This is about Monta Vista Software, a company that develops software for
embedded platforms, based on the Linux kernel and (possibly) other GPL
software.
In its official FAQ page, http://www.mvista.com/products/faq.html#q9 , it
says:
"A: The GNU General Public License (GPL) is very specific about the
obligations imposed on developers leveraging Open Source. If you
deploy/redistribute program binaries derived from source code licensed
under the GPL, you must
Supply the source code to derived GPL code or Make an offer (good for 3
years) to supply the source code
Retain all licensing / header information, copyright notices, etc. in
those sources
Redistribute the text of the GPL with the binaries and/or source code
Note that your obligation is strictly to the recipients of binaries
(i.e., your customers). You have no responsibility to the "community" at
large."
Its the last sentence that is of concern. Does this mean no 3rd
party (who is not a customer) can get the GPL source code part of their
products ? Seems like a GPL violation of clause 3b in
http://www.gnu.org/licenses/gpl.html .
In addition:
1. There is no linux kernel source in ftp://ftp.mvista.com/
2. The download page http://www.mvista.com/previewkit/index.html does not
claim to offer any source code at all.
thanks
- Nil
IANAL and I'm sure someone <couRMSgh> will slap me around if this is
incorrect in any real way.
On Wed, 2003-02-05 at 06:58, Nilmoni Deb wrote:
> Note that your obligation is strictly to the recipients of binaries
> (i.e., your customers). You have no responsibility to the "community" at
> large."
>From the GPL faq:
http://www.gnu.org/licenses/gpl-faq.html#WhatDoesWrittenOfferValid
What does this "written offer valid for any third party" mean? Does that
mean everyone in the world can get the source to any GPL'ed program no
matter what?
"Valid for any third party" means that anyone who has the offer
is entitled to take you up on it.
If you commercially distribute binaries not accompanied with
source code, the GPL says you must provide a written offer to
distribute the source code later. When users non-commercially
redistribute the binaries they received from you, they must pass
along a copy of this written offer. This means that people who
did not get the binaries directly from you can still receive
copies of the source code, along with the written offer.
The reason we require the offer to be valid for any third party
is so that people who receive the binaries indirectly in that
way can order the source code from you.
...so if you have no binaries, you aren't entitled to the source code.
Many commercial Linux companies offer it anyway (witness Tivo, where the
sources are hard to 'click to' but google will turn up a reasonably well
organized set of GPL sources used on their devices.) They -could-
insist that you prove you are a customer; they -cannot- insist that you
prove everyone -you- give it to is a customer..
> In addition:
>
> 1. There is no linux kernel source in ftp://ftp.mvista.com/
Call them, identify yourself as a customer, -prove it- and then
complain..
> 2. The download page http://www.mvista.com/previewkit/index.html does not
> claim to offer any source code at all.
See above.. actually. Scratch that - if they are offering 'public
previews' (haven't hit the site yet) then contact them for sources...
you DID try talking to them first, BEFORE trying to start a witchhunt,
right?
> thanks
> - Nil
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
On Wed, 2003-02-05 at 06:58, Nilmoni Deb wrote:
> Its the last sentence that is of concern. Does this mean no 3rd
> party (who is not a customer) can get the GPL source code part of their
> products ? Seems like a GPL violation of clause 3b in
> http://www.gnu.org/licenses/gpl.html .
The GPL only requires that source accompanies any distributed binaries.
In other words, with distribution comes the requirement of source.
MontaVista provides all the source to all GPL components as part of
their products.
The GPL places no requirement on giving source away to the world at
large, which would be silly. If you do not distribute anything, you do
not have to make available the source. If you do, you need to give the
recipients the source as well.
That line is in the FAQ because many customers are concerned of their
obligations with GPL software.
Robert Love
On Wed, Feb 05, 2003 at 11:58:23AM +0000, Nilmoni Deb wrote:
>
> This is about Monta Vista Software, a company that develops software for
> embedded platforms, based on the Linux kernel and (possibly) other GPL
> software.
The statements arew right. Of course it's not nice of them and 'd suggest
to refuse any cooperation with mvista for every kernel hackers that cares.
It would also be cool to get that into the press for those of you that have
the contacts.. :)
On 5 Feb 2003, Robert Love wrote:
> On Wed, 2003-02-05 at 06:58, Nilmoni Deb wrote:
>
> > Its the last sentence that is of concern. Does this mean no 3rd
> > party (who is not a customer) can get the GPL source code part of their
> > products ? Seems like a GPL violation of clause 3b in
> > http://www.gnu.org/licenses/gpl.html .
>
> The GPL only requires that source accompanies any distributed binaries.
> In other words, with distribution comes the requirement of source.
> MontaVista provides all the source to all GPL components as part of
> their products.
>
> The GPL places no requirement on giving source away to the world at
> large, which would be silly. If you do not distribute anything, you do
> not have to make available the source. If you do, you need to give the
> recipients the source as well.
U missed the point. This is not the case of "If you do not distribute
anything, you do not have to make available the source." because the
vendor is distributing the binaries and source to the customers just fine.
The issue is whether any 3rd party (who is not purchasing the binaries)
has the right to get the source too as per clause 3b of
http://www.gnu.org/licenses/gpl.html . More specifically, is it true that
if u obey clause 3a, u don't have to follow clause 3b ?
>
> That line is in the FAQ because many customers are concerned of their
> obligations with GPL software.
>
> Robert Love
>
>
[Ooh, another CMU person. Hi.]
Disclaimer: this is not legal advice, and I do not speak for my
employer. I'm also not responsible for policies, so please don't vent
at me or I'll stop trying to be helpful.
