I was not free to make use of closed source products, or
kernel modules as part of my operating environment, I would not be able
to contribute to Linux development. The point is more visible for those
in this mailing list who contribute a substantial amount.

>     I'm saying that if you truly have a just cause, you don't need a hammer
>     or a sickle to force people to see things your way. Intelligent people
>     will have no choice but to follow your lead.
> Since our views have little in common with Communism, it is remarkable
> that our enemies sometimes call us Communists.  Perhaps they do this
> because it is easier to attack Communism than confront our real views.

Actually I put the hammer/sickle reference above as an amusement, and
not a claim that you are communist. (Although -- as I understand it, the
more perfect society that would allow GPL to thrive *is* a communist
society -- which would be wonderful, if such a political system could
be proven to be feasible)

>     Visionaries should have faith in their own vision.
> Real visionaries know that just having a vision does not change
> society.  Sustained effort is necessary.

Traditional (not *REAL*) visionaries that have incomplete visions, where
the vision cannot be easily transferred from person to person without
mutating, or losing its scope, are forced to sustain their efforts as
a physical effort.

*REAL* visions are like viruses. All you need to do is transfer the vision
from one person to the next, and each time the image is transferred, the
next person takes up the cross as their own, with all of the energy and
motivation that the original visionary possessed.

You don't need to threaten people, or demean their methods in order to
propagate a *REAL* vision. Transfer of the vision is all that is necessary.
The people will see the light.

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Richard Stallman]

    In the *mean time*, I need a practical response, and not a claim that
    'in a more perfect world, such and such'. If binary-only modules are
    illegal, then I lose the freedom to legally use Linux + ClearCase MVFS.

You would be unable to do that unless/until someone (perhaps you?)
wrote a free implementation of MVFS for Linux.

Whether you can use GNU/Linux with ClearCase's non-free MVFS
implementation is not terribly important for our community.  I would
not base my decisions on that factor.  It is not worth changing a
license on a program just so that a small segment of computer users
might use a program *today* rather than waiting.

Enabling you and others to access MVFS with free software is the right
solution.  It may take longer, but it really solves the problem.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Richard Stallman]

    Traditional (not *REAL*) visionaries that have incomplete visions, where
    the vision cannot be easily transferred from person to person without
    mutating, or losing its scope, are forced to sustain their efforts as
    a physical effort.

    *REAL* visions are like viruses. All you need to do is transfer the vision
    from one person to the next, and each time the image is transferred, the
    next person takes up the cross as their own, with all of the energy and
    motivation that the original visionary possessed.

These definitions equate "real" with "complete" with "transmitted
perfectly reliability".  I don't agree with that equation, but you and
I both observe that the vision of free software transmits with less
than perfect reliability.

Many successful social movements have been less than "real" in your
terms.   Success requires persistent work.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[venom]

On Sat, 4 Jan 2003, Richard Stallman wrote:

> Date: Sat, 04 Jan 2003 18:44:38 -0500
> From: Richard Stallman <rms@gnu.org>
> To: mark@mark.mielke.cc
> Cc: billh@gnuppy.monkey.org, paul@clubi.ie, riel@conectiva.com.br,
>      Hell.Surfers@cwctv.net, linux-kernel@vger.kernel.org
> Subject: Re: Why is Nvidia given GPL'd code to use in non-free drivers?
>
>     I have the freedom to use Linux and ClearCase. If closed source modules
>     were to be disallowed, it would be illegal for me to use this configuration,
>     and I would be forced to use HP-UX or Solaris, and not Linux.
>
> You can't have freedom while using ClearCase, because it is non-free
> software.  What we really need is a free replacement for it.  Will
> people write one?  Our main influence on whether people do this is by
> what we say.  A strong Free Software Movement will inspire more people
> to reject non-free software and write free replacements.
>

Seeing this duscussion, I think that it will go for a lot of time, and
none will change his own position. Because I "mostly" agree with free software
goals, (not about this discussion, where I have a different opinion)
I would like to make some points why it will not possible to find an agreement
point beetwen contendants.

That is possibly one of the most interesting point.

- I am free because I have the freedom of choicheing the software I
  do prefer, even non free-software nor open source

- I am free because I use just free-software that is the instrumentum
  that warrants my freedom

Untill the real meaning of what is intended by being free will not be
defined, it is very difficoult to avoid this dual opposition.

That is a very important point, because you won't be able to
have a clear discussion with many of the subscribers of this mailing list,
simply because the mean a different application of freedom.

> Allowing non-free modules (whether they are open-source or not)
> weakens the impetus for people to make free extensions to Linux.  The
> general attitude Linux developers take towards non-free software also
> weakens it.  Your own message, citing this gap in Linux, will tend to
> discourage people from working to close the gap.
>

To be honest this is just partially true. Please consider the new
modules infrastructure with workqueue in 2.5 kernels.
Non GPL modules have a big
penalty, because they cannot create their own queue, but have to use a default
one.

> All else being equal, I'm glad that you use a variant of the GNU
> system, but what system you use is not really important except to you.
> If you used HP-UX or Solaris, it would be your loss, not our
> community's loss.

This, as you answer to Mark, is another point that makes me curious.
If, just supposing,
I am using bash emacs gcc and other free software on HP-UX,
or AIX, or
Solaris, to do my work.
I am not complitelly free, because the kernel and some of
the utility of the system are not free.

then what is exactly my condition?
I should be half-free. And that seems mutch to me quite similar to the
Aristotelian distiction beetwen sapiens (free) and savius (half free).

But then free software is ported, rightly, on all platform, included
M$ (my syster uses bash and emacs and TeX in W2000, when we talk about
tastes...)


> Since our views have little in common with Communism, it is remarkable
> that our enemies sometimes call us Communists.  Perhaps they do this
> because it is easier to attack Communism than confront our real views.

Who call you comunist is simply showing that he is complitelly
ignorant about what comunism is.

I suppose they use the word comunism because they
give to it a bad meaning. It is just a loose of time to consider
argumentation of people who use the "comunist" definition just
because they suppose this word should make a bad impression,
and maybe should scare.


[of topic]
Socialism, socialdemocracy and comunisms (there are more than one comunism)
are philosofical and
political systems that have really nothing to do with the
idea that most of people in the world have about them, because of
the leninist comunism (see "the three theories of socialdemocracy, written by
Lenin). But I see in this the bad influence of
a cultural propaganda made in USA in the fifties.
Where I live there has been the time when propaganda said that comunist
eat children. Well, in Italy the comunist party was around
30% by that times, and we never saw this cannibalism.

Half Europe has social and social democratic government, and nowhere
there is a law against private property. And nowhere in Europe
the modern comunist parties are proposing an abolition of private property.
[/of topic]



>
> It is the system of non-free software that resembles Stalinism.  For
> more about this, see http://www.gnu.org/philosophy/why-free.html.