On Wed, Feb 05, 2003 at 11:58:23AM +0000, Nilmoni Deb wrote:
>
> This is about Monta Vista Software, a company that develops software for
> embedded platforms, based on the Linux kernel and (possibly) other GPL
> software.
>
> In its official FAQ page, http://www.mvista.com/products/faq.html#q9 , it
> says:
>
> "A: The GNU General Public License (GPL) is very specific about the
> obligations imposed on developers leveraging Open Source. If you
> deploy/redistribute program binaries derived from source code licensed
> under the GPL, you must
>
>
> Supply the source code to derived GPL code or Make an offer (good for 3
> years) to supply the source code
>
> Retain all licensing / header information, copyright notices, etc. in
> those sources
>
> Redistribute the text of the GPL with the binaries and/or source code
>
> Note that your obligation is strictly to the recipients of binaries
> (i.e., your customers). You have no responsibility to the "community" at
> large."
>
>
>
>
>
> Its the last sentence that is of concern. Does this mean no 3rd
> party (who is not a customer) can get the GPL source code part of their
> products ? Seems like a GPL violation of clause 3b in
> http://www.gnu.org/licenses/gpl.html .
I suggest you read clause 3b again:
3. You may copy and distribute the Program (or a work based on it, under Section 2) in object
code or executable form under the terms of Sections 1 and 2 above provided that you also do
one of the following:
* b) Accompany it with a written offer, valid for at least three years, to give any third
party, for a charge no more than your cost of physically performing source distribution,
a complete machine-readable copy of the corresponding source code, to be distributed
under the terms of Sections 1 and 2 above on a medium customarily used for software
interchange; or,
If you don't distribute a binary to someone, then you are under no
obligation to distribute source to them. The GPL's always worked that
way. One can't restrict what one's customers do with the source, but
that doesn't oblige one to give it away.
> In addition:
>
> 1. There is no linux kernel source in ftp://ftp.mvista.com/
Nor any reason it should be there.
> 2. The download page http://www.mvista.com/previewkit/index.html does not
> claim to offer any source code at all.
I'm told that the preview kits do include kernel source, although I
haven't checked for myself in a couple of months.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
On Wed, 2003-02-05 at 12:18, Nilmoni Deb wrote:
> U missed the point. This is not the case of "If you do not distribute
> anything, you do not have to make available the source." because the
> vendor is distributing the binaries and source to the customers just fine.
> The issue is whether any 3rd party (who is not purchasing the binaries)
> has the right to get the source too as per clause 3b of
> http://www.gnu.org/licenses/gpl.html . More specifically, is it true that
> if u obey clause 3a, u don't have to follow clause 3b ?
No, you are missing the point.
Clause 3 says "...provided that you also do one of the following"
And 3(a) is done, as the source is included with the binary.
Further, all of the source is available anyhow in various open source
projects.
Besides, 3(b) says one must accompany the binary with a written offer.
How do you know that is not being done? It says nothing about source
available on a website. But 3(b) is not a requirement since 3(a) is
met.
Robert Love
On Wed, Feb 05, 2003 at 12:04:36PM -0500, Disconnect wrote:
> What does this "written offer valid for any third party" mean? Does that
> mean everyone in the world can get the source to any GPL'ed program no
> matter what?
> "Valid for any third party" means that anyone who has the offer
> is entitled to take you up on it.
>
> If you commercially distribute binaries not accompanied with
> source code, the GPL says you must provide a written offer to
> distribute the source code later. When users non-commercially
> redistribute the binaries they received from you, they must pass
> along a copy of this written offer. This means that people who
> did not get the binaries directly from you can still receive
> copies of the source code, along with the written offer.
Has any user of a mvista-powered embedded device ever seen this offer?..
I guess many of their customers "forgot" that
IANAL.
In fact, most of the people on this list are not lawyers, so maybe
we're not the most appropriate people to answer your query.
On Wed, Feb 05, 2003 at 11:58:23AM +0000, Nilmoni Deb wrote:
> Its the last sentence that is of concern. Does this mean no 3rd
> party (who is not a customer) can get the GPL source code part of their
> products ? Seems like a GPL violation of clause 3b in
> http://www.gnu.org/licenses/gpl.html .
Section 3 seems to apply when a binary is redistributed, and my reading
of section 3b leads me to understand that it would only apply if all
of the following are satisfied:
1. A distributed a binary image to B.
2. A accompanied it with a written offer under section 3b.
3. A did not supply the source code corresponding to that
binary as per 3a.
> In addition:
>
> 1. There is no linux kernel source in ftp://ftp.mvista.com/
>
> 2. The download page http://www.mvista.com/previewkit/index.html does not
> claim to offer any source code at all.
Does the site allow you to download a binary kernel image? If the
answer is "no" then there isn't a problem here since section 3 does
not apply (no distribution of the program in object code or
executable form).
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
Nilmoni Deb wrote:
> Its the last sentence that is of concern. Does this mean no 3rd
> party (who is not a customer) can get the GPL source code part of their
> products ? Seems like a GPL violation of clause 3b in
> http://www.gnu.org/licenses/gpl.html .
I include clause 3 in its entirety below, emphasis mine:
3. You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the following:
----------------------
* a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections 1 and
2 above on a medium customarily used for software interchange; or,
* b) Accompany it with a written offer, valid for at least three
years, to give any third party, for a charge no more than your cost of
physically performing source distribution, a complete machine-readable
copy of the corresponding source code, to be distributed under the terms
of Sections 1 and 2 above on a medium customarily used for software
interchange; or,
* c) Accompany it with the information you received as to the offer
to distribute corresponding source code. (This alternative is allowed
only for noncommercial distribution and only if you received the program
in object code or executable form with such an offer, in accord with
Subsection b above.)