I partially agree. I would say that closed source software sometimes
remembers me stalinism.
When it comes to other non free-software, but anyway open source
licenses it depends of the license terms.

I do suppose that at the beginning some of your oppositors
used the term comunist simply because the feeling of rigidity
that they could smell from your words. Or maybe because the comunist party
is known because of the strong discipline of its members, where
they could discuss very hard on topics, but when a decision was made, all
members were defending and sustaining and sharing the opinion of the party.

You cannot expect that everyone should share complitelly your opinions,
but, as it is normal, he could share some point, and be critical about some
other. Tha is the way that culture make its own progress.

>
> Real visionaries know that just having a vision does not change
> society.  Sustained effort is necessary.
>

That is exactly what is your right to do, and
personally I thank you for this.
But sustained effort does not mean to make a war against
every minimal disagreement from a walled opinion.
And that is why this discussion will never end even with
people on lkml, who should/could be the ones nearest to share
at less the most part of your view.

Luigi Genoni

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Geert Uytterhoeven]

On Sat, 4 Jan 2003, Richard Stallman wrote:
>     I have the freedom to use Linux and ClearCase. If closed source modules
>     were to be disallowed, it would be illegal for me to use this configuration,
>     and I would be forced to use HP-UX or Solaris, and not Linux.
> 
> You can't have freedom while using ClearCase, because it is non-free
> software.  What we really need is a free replacement for it.  Will
> people write one?  Our main influence on whether people do this is by

People are already working on it: http://www.advogato.org/proj/Katie/

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Henning P. Schmiedehausen]

Mark Mielke <mark@mark.mielke.cc> writes:

>I have the freedom to use Linux and ClearCase. If closed source modules
>were to be disallowed, it would be illegal for me to use this configuration,
>and I would be forced to use HP-UX or Solaris, and not Linux.

No, it wouldn't. Thats' what most people don't understand. You
wouldn't have a license to GIVE AWAY a system which consists of Linux
kernel and MVFS object module.

You definitely have a license to get a Linux system, install it, run
it and install on it every piece of software you like. If you install
MVFS, there is nothing in the GPL to prevent you from this. Neither in
the GPL nor in the Linux-modified version of "you may load binary
modules" GPL. This is your personal decision of your personal
system. If you install a module that is binary only, fine.

GPL is about SOFTWARE DISTRIBUTION. Not about SOFTWARE USAGE. I know
(and I read this in many of his postings) that RMS likes to blur this
point into "if it is not free, you must not use it with GPL software",
but this is simply _NOT TRUE_. It is your personal freedom to choose
and use a binary module. If you redistribute it, you may take freedom
from the recipient away and this prohibits the GPL. But not your
personal usage.

Sheesh. I have lots of kernel modules in current use which will never
be released outside the scope of my own boxes. That's no breach of the
GPL. You'll never be able to acquire either a source or a binary code
license. This is my code. You cannot have it. My freedom to decide so. 
End of story.

	Regards
		Henning


-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     hps@intermeta.de

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   info@intermeta.de
D-91054 Buckenhof     Fax.: 09131 / 50654-20   

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Andre Hedrick]

On Sun, 5 Jan 2003, Henning P. Schmiedehausen wrote:

> Mark Mielke <mark@mark.mielke.cc> writes:
> 
> >I have the freedom to use Linux and ClearCase. If closed source modules
> >were to be disallowed, it would be illegal for me to use this configuration,
> >and I would be forced to use HP-UX or Solaris, and not Linux.
> 
> No, it wouldn't. Thats' what most people don't understand. You
> wouldn't have a license to GIVE AWAY a system which consists of Linux
> kernel and MVFS object module.
> 
> You definitely have a license to get a Linux system, install it, run
> it and install on it every piece of software you like. If you install
> MVFS, there is nothing in the GPL to prevent you from this. Neither in
> the GPL nor in the Linux-modified version of "you may load binary
> modules" GPL. This is your personal decision of your personal
> system. If you install a module that is binary only, fine.
> 
> GPL is about SOFTWARE DISTRIBUTION. Not about SOFTWARE USAGE. I know
> (and I read this in many of his postings) that RMS likes to blur this
> point into "if it is not free, you must not use it with GPL software",
> but this is simply _NOT TRUE_. It is your personal freedom to choose
> and use a binary module. If you redistribute it, you may take freedom
> from the recipient away and this prohibits the GPL. But not your
> personal usage.

Sweet, and true.

The effect is shipping a binary alone without its associated kernel is the
distribution of a product independent.

One other point, about RMS and FSF ... They have no stake or holding in
the linux kernel, only a license about distribution.

So when the poke there nose in this issue and interfere in the operations
and business, they are exposing themselves to litigation.  This was an
interesting point made to me and I think it needs research.

Something else that needs research is a linuxgram story which I am having
trouble tracking down.  It has something to say about a FSF/GPL
certification audit for $25,000.00, regardless if it is open or closed.
If this is true, a conflict of interest, and a huge grey area is worthly
of investigation.  The questions to ask:

Who has requested certification?
Who has passed?
Who has failed?
Has any one failed?

If the last question is answered by "none" ...

> Sheesh. I have lots of kernel modules in current use which will never
> be released outside the scope of my own boxes. That's no breach of the
> GPL. You'll never be able to acquire either a source or a binary code
> license. This is my code. You cannot have it. My freedom to decide so. 
> End of story.

Thanks is has been fun and informative.

Cheers,

Andre Hedrick
LAD Storage Consulting Group

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Mark Mielke]

On Sun, Jan 05, 2003 at 09:33:48PM +0000, Henning P. Schmiedehausen wrote:
> Mark Mielke <mark@mark.mielke.cc> writes:
> >I have the freedom to use Linux and ClearCase. If closed source modules
> >were to be disallowed, it would be illegal for me to use this configuration,
> >and I would be forced to use HP-UX or Solaris, and not Linux.
> No, it wouldn't. Thats' what most people don't understand. You
> wouldn't have a license to GIVE AWAY a system which consists of Linux
> kernel and MVFS object module.
> ...

Just to point out - I said "If closed source modules were to be disallowed,".

Also - the question isn't whether closed source modules can be distributed,
as much as "can closed source modules that could not be compiled without GPL
source code (header files), be distributed?"

RMS wishes my configuration (Linux + ClearCase MVFS) to be illegal,
because he wishes to enforce an all-free ("free" as defined by RMS)
final product, and the existence of closed-source hardware drivers
(nVidia) or software extensions (ClearCase MVFS) are in the way of
this goal.