Note that you only have to do one of the three. If you ship source with
the product, you're covered. If you don't ship source, you can ship a
written offer for the source. In this case, the written offer can be
redeemed by *any third party*. Finally, if you're noncommercial you can
use the third option.
Chris
--
Chris Friesen | MailStop: 043/33/F10
Nortel Networks | work: (613) 765-0557
3500 Carling Avenue | fax: (613) 765-2986
Nepean, ON K2H 8E9 Canada | email: [email protected]
>>>>> "rl" == Robert Love <[email protected]> writes:
rl> Further, all of the source is available anyhow in various open
rl> source projects.
I failed to find any source for Montavista XIP implementation, which
they claim they have put on their "Consumer Electronics" release.
Mind you, I do know that Montavista has no obbligation of giving me
that piece of source, as I haven't bought any of their products (I
would if I could gather some technical papers on their implementation
of XIP) but it would be nice if anyone knew something about this
issue.
Best regards,
andrea
--
Andrea Glorioso [email protected]
Binary Only http://www.binary-only.com/
Via A. Zanolini, 7/b Tel: +39-348.921.43.79
40126 Bologna Fax: +39-051-930.31.133
>>>>> "dj" == Daniel Jacobowitz <[email protected]> writes:
dj> I'm told that the preview kits do include kernel source,
dj> although I haven't checked for myself in a couple of months.
They do, although their preview kit is severely brain-damaged
(registration + "encrypted" ISO image + decodification of shell
scripts if you don't use Red Hat, SuSE or Mandrake).
bye,
andrea
--
Andrea Glorioso [email protected]
Binary Only http://www.binary-only.com/
Via A. Zanolini, 7/b Tel: +39-348.921.43.79
40126 Bologna Fax: +39-051-930.31.133
On Wed, Feb 05, 2003 at 06:38:42PM +0100, [email protected] wrote:
> >>>>> "dj" == Daniel Jacobowitz <[email protected]> writes:
>
> dj> I'm told that the preview kits do include kernel source,
> dj> although I haven't checked for myself in a couple of months.
>
> They do, although their preview kit is severely brain-damaged
> (registration + "encrypted" ISO image + decodification of shell
> scripts if you don't use Red Hat, SuSE or Mandrake).
So that you got the binaries now could you check where their written offer
is and post it to lkml? Afterwards anyone on this list could ask them
for their source (and at least I want to).
Btw, what do people think of a regular list where anyone who got such
an offer can post it so anyone else can get source aswell?
On Wed, 5 Feb 2003 [email protected] wrote:
> >>>>> "rl" == Robert Love <[email protected]> writes:
>
> rl> Further, all of the source is available anyhow in various open
> rl> source projects.
>
> I failed to find any source for Montavista XIP implementation, which
> they claim they have put on their "Consumer Electronics" release.
Hey, it's one of my many duties to release this XIP code to the community.
In other words, Monta Vista do actually pay me to give that source away.
It's just not up to the cleanliness level I expect of community source even
if customers usually don't care, and I didn't have time to clean it up yet.
But if you really insist I can give you a raw patch.
I find it quite saddening that a lot of people only care about making up shit
while the company in question is devoting a lot of money in salary to hire
kernel developers full time whose work has always been merged to community
trees so far.
Nicolas
>>>>> "ch" == Christoph Hellwig <[email protected]> writes:
ch> So that you got the binaries now could you check where their
ch> written offer is and post it to lkml? Afterwards anyone on
ch> this list could ask them for their source (and at least I want
ch> to).
I didn't explain myself. The preview kit contains all the relative
source, as far as I can tell (I didn't study it thoroughly yet), and
you can get it quite easily by registering on Montavista site, so you
don't need me to post the written form (which I couldn't find, by the
way, but see above about quick examination). I didn't find any source
code for their XIP implementation, though, so I'm still in the process
of verifying whether they provided it in binary form or quite simply
it's not provided in the preview kit. I think the latter option is
true, since all various Montavista offerings, CEE amongst them, all
link to the same "preview kit". Anyway, I'm in touch with a
representative from Montavista who is gathering "technical info" on
this matter
bye,
andrea
--
Andrea Glorioso [email protected]
Binary Only http://www.binary-only.com/
Via A. Zanolini, 7/b Tel: +39-348.921.43.79
40126 Bologna Fax: +39-051-930.31.133
Just a quick clarification before this thing explodes in another "ten
thousan messages" thread.
>>>>> "np" == Nicolas Pitre <[email protected]> writes:
np> On Wed, 5 Feb 2003 [email protected] wrote:
>> >>>>> "rl" == Robert Love <[email protected]> writes:
>>
rl> Further, all of the source is available anyhow in various open
rl> source projects.
>> I failed to find any source for Montavista XIP implementation,
>> which they claim they have put on their "Consumer Electronics"
>> release.
np> Hey, it's one of my many duties to release this XIP code to
np> the community. In other words, Monta Vista do actually pay me
np> to give that source away.
That's wonderful (no kidding).
np> It's just not up to the cleanliness level I expect of
np> community source even if customers usually don't care, and I
np> didn't have time to clean it up yet. But if you really insist
np> I can give you a raw patch.
It would be quite interesting for me, sincerely. Feel free to send me
the patch if you can.
np> I find it quite saddening that a lot of people only care about
np> making up shit while the company in question is devoting a lot
np> of money in salary to hire kernel developers full time whose
np> work has always been merged to community trees so far.