If he succeeds, I may lose the freedom to effectively use Linux,
because I don't *mind* buying good software, and I don't *mind* if it
is closed source. Why don't I mind? Because, with few exceptions,
closed source software for expensive price tags tends to be better, or
fuller in some way, in my experience. We live in a capitalist society.
Pretending that capitalism can be avoided is... not realistic.

mark

-- 
mark@mielke.cc/markm@ncf.ca/markm@nortelnetworks.com __________________________
.  .  _  ._  . .   .__    .  . ._. .__ .   . . .__  | Neighbourhood Coder
|\/| |_| |_| |/    |_     |\/|  |  |_  |   |/  |_   | 
|  | | | | \ | \   |__ .  |  | .|. |__ |__ | \ |__  | Ottawa, Ontario, Canada

  One ring to rule them all, one ring to find them, one ring to bring them all
                       and in the darkness bind them...

                           http://mark.mielke.cc/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[David Schwartz]

On Fri, 03 Jan 2003 15:30:32 -0500, Richard Stallman wrote:

>You don't really have freedom now, if you need a non-free module.  In
>the long run, your best chance of being able to use a fully free
>GNU/Linux system on the hardware you use is if we stand firm together
>for the freedom of the system.

	But I don't want to coerce other people into providing me with my freedom.

>If open source is so good, companies with closed source products will
>change.

	Yes, even without being coerced and pressured to do so by restrictive 
licenses.

>I don't support the open source movement, but I know what they say
>about this.  They say that open source usually leads to more powerful
>and reliable software.  Nothing assures us that will persuade all
>companies to adopt the practice.  You have simplified their position
>to a point where they would not recognize it.

	The point is not to persuade companies to adopt the practice. The point is 
to show that the practice is superior and let the companies that adopt it 
prosper and those who don't fail. The GPL weakens this position by providing 
proprietary software with an excuse.

	It is roughly comparable to the United States embargo on Cuba. We want Cuba 
to change, so we don't let them use any of our stuff. Force them to be free, 
or we won't touch 'em.

	It fails for the same reason. If you believe in freedom, set the example. 
Set people free. Defend fair use, first sale, and a very strict definition of
a derived work.

	That's real freedom, and the GPL works against it by attempting to coerce 
freedom by using legal tools that problably shouldn't exist.

>You seem to be saying that we should sit back and let these inevitable
>forces either convince all companies to make software free--or not.

	No. We're saying that we shouldn't try to rig the system. We should allow 
free software to win in a fair fight. Not by using vicious legal tools to 
coerce others to provide your freedom.

>If we had such a passive attitude, no free system would exist.
>GNU/Linux exists because of people who were willing to work to have
>freedom.  Freedom does not yet prevail, and we have plenty more work
>to do to make that happen.  And after we fully have freedom, we will
>still have to work, to make sure we don't lose it.

	Defending shrink wrap licensing agreements, arguing to weaken fair use and 
first sale doctrines, and arguing that if you include a header it's a derived 
work is a strange way to defend intellectual freedom.

	DS

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Richard Stallman]

	    Defending shrink wrap licensing agreements, arguing to weaken fair use and 
    first sale doctrines, and arguing that if you include a header it's a derived 
    work is a strange way to defend intellectual freedom.

Those are not my views.  Are you confusing me with someone else?

    >If open source is so good, companies with closed source products will
    >change.

    Yes, even without being coerced and pressured to do so by restrictive 
    licenses.

The Open Source Movement says that will happen; when it does, that's
good, but if we had relied on that to give us freedom, we wouldn't
have any free operating systems today.

In the Free Software Movement we think freedom is worth working for.
If companies don't choose to respect our freedom, we don't cite that
and say "it's hopeless" and we don't say that makes non-freedom ok.
We write free replacements and build freedom for ourselves--and for
you.

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[David Schwartz]

On Sat, 04 Jan 2003 18:44:58 -0500, Richard Stallman wrote:

>Defending shrink wrap licensing agreements, arguing to weaken
>fair use and
>first sale doctrines, and arguing that if you include a header it's
>a derived
>work is a strange way to defend intellectual freedom.

>Those are not my views.  Are you confusing me with someone else?

	Then please explain to me how the GPL comes to apply to a person who
did not agree to it as a condition of receiving a copyrighted work. 
Please explain to me why you think that the GPL should have applied 
to kernel modules that only include header files.

	You may not explicitly endorse the obvious logical consequences of 
your views, but you are still responsible for them.

>>If open source is so good, companies with closed source products
>>will
>>change.

>Yes, even without being coerced and pressured to do so by
>restrictive
>licenses.

>The Open Source Movement says that will happen; when it does, that's
>good, but if we had relied on that to give us freedom, we wouldn't
>have any free operating systems today.

	That's a lot better than trying to arm twist others in to providing 
our freedom to use their works. When you talk about forcing a person 
to distribute the source code to a derived work, you are only talking
about their control over what they added. When a person creates a 
derived work of an open source work, all they have to offer is the 
value they added. In the name of freedom, you take their control over
their work from them.

	This is the same "freedom" that socialism promises the workers. They
call it the freedom to own the machinery they use to produce. 
Analogously, this "freedom" is really just the loss of the freedom of
ownership.

>In the Free Software Movement we think freedom is worth working for.
>If companies don't choose to respect our freedom, we don't cite that
>and say "it's hopeless" and we don't say that makes non-freedom ok.
>We write free replacements and build freedom for ourselves--and for
>you.

	This is false for two reasons:

	1) The difference between the GPL and the BSD license is the GPL 
license *compels* source distribution. You can't compel someone else 
to make you free. It's just not going to work.

	2) To make the GPL enforceable, you need to argue for a very loose 
definition of a derived work and you need to argue that a license can 
be enforceable even if it's not negotiated or explicitly agreed to 
prior to distribution. This will have the net effect of reducing 
everyone's freedom in very real ways.

	DS

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Wolfgang Walter]

On Sunday 05 January 2003 01:17, David Schwartz wrote:
> On Sat, 04 Jan 2003 18:44:58 -0500, Richard Stallman wrote:
> >Defending shrink wrap licensing agreements, arguing to weaken
> >fair use and
> >first sale doctrines, and arguing that if you include a header it's
> >a derived
> >work is a strange way to defend intellectual freedom.
> >
> >Those are not my views.  Are you confusing me with someone else?
>
> 	Then please explain to me how the GPL comes to apply to a person who
> did not agree to it as a condition of receiving a copyrighted work.
> Please explain to me why you think that the GPL should have applied
> to kernel modules that only include header files.
>

You seem not to understand copyright.

The GPL does not affect the user of the software. If you have bought a copy of 
Red Linux distribution cd i.a. it is not necessary to accept the GPL (or BSD 
or whatever license) to use the software. You may sell your received copy 
when ever you want to ever you want for whatever price you can get - if you 
do not keep a copy. As you can do with microsoft windows - if you bought it 
(and did not licensed it from microsoft). I.a. it is not necessary to provide 
source code because it is Red Hat which a) made the copy and b) did so by 
accepting the GPL. 

But if you want to make and use or distribute copies of that CD or distributed 
works, well, then you must get explicit permission from the copyright owners 
- as you would have to for any copyrightable work. This is so because of 
copyright law. If you buy the software you only have the right to use it. You 
do not have by default the right to distribute copies, make or distribute 
derived works etc.

If the CD would be a copy of microsoft windows you would have to negotiate 
with microsoft - probably they would not allow that you distribute a derived 
version.

Now the authors of the software on the Red Hat CD make you an offer: you may 
accept the GPL. If you do so, they allow you to make and distribute copies or 
derived works under certain conditions. You don't have to accept the GPL. If 
you do not, you may try to negotiate for other terms with the copyright 
holders.

> 	That's a lot better than trying to arm twist others in to providing
> our freedom to use their works. When you talk about forcing a person
> to distribute the source code to a derived work, you are only talking
> about their control over what they added. When a person creates a

Do you understand? You are not allowed to produce derived works without 
permission of the copyright owner. He may do so under what consitions he want 
(or simply does not allow you to do so at all).

With the GPL the copyright owner(s) of the work grants you the right to do so 
under certain conditions described under the GPL. One right is to produce 
derived works at all and 

Have you ever got permission from microsoft or adobe to produce derived works 
from windows 2000 or photoshop?

> derived work of an open source work, all they have to offer is the
> value they added. In the name of freedom, you take their control over
> their work from them.

No, they allow you to do the work at all. By default you would not be allowed 
to add value at all.

>
> 	This is the same "freedom" that socialism promises the workers. They
> call it the freedom to own the machinery they use to produce.
> Analogously, this "freedom" is really just the loss of the freedom of
> ownership.

No. The authors of the work has with your words - the "ownership" of his work. 
Law says that one facete of that "ownership" is that he may allow or forbid 
derived works. And if he allows someone to produce a derived work its under 
his conditions.

The GPL does not restrict you. Contrary, it uses copyright law to establish a 
pool of software with much more freedom as copyright law gives to you. It 
only does not give you so much freedom to deny other people the same 
freedoms. The copyright holders can only do so, because copyright law itself 
give you none of these rights at all.

Richard Stallmann - if I understand him right - beliefs that the rights the 
GPL grants should be granted (for software) by copyright law itself and 
therefor granted for every software. You may argument about that.

But you cant't argument that an author as owner of his work should use a less 
restrictive license than the GPL so you can make a derived work and 
distribute it under a more restrictive license than GPL. Why should he want 
to allow that at all (a lot of peoply allow that choosing a BSD-license - 
nice gift)? Nobody can force him to do so. Its not the GPL which restricts 
your freedom, it is copyright law and the author(s) of the work you want to 
made a derived work from.

It is simply impossible that authors have control under which condition 
derived works may be made from their works AND in the same time have the 
right to made derived works from works of other authors without control of 
these authors.


By the way: a completely different question is if a work is a derived work. Is 
a driver for nvidia a derived work. Well, the GPL can not define that, of 
course - because it only applies to derived works. No license can that. A 
licence may state what it will not regard as a derived work.

The courts decide what is a derived work. The courts decide that your book 
with a main character named Harry Potter, wizard studying in Hogwards, is a 
derived work from 4 those well known Rowling-books and that you may not 
distribute it without permission.


Greetings,

Wolfgang Walter
-- 
Wolfgang Walter
Studentenwerk München
Anstalt des öffentlichen Rechts
EDV
Leopoldstraße 15
80802 München
Tel: +49 89 38196-276
Fax: +49 89 38196-144
wolfgang.walter@studentenwerk.mhn.de
http://www.studentenwerk.mhn.de/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[David Schwartz]

On Sun, 5 Jan 2003 05:39:35 +0100, Wolfgang Walter wrote:
>On Sunday 05 January 2003 01:17, David Schwartz wrote:
>>On Sat, 04 Jan 2003 18:44:58 -0500, Richard Stallman wrote:
>>>Defending shrink wrap licensing agreements, arguing to weaken
>>>fair use and
>>>first sale doctrines, and arguing that if you include a header
>>>it's
>>>a derived
>>>work is a strange way to defend intellectual freedom.
>>>
>>>Those are not my views.  Are you confusing me with someone else?
>>
>>    Then please explain to me how the GPL comes to apply to a 
person
>>who
>>did not agree to it as a condition of receiving a copyrighted work.
>>Please explain to me why you think that the GPL should have applied
>>to kernel modules that only include header files.

>You seem not to understand copyright.

	I don't see what gives you this impression.

>The GPL does not affect the user of the software. If you have bought
>a copy of
>Red Linux distribution cd i.a. it is not necessary to accept the GPL
>(or BSD
>or whatever license) to use the software. You may sell your received
>copy
>when ever you want to ever you want for whatever price you can get -
>if you
>do not keep a copy. As you can do with microsoft windows - if you
>bought it
>(and did not licensed it from microsoft). I.a. it is not necessary
>to provide
>source code because it is Red Hat which a) made the copy and b) did
>so by
>accepting the GPL.

	Right.

>But if you want to make and use or distribute copies of that CD or
>distributed
>works, well, then you must get explicit permission from the
>copyright owners
>- as you would have to for any copyrightable work. This is so
>because of
>copyright law. If you buy the software you only have the right to
>use it. You
>do not have by default the right to distribute copies, make or
>distribute
>derived works etc.

	You have those rights the law gives you and those rights that the 
copyright holder chooses to give you in the transfer agreement.

>Now the authors of the software on the Red Hat CD make you an offer:
>you may
>accept the GPL. If you do so, they allow you to make and distribute
>copies or
>derived works under certain conditions. You don't have to accept the
>GPL. If
>you do not, you may try to negotiate for other terms with the
>copyright
>holders.

	Sounds like every shrink wrap agreement in the world. You already 
have the thing you want to license, the licensee simply refuses to 
grant you the rights to that thing you already have unless you agree 
to a license that you are not free to negotiate.

>>    That's a lot better than trying to arm twist others in to
>>providing
>>our freedom to use their works. When you talk about forcing a
>>person
>>to distribute the source code to a derived work, you are only
>>talking
>>about their control over what they added. When a person creates a

>Do you understand? You are not allowed to produce derived works
>without
>permission of the copyright owner. He may do so under what
>conditions he want
>(or simply does not allow you to do so at all).

	This is the same for use. If Microsoft wants to, they can impose any
terms in the EULA that they want.

>With the GPL the copyright owner(s) of the work grants you the right
>to do so
>under certain conditions described under the GPL. One right is to
>produce
>derived works at all and

>Have you ever got permission from microsoft or adobe to produce
>derived works
>from windows 2000 or photoshop?

	Microsoft doesn't try to argue that every document I write in 
Windows 2000 is a derived work. Photoshop doesn't argue that every 
image I create in photoshop is a derived work.

	All you can do with a header file is include it in your own code. 
All you can do with photoshop is produce photoshop files. Adobe 
doesn't argue that photoshop-created images are derived works. 
Stallman *does* argue that Linux binary modules are derived works.