As far as I'm concerned, I just wrote:
- that Montavista claims they implemented XIP as part of their
"Consumer Electronics Edition" package;
- that source code is included in Montavista preview kit, but that I
couldn't find any XIP reference in it;
- that probably XIP is a part of the CEE edition but *not* of the
preview kit, hence the lack of source code for that specific part in
the ISO I downloaded; incidentally, that also means that right now I'm
not entitled to have source code for Montavista XIP implementation
(and I stated that clearly);
I personally don't have anything against Montavista, and I think
they're doing a good work (I know that you have to balance between
distributing source code and having your old-minded customers run away
in fear). I still think that having an encrypted ISO image is pure
brain damage, because it doesn't buy Montavista anything as far as I
can tell and it just slows down casual observers - which, by the way,
are the kind of people Montavista is supposed to attract. :)
Let's straighten out this issue. Maybe a public development tree of
Montavista kernels, with a big red flag saying "don't use this on
production systems" would be enough.
Best regards,
--
Andrea Glorioso [email protected]
Binary Only http://www.binary-only.com/
Via A. Zanolini, 7/b Tel: +39-348.921.43.79
40126 Bologna Fax: +39-051-930.31.133
On Wed, Feb 05, 2003 at 12:57:49PM -0500, Nicolas Pitre wrote:
> I find it quite saddening that a lot of people only care about making up shit
> while the company in question is devoting a lot of money in salary to hire
> kernel developers full time whose work has always been merged to community
> trees so far.
Blah. Why can't mvista put up the source to their products on some ftp
server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
this discussion. mvista doesn't do anything illegal, but they certainly
don't play nice.
On Wed, 2003-02-05 at 11:58, Nilmoni Deb wrote:
> Note that your obligation is strictly to the recipients of binaries
> (i.e., your customers). You have no responsibility to the "community" at
> large."
This is correct. Its actually very important. A lot of GPL software is
created by a small company for another. It would be completely unfair
for that small company to be expected to ship stuff to everyone. Their
customer may choose to but then they must distribute sources and so in
turn.
Montavista feed a fair bit of stuff back into the kernel, especially at
the mips end of the universe.
On Wed, Feb 05, 2003 at 01:17:57PM -0500, Nicolas Pitre wrote:
> Would kernel hackers care if Monta Vista stopped contributing its kernel
> development to the community entirely, but give it only to its customers?
The only mvista people that I see actively contributing are those that were
part of the linux community even before working for mvista..
> Now it should be said that Monta Vista is giving away to the community most
> of its kernel changes, and even preaching for their inclusion in Linus' tree
> or the different architecture specific projects. That's beyond the GPL
> requirements. When that doesn't happen, it's usually because the code is
> either too ugly for general acceptance, of low usefulness but to one
> specific customer, or because the customer asked not to publish, or due to a
> lack of time to process say source.
Just put up your damn source tree so people can look at it like every one else
does. That is the whole point!
Redhat does _not_ put up the source or binaries to redhat advanced
server on their web site.
THanks
-steve
Christoph Hellwig wrote:
>On Wed, Feb 05, 2003 at 12:57:49PM -0500, Nicolas Pitre wrote:
>
>
>>I find it quite saddening that a lot of people only care about making up shit
>>while the company in question is devoting a lot of money in salary to hire
>>kernel developers full time whose work has always been merged to community
>>trees so far.
>>
>>
>
>Blah. Why can't mvista put up the source to their products on some ftp
>server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
>this discussion. mvista doesn't do anything illegal, but they certainly
>don't play nice.
>
>-
>To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>the body of a message to [email protected]
>More majordomo info at http://vger.kernel.org/majordomo-info.html
>Please read the FAQ at http://www.tux.org/lkml/
>
>
>
>
>
On Wed, 5 Feb 2003, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 11:58:23AM +0000, Nilmoni Deb wrote:
> >
> > This is about Monta Vista Software, a company that develops software for
> > embedded platforms, based on the Linux kernel and (possibly) other GPL
> > software.
>
> The statements arew right. Of course it's not nice of them and 'd suggest
> to refuse any cooperation with mvista for every kernel hackers that cares.
Would kernel hackers care if Monta Vista stopped contributing its kernel
development to the community entirely, but give it only to its customers?
Customers always had the source for every GPL binaries being distributed.
That's for section 3a of the GPL so things are clear.
Now it should be said that Monta Vista is giving away to the community most
of its kernel changes, and even preaching for their inclusion in Linus' tree
or the different architecture specific projects. That's beyond the GPL
requirements. When that doesn't happen, it's usually because the code is
either too ugly for general acceptance, of low usefulness but to one
specific customer, or because the customer asked not to publish, or due to a
lack of time to process say source.
> It would also be cool to get that into the press for those of you that have
> the contacts.. :)
Childish.
Nicolas
On Wed, Feb 05, 2003 at 11:24:32AM -0700, Steven Dake wrote:
> Redhat does _not_ put up the source or binaries to redhat advanced
> server on their web site.
It does put the source on the ftp site.
On Wed, 5 Feb 2003, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 01:17:57PM -0500, Nicolas Pitre wrote:
> > Would kernel hackers care if Monta Vista stopped contributing its kernel
> > development to the community entirely, but give it only to its customers?
>
> The only mvista people that I see actively contributing are those that were
> part of the linux community even before working for mvista..
And guess why they've been hired?
> > Now it should be said that Monta Vista is giving away to the community most
> > of its kernel changes, and even preaching for their inclusion in Linus' tree
> > or the different architecture specific projects. That's beyond the GPL
> > requirements. When that doesn't happen, it's usually because the code is
> > either too ugly for general acceptance, of low usefulness but to one
> > specific customer, or because the customer asked not to publish, or due to a
> > lack of time to process say source.