	To support the GPL's ability to regulate the distribution of derived
works you would have to argue that Adobe's EULA could legitimately 
prohibit you from distributing images you create with photoshop. Far 
smarter for advocates of freedom to argue that this is fair use and 
the argument that such works are derived is bullcrap.

>>derived work of an open source work, all they have to offer is the
>>value they added. In the name of freedom, you take their control
>>over
>>their work from them.

>No, they allow you to do the work at all. By default you would not
>be allowed
>to add value at all.

	Yes, but this is *use*, which is what the GPL is *not* supposed to 
stop. How can you use photoshop except to create images with it? How 
can you use a header file except to include it in your own code. I 
argue that we should take the position that this type of normal use 
does not create a derived work any more than reading a novel makes 
your brain a derived work of that novel.

>>    This is the same "freedom" that socialism promises the workers.
>>They
>>call it the freedom to own the machinery they use to produce.
>>Analogously, this "freedom" is really just the loss of the freedom
>>of
>>ownership.

>No. The authors of the work has with your words - the "ownership" of
>his work.
>Law says that one facete of that "ownership" is that he may allow or
>forbid
>derived works. And if he allows someone to produce a derived work
>its under
>his conditions.

	The argument is over what is a derived work, what constitutes 
"using" a header file, and what constitutes agreement to a contract.

>But you cant't argument that an author as owner of his work should
>use a less
>restrictive license than the GPL so you can make a derived work and
>distribute it under a more restrictive license than GPL. Why should
>he want
>to allow that at all (a lot of peoply allow that choosing a BSD-
>license -
>nice gift)? Nobody can force him to do so. Its not the GPL which
>restricts
>your freedom, it is copyright law and the author(s) of the work you
>want to
>made a derived work from.

	I guess I haven't made myself clear. My argument is not specifically
with the GPL except in the sense that it requires people who support 
it to take anti-freedom positions with respect to fair use, derived 
works, first sale, and other important issues where actual 
information freedom is at stake.

	I'm afraid I'm too tired right now to respond to the rest of your 
argument. I hope I didn't miss anything importasnt.

	DS

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Wolfgang Walter]

On Sunday 05 January 2003 06:35, David Schwartz wrote:
> On Sun, 5 Jan 2003 05:39:35 +0100, Wolfgang Walter wrote:
> >On Sunday 05 January 2003 01:17, David Schwartz wrote:
> >>On Sat, 04 Jan 2003 18:44:58 -0500, Richard Stallman wrote:

> 	Sounds like every shrink wrap agreement in the world. You already
> have the thing you want to license, the licensee simply refuses to
> grant you the rights to that thing you already have unless you agree
> to a license that you are not free to negotiate.
>

A shrink wrap agreement is something completely different.  You must 
differiate between using software and the exploitation right of the 
copyright-user.

Say you buy a book. Reading it ist usage. Destroing it is usage. But writing a 
book which contains part of this book is not using it. Lending it in public 
libraries is not usage. Making copies and distribute them is not usage.

You don't need a license from the author to use the book.

A shrink wrap license agreement (or EULA) tries to restrict your rights to USE 
your bought copy THOUGH you didn't bought it from the person who wants you to 
do so and AFTER you bought it. With the book-example: you may only read it by 
night and you are not allowed to speak bad of it.

In Germany these shrink wrap license agreement and EULAs are simply invalid. I 
don't need a license to use the software I bought. Even though I have to 
click on "I Accept" or "I Agree" - this means nothing as I have to, to 
install it, and I have to install it, to use it.

Of course copyright law is different from country to country but this so in 
most countries.

In Germany microsoft tried to inhibit that peoply sell there copy of windows 
bought with a new computer (based on there EULA which declared this copy as 
OEM and only valid together with this computer). They failed of course - 
there is no license-agreement between the owner of this windows-copy and 
microsoft. I didn't license the copy, I bought it and own it. And to own 
software is enough to use it. They can't restrrict unilaterally my right to 
use it.

If I now use the software-update of windows, things get different. Then I 
conclude a license agreement with them.

It is the same with a Red Hat CD. To use the software you don't need a 
license. They cannot restrcit you in your legal rights as a user.

Back to nvidia: if nvidea-drivers are derived work from the kernel I don't 
now. By itself probably not if they don't use kernel-code. Does a user commit 
a copyright infringement if he loads them as module? Probably not because it 
is using the software (the kernel can and does load this module, you don't 
need to modify it). Does the use infringe the GPL? Hmm, as long as he uses 
the kernel-binary he bought and this kernel provides the mechanism to load 
the module he don't need to accept the GPL.

If he has to modify the kernel to load the module, then of course he has to 
accept the GPL because modifing the kernel is not using it. And then the GPL 
may forbid him to do so.

You see what would be the way to effectivly forbid non GPL-modules by a user: 
a) force the user so he has to modify the kernel OR b) force a module writer 
to include copyrighted material.

Maybe it is enough that the loader-mechanism in user-space and that this tool 
is not part of the kernel to make it legally a modification of the kernel - 
but I doubt it. But kernel-developpers may check that the module includes a 
poetry they wrote and which is part of the kernel-code.

I see (you state that below) that you think that using header files in 
software-projects is not making a derived work from those header files but 
instead using them.

I don't know what a court will decide. But I think this does not hold for 
header files as it does not hold for runtime libraries etc.

But of course you can reverse engineer and write your ones. Reverse 
engineering is rather easy with open source.

> 	This is the same for use. If Microsoft wants to, they can impose any
> terms in the EULA that they want.
>

No - not in most countries, not in the EU. If you don't conclude an agreement 
which microsoft which most people don't do. You buy a computer with windows 
2000 - you don't have an agreement with microsoft and they can't unilaterally 
force you to do so by effectifly making the product unusuable.

This is even so in most states of USA.

> 	Microsoft doesn't try to argue that every document I write in
> Windows 2000 is a derived work. Photoshop doesn't argue that every
> image I create in photoshop is a derived work.
>

Hmm, does OpenOffice that? Does Gimp that? No, of course not.

> 	All you can do with a header file is include it in your own code.
> All you can do with photoshop is produce photoshop files. Adobe
> doesn't argue that photoshop-created images are derived works.

They do. If you use there cliparts-collection you may produce derived works 
(it depends on the clipart) and the you NEED the agreement of the 
copyright-owners.

If you produce a PDF an you include copies of fonts you NEED the permission of 
the copyright holders of those fonts (if the fonts are copyrighted).

> Stallman *does* argue that Linux binary modules are derived works.

I don't know if he does.

If the source code of binary-modules do not contain copyrighted material from 
the kernel they probably not derived works. Loading the module into the 
kernel by the user may produce a derived work. Putting kernel and modules 
together in a distribution may produce a derived work.