>
> Just put up your damn source tree so people can look at it like every one else
> does. That is the whole point!
See my other mail for an answer to this.
Nicolas
On Wed, 5 Feb 2003, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 12:57:49PM -0500, Nicolas Pitre wrote:
> > I find it quite saddening that a lot of people only care about making up shit
> > while the company in question is devoting a lot of money in salary to hire
> > kernel developers full time whose work has always been merged to community
> > trees so far.
>
> Blah. Why can't mvista put up the source to their products on some ftp
> server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
> this discussion.
Maybe because MV is trying to avoid the financial difficulties many of the
companies quoted above have?
> mvista doesn't do anything illegal, but they certainly don't play nice.
MontaVista contribute 98% of their modifications to GPL projects back into
their respective public repositories. That's all a hacker like you should
care about. For the rest of the people in the embedded world that don't
want to gather a distribution by themselves tailored to their commercial
product then that work comes with a price.
Nicolas
On Wed, Feb 05, 2003 at 01:31:43PM -0500, Nicolas Pitre wrote:
> MontaVista contribute 98% of their modifications to GPL projects back into
> their respective public repositories. That's all a hacker like you should
> care about. For the rest of the people in the embedded world that don't
> want to gather a distribution by themselves tailored to their commercial
> product then that work comes with a price.
Anyone with a bit technical knowledge wouldn't use the shit provide by mvista
anyway. The point is to have the trees for comparism, not for actually using
them.
On Wed, Feb 05, 2003 at 06:20:25PM +0000, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 01:17:57PM -0500, Nicolas Pitre wrote:
> > Would kernel hackers care if Monta Vista stopped contributing its kernel
> > development to the community entirely, but give it only to its customers?
>
> The only mvista people that I see actively contributing are those that were
> part of the linux community even before working for mvista..
>
> > Now it should be said that Monta Vista is giving away to the community most
> > of its kernel changes, and even preaching for their inclusion in Linus' tree
> > or the different architecture specific projects. That's beyond the GPL
> > requirements. When that doesn't happen, it's usually because the code is
> > either too ugly for general acceptance, of low usefulness but to one
> > specific customer, or because the customer asked not to publish, or due to a
> > lack of time to process say source.
>
> Just put up your damn source tree so people can look at it like every one else
> does. That is the whole point!
Chris,
Nicolas submits a hell of a lot through me.
Nicolas has been around in the Linux community pre-Montavista. He
maintained the SA1100 patches, and worked to integrate them into my
patches for a fair number of years, from 2.2 right through late 2.4
times, including after he started working for Montavista. Lately, he
has become responsible for the PXA250 code.
In other words, Nicolas actively participates in the communities
concerning the areas that he works in.
The fact you don't see his submissions or participation is not something
you can or should be using to leverage your point.
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
On Wed, Feb 05, 2003 at 01:41:04PM -0500, Nicolas Pitre wrote:
> And you really think that those without that technical knowledge who are
> happy to pay for that "shit" today will continue to do so if they just could
> run away with it for free?
surprisingly people still buy RHAS or UL even if they can download the source
for free. Or buy gcc support contracts at Cygnus/RedHat even if they get the
source for free.
On 5 Feb 2003, Robert Love wrote:
> On Wed, 2003-02-05 at 12:18, Nilmoni Deb wrote:
>
> > U missed the point. This is not the case of "If you do not distribute
> > anything, you do not have to make available the source." because the
> > vendor is distributing the binaries and source to the customers just fine.
> > The issue is whether any 3rd party (who is not purchasing the binaries)
> > has the right to get the source too as per clause 3b of
> > http://www.gnu.org/licenses/gpl.html . More specifically, is it true that
> > if u obey clause 3a, u don't have to follow clause 3b ?
>
> No, you are missing the point.
I am not going to argue about that. But the last sentence of my previous
email "is it true that if u obey clause 3a, u don't have to follow clause
3b ?" was the point I wanted to clarify which wasn't answered at that
time.
> Clause 3 says "...provided that you also do one of the following"
>
> And 3(a) is done, as the source is included with the binary.
>
> Further, all of the source is available anyhow in various open source
> projects.
>
> Besides, 3(b) says one must accompany the binary with a written offer.
> How do you know that is not being done? It says nothing about source
> available on a website.
Source available on web was just an example. I just could not find a link
on the site which says something like "Apply/register to get source code
on cd/dvd/ftp/http/bla_bla media". Basically, 3b refers to 3rd party
(who is not getting/buying the binaries) and my question was about that.
> But 3(b) is not a requirement since 3(a) is
> met.
And u have answered my question now. Thanks for the clarification.
>
> Robert Love
>
>
On Wed, 5 Feb 2003, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 01:31:43PM -0500, Nicolas Pitre wrote:
> > MontaVista contribute 98% of their modifications to GPL projects back into
> > their respective public repositories. That's all a hacker like you should
> > care about. For the rest of the people in the embedded world that don't
> > want to gather a distribution by themselves tailored to their commercial
> > product then that work comes with a price.
>
> Anyone with a bit technical knowledge wouldn't use the shit provide by mvista
> anyway. The point is to have the trees for comparism, not for actually using
> them.
And you really think that those without that technical knowledge who are
happy to pay for that "shit" today will continue to do so if they just could
run away with it for free?
Yet I'm amazed about your own insistence to compare your work with "shit".
I always thought you were a highly qualified kernel hacker.
Nicolas
Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 01:31:43PM -0500, Nicolas Pitre wrote:
>
>>MontaVista contribute 98% of their modifications to GPL projects back into
>>their respective public repositories. That's all a hacker like you should
>>care about. For the rest of the people in the embedded world that don't
>>want to gather a distribution by themselves tailored to their commercial
>>product then that work comes with a price.
>
>
> Anyone with a bit technical knowledge wouldn't use the shit provide by mvista
> anyway. The point is to have the trees for comparism, not for actually using
> them.
Nice attitude. "Your shit sucks...but I want to see it all right now
anyway." No one seems to be arguing that MV is doing anything illegal,
only that they may not in all cases be bending over backwards to please
a few non-customers.
Why would anyone go out of their way to please someone who spouts
such derogatory crap about them?
--
Ben Greear <[email protected]> <Ben_Greear AT excite.com>
President of Candela Technologies Inc http://www.candelatech.com
ScryMUD: http://scry.wanfear.com http://scry.wanfear.com/~greear
On Wed, 2003-02-05 at 18:24, Steven Dake wrote:
> Redhat does _not_ put up the source or binaries to redhat advanced
> server on their web site.
The AS kernel sources are up on http://www.kernelnewbies.org/kernels/ if
anyone needs them. I think all the relevant stuff is in base kernels
nowdays. They collect kernel source sets so if anyone wants to send them
the ones they don't have - be it montavista, unitedlinux or whatever
they'll be glad to add them to the browsable set.
Alan
On Wed, Feb 05, 2003 at 07:41:46PM +0000, Alan Cox wrote:
> On Wed, 2003-02-05 at 18:24, Steven Dake wrote:
> > Redhat does _not_ put up the source or binaries to redhat advanced
> > server on their web site.
>
> The AS kernel sources are up on http://www.kernelnewbies.org/kernels/ if
> anyone needs them.
Or at ftp.redhat.com/pub/redhat/linux/enterprise (or in the updates section
for the current versions).
On Wed, 5 Feb 2003, Daniel Jacobowitz wrote:
> [Ooh, another CMU person. Hi.]
Is CMU in a black list ? Never mind ....
>
> Disclaimer: this is not legal advice, and I do not speak for my
> employer. I'm also not responsible for policies, so please don't vent
> at me or I'll stop trying to be helpful.
as u wish
>
> On Wed, Feb 05, 2003 at 11:58:23AM +0000, Nilmoni Deb wrote:
> >
> > This is about Monta Vista Software, a company that develops software for
> > embedded platforms, based on the Linux kernel and (possibly) other GPL
> > software.
> >
> > In its official FAQ page, http://www.mvista.com/products/faq.html#q9 , it
> > says:
> >
> > "A: The GNU General Public License (GPL) is very specific about the
> > obligations imposed on developers leveraging Open Source. If you
> > deploy/redistribute program binaries derived from source code licensed
> > under the GPL, you must
> >
> >
> > Supply the source code to derived GPL code or Make an offer (good for 3
> > years) to supply the source code
> >
> > Retain all licensing / header information, copyright notices, etc. in
> > those sources
> >
> > Redistribute the text of the GPL with the binaries and/or source code
> >
> > Note that your obligation is strictly to the recipients of binaries
> > (i.e., your customers). You have no responsibility to the "community" at
> > large."
> >
> >
> >
> >
> >
> > Its the last sentence that is of concern. Does this mean no 3rd
> > party (who is not a customer) can get the GPL source code part of their
> > products ? Seems like a GPL violation of clause 3b in
> > http://www.gnu.org/licenses/gpl.html .
>
> I suggest you read clause 3b again:
> 3. You may copy and distribute the Program (or a work based on it, under Section 2) in object
> code or executable form under the terms of Sections 1 and 2 above provided that you also do
> one of the following:
> * b) Accompany it with a written offer, valid for at least three years, to give any third
> party, for a charge no more than your cost of physically performing source distribution,
> a complete machine-readable copy of the corresponding source code, to be distributed
> under the terms of Sections 1 and 2 above on a medium customarily used for software
> interchange; or,
>
> If you don't distribute a binary to someone, then you are under no
> obligation to distribute source to them. The GPL's always worked that
> way. One can't restrict what one's customers do with the source, but
> that doesn't oblige one to give it away.
Actually, my question has been answered and the answer is: if the vendor
complies with 3a, then the vendor does not have to worry about 3b anymore.
> > In addition:
> >
> > 1. There is no linux kernel source in ftp://ftp.mvista.com/
>
> Nor any reason it should be there.
>
> > 2. The download page http://www.mvista.com/previewkit/index.html does not
> > claim to offer any source code at all.
>
> I'm told that the preview kits do include kernel source, although I
> haven't checked for myself in a couple of months.
>
> --
> Daniel Jacobowitz
> MontaVista Software Debian GNU/Linux Developer
On Wed, Feb 05, 2003 at 01:54:36PM -0500, Dana Lacoste wrote:
> On Wed, 2003-02-05 at 13:15, Christoph Hellwig wrote:
>
> > Blah. Why can't mvista put up the source to their products on some ftp
> > server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
> > this discussion. mvista doesn't do anything illegal, but they certainly
> > don't play nice.
>
> The source code is available to those with the binaries,
> just as the GPL requires. AFAIK everyone's said so far
> that they not only meet the GPL requirements they feed
> the changes they do make back to the community as well
> as to their customers.
everyone including me, right. I can't force them, but I can be pissed
if they don't play nice. And I am.
On 5 Feb 2003, Alan Cox wrote:
> On Wed, 2003-02-05 at 11:58, Nilmoni Deb wrote:
> > Note that your obligation is strictly to the recipients of binaries
> > (i.e., your customers). You have no responsibility to the "community" at
> > large."