Using kernel header files to produce the binary is very probably making a 
derived work. But it would be rather hard to prove that - as it is so easy to 
reverse engineer open source software and write your own header files.

>
> 	To support the GPL's ability to regulate the distribution of derived
> works you would have to argue that Adobe's EULA could legitimately
> prohibit you from distributing images you create with photoshop. Far

A image produced with photoshop is not a derived work. It does not contain 
photoshop. If you use a nice picture they delivered with photoshop as base 
then of course you may need a license.

> smarter for advocates of freedom to argue that this is fair use and
> the argument that such works are derived is bullcrap.

Fair use is something different. Fair use is about exploitation right without 
permission of the copyright holder. I.e. you may cite a book in your book 
(but you may not print a whole page or so). Making a copy of a book for 
private use without permission of the copyright holder. (In Germany i.a. you 
pay for this right: on every copy-device as cd-burners, printers, and on 
memories like harddiscs, blank CDs, etc. there are fees). 

For software there is almost no fair use in the EU. I.e. the right for private 
copies does not exist. On the other habd there are other explicit rights, 
i.e. to decompilate software to see how it works.

> 	Yes, but this is *use*, which is what the GPL is *not* supposed to
> stop. How can you use photoshop except to create images with it? How
> can you use a header file except to include it in your own code. I

You can read it. You can use that knowledge to write your own.

A

#include "linux/blabla.h"

does not make your file a derived work as long as you do not distribute those 
files with your file.

The one who compiles it using the kernel header files makes a derived work - 
the binary is a derived work. But thats my opinion. You thinks that it is 
using them.


Greetings,

Wolfgang Walter

-- 
Wolfgang Walter
Studentenwerk München
Anstalt des öffentlichen Rechts
EDV
Leopoldstraße 15
80802 München
Tel: +49 89 38196-276
Fax: +49 89 38196-144
wolfgang.walter@studentenwerk.mhn.de
http://www.studentenwerk.mhn.de/

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[David Schwartz]

On Mon, 6 Jan 2003 17:24:21 +0100, Wolfgang Walter wrote:
>On Sunday 05 January 2003 06:35, David Schwartz wrote:
>>On Sun, 5 Jan 2003 05:39:35 +0100, Wolfgang Walter wrote:
>>>On Sunday 05 January 2003 01:17, David Schwartz wrote:
>>>>On Sat, 04 Jan 2003 18:44:58 -0500, Richard Stallman wrote:

>>Sounds like every shrink wrap agreement in the world. You
>>already
>>have the thing you want to license, the licensee simply refuses to
>>grant you the rights to that thing you already have unless you
>>agree
>>to a license that you are not free to negotiate.

>A shrink wrap agreement is something completely different.  You must
>differentiate between using software and the exploitation right of 
>the copyright-user.

	You can aim this criticism at many other people in this 
conversation, but not me. I think I'm the only one who does 
differentiate clearly.	

>Say you buy a book. Reading it ist usage.

	Really? How do you read a book? You bounce a light off it and make a 
copy of the book on your retina, right? In other words, you use 
things by copying them.

>Destroing it is usage. But
>writing a
>book which contains part of this book is not using it. Lending it in
>public
>libraries is not usage. Making copies and distribute them is not
>usage.

	Yes, copies *of* *that* *book*. But when you make a copy of the book 
on your retina, your eyes and brain are not a derived work. When you 
use photoshop, the graphics you create are not a derived work.

	I submit that the *only* way to use a header file is to include it 
in a source file, and compile and copy the resultant output. Note 
that you cannot run a program without copying it. It's physically 
impossible.

>You don't need a license from the author to use the book.

	Right, because we recognize that a graphic created with photoshop is 
*not* a derived work of photoshop. A brain that has read a book is 
not a derived work of that book. Similary, a program whose source 
code includes a header file should not be considered a derived work 
of that header file.

>A shrink wrap license agreement (or EULA) tries to restrict your
>rights to USE
>your bought copy THOUGH you didn't bought it from the person who
>wants you to
>do so and AFTER you bought it. With the book-example: you may only
>read it by
>night and you are not allowed to speak bad of it.

	Tell me how you use a computer program without copying it. Please, 
do that. How do you use a CD without making a copy of the data on it?

	Use and copying are the same when it comes to information. There is 
no other way to use information. This is why it's critical to 
strengthen fair use, first sale, and necessary step type defenses.

	You can't use a header file without including it in source code. You 
can't use the resultant object file without copying it. Thus these 
*must* be fair uses.

>In Germany microsoft tried to inhibit that peoply sell there copy of
>windows
>bought with a new computer (based on there EULA which declared this
>copy as
>OEM and only valid together with this computer). They failed of
>course -
>there is no license-agreement between the owner of this windows-copy
>and
>microsoft. I didn't license the copy, I bought it and own it. And to
>own
>software is enough to use it. They can't restrrict unilaterally my
>right to
>use it.

	How can you use Windows without copying it from your hard drive into 
memory? Copying is using. Using is copying.

>If he has to modify the kernel to load the module, then of course he
>has to
>accept the GPL because modifing the kernel is not using it. And then
>the GPL
>may forbid him to do so.

	Modifying the kernel is not using it? A copy of the kernel is RAM is 
different from a copy of the kernel on hard drive. This is a 
transformative modification. You cannot use the linux kernel without 
modifying it. And guess what? When you run the Linux kernel, are the 
data structure you thereby create in your memory derived works?

>I see (you state that below) that you think that using header files
>in
>software-projects is not making a derived work from those header
>files but
>instead using them.

	Tell me how else you can use them. Please. Go ahead. Tell me any 
other way to use a header file other than to include it in a source 
file, compile that source file, and then copy the resulting 
executable.

>>Stallman *does* argue that Linux binary modules are derived works.

>I don't know if he does.

	Then he is arguing to weaken fair use, first sale, and necessary 
step type principles. These are far more important than the GPL.

>If the source code of binary-modules do not contain copyrighted
>material from
>the kernel they probably not derived works. Loading the module into
>the
>kernel by the user may produce a derived work. Putting kernel and
>modules
>together in a distribution may produce a derived work.

	No, because these are all necessary steps. A necessary step to use 
is always use. You cannot use the kernel without copying it into 
memory. You cannot use the kernel without feeding it information and 
having it produce structure in memory.

	Stallman is out to destroy fair use. Whether you knows it or admits 
it or not.

>Using kernel header files to produce the binary is very probably
>making a
>derived work. But it would be rather hard to prove that - as it is
>so easy to
>reverse engineer open source software and write your own header
>files.

	So tell me how else you use kernel header files. What else can you 
do with a header file?!

>>    To support the GPL's ability to regulate the distribution of
>>derived
>>works you would have to argue that Adobe's EULA could legitimately
>>prohibit you from distributing images you create with photoshop.

>A image produced with photoshop is not a derived work. It does not
>contain
>photoshop. If you use a nice picture they delivered with photoshop
>as base
>then of course you may need a license.