>
> This is correct. Its actually very important. A lot of GPL software is
> created by a small company for another. It would be completely unfair
> for that small company to be expected to ship stuff to everyone. Their
> customer may choose to but then they must distribute sources and so in
> turn.
While one issue stands resolved (that a vendor complying with clause 3a of
GPL 2.0 does not have to comply with 3b), the GPL may have been
misprepresented by MontaVista, as per the opinion of a FSF member (Dave
Turner via RT <[email protected]>):
-------- EXCERPT STARTS ---------
> Note that your obligation is strictly to the recipients of binaries
> (i.e., your customers). You have no responsibility to the "community" at
> large."
>
>
> Its the last sentence that is of concern. Does this mean no 3rd
> party (who is not a customer) can get the GPL source code part of their
> products ?
Actually, they're wrong -- if they choose (3)(b), their offer must be
open to all third parties. And they're wrong about who their
"obligation" is to -- legally speaking, their license comes from the
copyright holder.
-------- EXCERPT ENDS ---------
> Montavista feed a fair bit of stuff back into the kernel, especially at
> the mips end of the universe.
On Wed, 5 Feb 2003, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 01:41:04PM -0500, Nicolas Pitre wrote:
> > And you really think that those without that technical knowledge who are
> > happy to pay for that "shit" today will continue to do so if they just could
> > run away with it for free?
>
> surprisingly people still buy RHAS or UL even if they can download the source
> for free.
Those distros are targetting a much wider customer base so it's less
critical if not everybody buy them. An again look at the financial
situation of many companies following the same model.
Again, if you want MontaVista work for free you just need to look at public
repositories, or search for public patch announcements on the appropriate
mailing lists. Hint: MontaVista deals with more than just x86.
> Or buy gcc support contracts at Cygnus/RedHat even if they get the
> source for free.
Oh but that's not the same. It's called consulting and MontaVista offers
that too.
Nicolas
On Wed, 2003-02-05 at 13:15, Christoph Hellwig wrote:
> Blah. Why can't mvista put up the source to their products on some ftp
> server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
> this discussion. mvista doesn't do anything illegal, but they certainly
> don't play nice.
The source code is available to those with the binaries,
just as the GPL requires. AFAIK everyone's said so far
that they not only meet the GPL requirements they feed
the changes they do make back to the community as well
as to their customers.
SO WHAT THE FUCK IS YOUR PROBLEM?
Why do you need access to a source tree
if you're not even using the product????????
It's nazi fascist BS like this that makes companies shy away from Linux.
--
Dana "hoping that the word 'nazi' kills this stupid thread off" Lacoste
Ottawa, Canada
Thanks I stand corrected, when I looked previously I couldn't find them,
but maybe I just didn't look hard enough
-steve
Christoph Hellwig wrote:
>On Wed, Feb 05, 2003 at 07:41:46PM +0000, Alan Cox wrote:
>
>
>>On Wed, 2003-02-05 at 18:24, Steven Dake wrote:
>>
>>
>>>Redhat does _not_ put up the source or binaries to redhat advanced
>>>server on their web site.
>>>
>>>
>>The AS kernel sources are up on http://www.kernelnewbies.org/kernels/ if
>>anyone needs them.
>>
>>
>
>Or at ftp.redhat.com/pub/redhat/linux/enterprise (or in the updates section
>for the current versions).
>
>
>
>
>
On Wed, Feb 05, 2003 at 02:02:38PM -0500, Nilmoni Deb wrote:
>
> On 5 Feb 2003, Alan Cox wrote:
>
> > On Wed, 2003-02-05 at 11:58, Nilmoni Deb wrote:
> > > Note that your obligation is strictly to the recipients of binaries
> > > (i.e., your customers). You have no responsibility to the "community" at
> > > large."
> >
> > This is correct. Its actually very important. A lot of GPL software is
> > created by a small company for another. It would be completely unfair
> > for that small company to be expected to ship stuff to everyone. Their
> > customer may choose to but then they must distribute sources and so in
> > turn.
>
> While one issue stands resolved (that a vendor complying with clause 3a of
> GPL 2.0 does not have to comply with 3b), the GPL may have been
> misprepresented by MontaVista, as per the opinion of a FSF member (Dave
> Turner via RT <[email protected]>):
>
> -------- EXCERPT STARTS ---------
>
> > Note that your obligation is strictly to the recipients of binaries
> > (i.e., your customers). You have no responsibility to the "community" at
> > large."
> >
> >
> > Its the last sentence that is of concern. Does this mean no 3rd
> > party (who is not a customer) can get the GPL source code part of their
> > products ?
>
> Actually, they're wrong -- if they choose (3)(b), their offer must be
> open to all third parties. And they're wrong about who their
> "obligation" is to -- legally speaking, their license comes from the
> copyright holder.
>
> -------- EXCERPT ENDS ---------
>
> > Montavista feed a fair bit of stuff back into the kernel, especially at
> > the mips end of the universe.
We don't deal with 3(b), actually. All our binary distributions
include source, a la 3(a).
It's generally considered polite to discuss your concerns with whoever
you're concerned with, instead of attempting to report them and rouse
public reaction, you know.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
On Wed, Feb 05, 2003 at 06:56:28PM +0000, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 01:54:36PM -0500, Dana Lacoste wrote:
> > On Wed, 2003-02-05 at 13:15, Christoph Hellwig wrote:
> >
> > > Blah. Why can't mvista put up the source to their products on some ftp
> > > server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
> > > this discussion. mvista doesn't do anything illegal, but they certainly
> > > don't play nice.
> >
> > The source code is available to those with the binaries,
> > just as the GPL requires. AFAIK everyone's said so far
> > that they not only meet the GPL requirements they feed
> > the changes they do make back to the community as well
> > as to their customers.