	Exactly. All photoshop can do is produce images. Therefore producing 
images with photoshop is use, barring some exceptional circumstance. 
(For example, if you take an image from their clip art.)

>>smarter for advocates of freedom to argue that this is fair use and
>>the argument that such works are derived is bullcrap.

>Fair use is something different. Fair use is about exploitation
>right without
>permission of the copyright holder. I.e. you may cite a book in your
>book
>(but you may not print a whole page or so). Making a copy of a book
>for
>private use without permission of the copyright holder. (In Germany
>i.a. you
>pay for this right: on every copy-device as cd-burners, printers,
>and on
>memories like harddiscs, blank CDs, etc. there are fees).

	Fair use includes any number of ways you can do things you might not 
normally be able to do. This includes 'necessary step' (this is why 
you can make a copy of a book on your retina) and 'firs sale' type 
rights.

>For software there is almost no fair use in the EU. I.e. the right
>for private
>copies does not exist.

	Then you can't use software at all. Installing from a CD is a 
private copy. Loading into memory is a private copy. You can't mean 
what you're saying. Either you're confused or the EU is utterly 
insane.

>The one who compiles it using the kernel header files makes a
>derived work -
>the binary is a derived work. But thats my opinion. You thinks that
>it is
>using them.

	How else can you use a header file other than to include it in a 
source file that you subsequently compile. This is how header files 
are intended to be used. This is like making a copy of a book on your 
retina. It's the only way to use it, so it *must* be use.

	These are the arguments the 'free software' (as in freedom) crowd 
should be making, not opposing.

	DS

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Steven Barnhart]

On Mon, 2003-01-06 at 17:04, David Schwartz wrote:
> 	You can aim this criticism at many other people in this 
> conversation, but not me. I think I'm the only one who does 
> differentiate clearly.	
> 
> >Say you buy a book. Reading it ist usage.
> 
> 	Really? How do you read a book? You bounce a light off it and make a 
> copy of the book on your retina, right? In other words, you use 
> things by copying them.
> 	Yes, copies *of* *that* *book*. But when you make a copy of the book 
> on your retina, your eyes and brain are not a derived work. When you 
> use photoshop, the graphics you create are not a derived work.
> 
> 	I submit that the *only* way to use a header file is to include it 
> in a source file, and compile and copy the resultant output. Note 
> that you cannot run a program without copying it. It's physically 
> impossible.
> 
> 	Right, because we recognize that a graphic created with photoshop is 
> *not* a derived work of photoshop. A brain that has read a book is 
> not a derived work of that book. Similary, a program whose source 
> code includes a header file should not be considered a derived work 
> of that header file.
Amazing how the topic changes so differently when it origionated as a
battle against NVidia. Amazing how sick I am of recieveing this topic in
my mailbox. Please end it somehow..there's no need for it anymore. It
has been concluded that we don't have to listen to rms if we don't want
too and that half of us will use proprietary software if we want to or
if their is no good free replacement. Ending it there.
-- 
Steven
sbarn03@softhome.net
GnuPG Fingerprint: 9357 F403 B0A1 E18D 86D5  2230 BB92 6D64 D516 0A94

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Georg Nikodym]

[-- Attachment #1: Type: text/plain, Size: 523 bytes --]

On Mon, 6 Jan 2003 14:04:15 -0800
David Schwartz <davids@webmaster.com> wrote:
 
> 	So tell me how else you use kernel header files. What else can
> 	you 
> do with a header file?!

Just some ideas:

1. Put 'em on t-shirts.
2. Read aloud to non-technical people as a cure for insonmia.
3. As song lyrics (just like the DeCSS decoder song).
4. Put 'em on wallpaper.
5. Print 'em on toiletpaper.
6. Print, shred and dispose as a was of disinforming dumpster divers.

I'm sure that the community can come up with more :-)

-g

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Mike Galbraith]

At 10:53 AM 1/7/2003 -0500, Georg Nikodym wrote:
>On Mon, 6 Jan 2003 14:04:15 -0800
>David Schwartz <davids@webmaster.com> wrote:
>
> >       So tell me how else you use kernel header files. What else can
> >       you
> > do with a header file?!
>
>Just some ideas:
>
>1. Put 'em on t-shirts.
>2. Read aloud to non-technical people as a cure for insonmia.
>3. As song lyrics (just like the DeCSS decoder song).
>4. Put 'em on wallpaper.
>5. Print 'em on toiletpaper.
>6. Print, shred and dispose as a was of disinforming dumpster divers.
>
>I'm sure that the community can come up with more :-)

These are great ideas.  I'm too busy chuckling to think up more... and :-/ 
I hope I won't be bored enough after I'm done laughing to try 8)

         -Mike

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Henning P. Schmiedehausen]

Wolfgang Walter <ml-linux-kernel@studentenwerk.mhn.de> writes:

>do not keep a copy. As you can do with microsoft windows - if you bought it 
>(and did not licensed it from microsoft). I.a. it is not necessary to provide 

Wrong. You buy a license from Microsoft and a media (CD/DVD). Read the
end user license agreement (EULA).  You're allowed to resell the media
but not the license. "Unfortunately" for Microsoft, this distinction
is illegal in free countries like Germany [1]. So M$ lost in court and
you can legally buy "OEM" versions for a fraction of the "boxed" price
and resell your licenses.

AFAIK, in the U.S. you cannot resell the license legally once you
accepted the EULA (i.e. opened the box). And dealers must not resell
unbundled OEM software.

Now, who's living in a free country again?

	Regards
		Henning

[1] The current german governement is actively trying to change this
part of the german law. That's what you get for electing morons who
might look good on TV. At this point, we're already on U.S. levels.

-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     hps@intermeta.de

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   info@intermeta.de
D-91054 Buckenhof     Fax.: 09131 / 50654-20   

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[David Schwartz]

On Sun, 5 Jan 2003 21:46:37 +0000 (UTC), Henning P. Schmiedehausen 
wrote:

>Wrong. You buy a license from Microsoft and a media (CD/DVD). Read
>the
>end user license agreement (EULA).  You're allowed to resell the
>media
>but not the license. "Unfortunately" for Microsoft, this distinction
>is illegal in free countries like Germany [1]. So M$ lost in court
>and
>you can legally buy "OEM" versions for a fraction of the "boxed"
>price
>and resell your licenses.

>AFAIK, in the U.S. you cannot resell the license legally once you
>accepted the EULA (i.e. opened the box). And dealers must not resell
>unbundled OEM software.

>Now, who's living in a free country again?

	We are. Freedom includes the freedom to set whatever terms you want 
on what other people do with what is yours. Of course, it also 
includes the freedom not to buy things that come with restrictive 
licenses.

	DS

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Henning P. Schmiedehausen]

Richard Stallman <rms@gnu.org> writes:

> [...]  They say that open source usually leads to more powerful and
> reliable software. [...]

Who are "they" and how do "they" put proof to this claim? This is one
of the biggest lies of the open source community that I read again and
again.

Most of the open source software works "about so". Do we have a bullet
proof USB stack? WLAN drivers? DHCP code? ACPI? APM (which is about
what? eight years old?)?

Open Source code most of the times is a collection of code of various
quality from people who needed to "scratch an itch" or put a hack in
to "support just my configuration".  

If you find a well designed and completely specified and developed
piece of open source software, you're almost sure to find a company or
an individual having been paid for developing it and the putting it
into open source.

Some open source code still gives me nightmares if I only think about
the filenames.

	Regards
		Henning

-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     hps@intermeta.de

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   info@intermeta.de
D-91054 Buckenhof     Fax.: 09131 / 50654-20   

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Alan Cox]

On Sun, 2003-01-05 at 21:17, Henning P. Schmiedehausen wrote:
> Most of the open source software works "about so". Do we have a bullet
> proof USB stack? WLAN drivers? DHCP code? ACPI? APM (which is about
> what? eight years old?)?

USB is getting there, but its certainly better than some others
WLAN yes - openap is superb stuff
DHCP - yes
ACPI - very recently become a truely open project so will I hope now
improve
APM - reliable for years, bios code (the nonfree bit) often very buggy

> If you find a well designed and completely specified and developed
> piece of open source software, you're almost sure to find a company or
> an individual having been paid for developing it and the putting it
> into open source.

I don't think its that clear. We have some extremely classy code done
for fun, or because people had the hardware, and some horrible code
people were paid to write.

Good code is about good engineers, and good engineers do things for many
different reasons and motivations. 

Alan

-- 
Alan Cox <alan@lxorguk.ukuu.org.uk>

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Henning P. Schmiedehausen]

Alan Cox <alan@lxorguk.ukuu.org.uk> writes:

>WLAN yes - openap is superb stuff

I didn't mention an open source access point. I already have a tried
and true one in hardware from Lucent. I meant a "driver which doesn't
lock up if it meets the WLAN card in an unusual configuration like
say, on an PCI/PCMCIA bridge in a desktop computer (yes, Windows 2000
screwed this one up, too. But they managed to fix it in SP1).

Or really supports all of the 802.11b power management modes if the
card is in managed mode (the WLAN card in my lap top sucks tremendous
amounts of current even if I don't do any data transfers. Under
windows the card goes to sleep and needs about 5% of the power). And
yes, the access point knows how to manage the card. =:-) That was the
whole point of buying an (start-1999) $1200 access point.

>DHCP - yes

Point taken. The ISC code seems to be the standards implementation.

>ACPI - very recently become a truely open project so will I hope now
>improve

"very recently". :-)

>APM - reliable for years, bios code (the nonfree bit) often very buggy

The APM code on my Laptop still can't figure out how to display the
battery level correctly all the time (it flips to "0%" for a few
seconds every five to ten minutes), so I can't use the "shut down if
below 5%" feature of apmd or my lap top would start shutting down
every five to ten minutes. Needless to tell that the Windows 2000 APM
has no such problem. BIOS? Really? (BTW: This is an Acer 710TE, one of
the best documented Linux laptops on the net). The buttons for
controlling the brilliance and contrast of the screen work fine in
console mode but not in X11. But the Func+F<n> buttons don't work at
all, because Linux considers "Func" the same as "ALT" (it is
not). Only if I don't have a virtual console on the F<n> key, it
works. It does work in console mode, though.

Yes, I know, I can map all this to work correctly with X11 key
mappings and I actually do know how to do it. But then again, I have
21+ years of computing experience, starting with self-soldered 8085
boards. My wife e.g. does not. 

The point is: With Linux I must (I can!) do all of this for myself.
For my wife, Windows 2000 does all the grunt work. So she uses Win2k a
nd I use Linux (but I have to support the Win2k for her. :-) After all 
she's not a CS major).

>> If you find a well designed and completely specified and developed
>> piece of open source software, you're almost sure to find a company or
>> an individual having been paid for developing it and the putting it
>> into open source.

>I don't think its that clear. We have some extremely classy code done
>for fun, or because people had the hardware, and some horrible code
>people were paid to write.

>Good code is about good engineers, and good engineers do things for many
>different reasons and motivations. 

You're definitely right. I tried to polarize a little. :-) 

	Regards
		Henning


-- 
Dipl.-Inf. (Univ.) Henning P. Schmiedehausen       -- Geschaeftsfuehrer
INTERMETA - Gesellschaft fuer Mehrwertdienste mbH     hps@intermeta.de

Am Schwabachgrund 22  Fon.: 09131 / 50654-0   info@intermeta.de
D-91054 Buckenhof     Fax.: 09131 / 50654-20   

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Xavier Bestel]

Le dim 05/01/2003 à 23:45, Henning P. Schmiedehausen a écrit :
> Alan Cox <alan@lxorguk.ukuu.org.uk> writes:
> 
> >WLAN yes - openap is superb stuff
> 
> I didn't mention an open source access point. I already have a tried
> and true one in hardware from Lucent. I meant a "driver which doesn't
> lock up if it meets the WLAN card in an unusual configuration like
> say, on an PCI/PCMCIA bridge in a desktop computer (yes, Windows 2000
> screwed this one up, too. But they managed to fix it in SP1).

My router has an Orinoco on a PCMCIA/ISA bridge (and I suppose the ISA
bus is itself bridged to the PCI bus) running with a stock 2.4.something
kernel. Never had a problem. But then, an example isn't a proof.

	Xav

Re: Why is Nvidia given GPL'd code to use in non-free drivers?

[Alan Cox]

On Sun, 2003-01-05 at 22:45, Henning P. Schmiedehausen wrote:
> Alan Cox <alan@lxorguk.ukuu.org.uk> writes:
> 
> >WLAN yes - openap is superb stuff
> 
> I didn't mention an open source access point. I already have a tried

I use openap for most of my card driving. Its much more resilient.

> >ACPI - very recently become a truely open project so will I hope now
> >improve
> 
> "very recently". :-)

Serious comment - Until very very recently Intel wouldn't take community
changes. Intels focus has also been on correctness, so changes to handle
things like broken MS AML 1.0 output haven't gone in - which burns some
toshiba users for example.

> >APM - reliable for years, bios code (the nonfree bit) often very buggy
> 
> The APM code on my Laptop still can't figure out how to display the
> battery level correctly all the time (it flips to "0%" for a few
> seconds every five to ten minutes), so I can't use the "shut down if

BIOS bug

> below 5%" feature of apmd or my lap top would start shutting down
> every five to ten minutes. Needless to tell that the Windows 2000 APM
> has no such problem. BIOS? Really? (BTW: This is an Acer 710TE, one of

yes - BIOS. Most likely btw your Windows setup is using the ACPI
interface.