>
> everyone including me, right. I can't force them, but I can be pissed
> if they don't play nice. And I am.
I fail to see any unfairness in their methods. They fulfil their
requirements under the GPL, providing the source to their customers.
They feed back relevant and useful changes they have made to the
sources to the main development effort -- which is not a requirement
under the GPL. I would suggest that this latter effort -- which is
probably significantly _more_ effort than shoving a tarball on an FTP
server somewhere -- is precisely "playing nice", and probably a lot
more helpful to the community than simply handing you a source
tarball. It has at least the advantage that the minority-interest
hacks don't get mixed up with the important mainstream(*) changes.
If you want to see the full and exact source for the MV products,
you can always either (a) obtain a binary distribution from MV, which
then entitles you to a copy of the relevant source, or (b) find
someone who already has a copy of that source, and ask them nicely for
it.
Given your comments about the MV source code in a previous email,
I'm not sure why you want to see it anyway, if it's as bad as you
claim.
Hugo.
(*) Mainstream for the appropriate problem space.
--
=== Hugo Mills: hugo@... carfax.org.uk | darksatanic.net | lug.org.uk ===
PGP key: 1C335860 from wwwkeys.eu.pgp.net or http://www.carfax.org.uk
--- Guards! Help! We're being rescued! ---
On Wed, 5 Feb 2003, Daniel Jacobowitz wrote:
> On Wed, Feb 05, 2003 at 02:02:38PM -0500, Nilmoni Deb wrote:
> >
> > On 5 Feb 2003, Alan Cox wrote:
> >
> > > On Wed, 2003-02-05 at 11:58, Nilmoni Deb wrote:
> > > > Note that your obligation is strictly to the recipients of binaries
> > > > (i.e., your customers). You have no responsibility to the "community" at
> > > > large."
> > >
> > > This is correct. Its actually very important. A lot of GPL software is
> > > created by a small company for another. It would be completely unfair
> > > for that small company to be expected to ship stuff to everyone. Their
> > > customer may choose to but then they must distribute sources and so in
> > > turn.
> >
> > While one issue stands resolved (that a vendor complying with clause 3a of
> > GPL 2.0 does not have to comply with 3b), the GPL may have been
> > misprepresented by MontaVista, as per the opinion of a FSF member (Dave
> > Turner via RT <[email protected]>):
> >
> > -------- EXCERPT STARTS ---------
> >
> > > Note that your obligation is strictly to the recipients of binaries
> > > (i.e., your customers). You have no responsibility to the "community" at
> > > large."
> > >
> > >
> > > Its the last sentence that is of concern. Does this mean no 3rd
> > > party (who is not a customer) can get the GPL source code part of their
> > > products ?
> >
> > Actually, they're wrong -- if they choose (3)(b), their offer must be
> > open to all third parties. And they're wrong about who their
> > "obligation" is to -- legally speaking, their license comes from the
> > copyright holder.
> >
> > -------- EXCERPT ENDS ---------
> >
> > > Montavista feed a fair bit of stuff back into the kernel, especially at
> > > the mips end of the universe.
>
> We don't deal with 3(b), actually. All our binary distributions
> include source, a la 3(a).
Good to know that.
> It's generally considered polite to discuss your concerns with whoever
> you're concerned with, instead of attempting to report them and rouse
> public reaction, you know.
This is not really about anybody's feelings or politeness. Nothing
personal is involved here. I reported the matter here because I wanted to
clarify certain claims made by certain vendor. Since it involves the
linux kernel, why not ask the folks at the kernel list itself ?
And contacting the vendor may not get me an unbiased opinion.
I am glad that one major issue has been clarified. As for rousing
"public reaction", the "public" is free is get aroused (or not aroused)
and that should not deter people from bringing issues to the forefront.
thanks
- Nil
>
> --
> Daniel Jacobowitz
> MontaVista Software Debian GNU/Linux Developer
>
So...could some kind soul with the source simply post it somewhere?
MontaVista doesn't have to provide it on their site. But they can't prevent
anyone with a copy from posting it. I've got some hardware from an old
project running hardhat linux and would like to upgrade.
They used to post it and it helped. I had a "frank exchange of views" with
the MontaVista marketing folks about their change in policy a while back.
Some questions about posting the source, though...
1. Can MontaVista add separate terms to their purchase agreement that
forbid/discourage/impede publishing the source?
2. What if MV mixes GPL and proprietary code in the directories to make
it difficult to post the GPL portion?
jeff
> > dj> I'm told that the preview kits do include kernel source,
> > dj> although I haven't checked for myself in a couple of months.
> >
> So that you got the binaries now could you check where their written offer
> is and post it to lkml? Afterwards anyone on this list could ask them
> for their source (and at least I want to).
>
> Btw, what do people think of a regular list where anyone who got such
> an offer can post it so anyone else can get source aswell?
> 1. Can MontaVista add separate terms to their purchase agreement
that
> forbid/discourage/impede publishing the source?
They can change their licence agreement, but this relates to the
entire licence (e.g. switch from GPL to proprietary licence, but only
if they authored the GPL'd code.) I am not sure of the difference
between purchase agreement and licence. Retrospectively
adding/amending an existing licence when it has already been issued
with customers is another thing - it probably depends on what a court
is willing to allow. As far as I know, there is little to go on when
trying to figure out if something is legal or within their rights or
not as far as software licences go.
One thing is for sure. The law provides for copyright protection and
no more. If something is outside the scope of copyright protection
*and there is no contract* (and a licence is *not* a contract) then
those licence agreements are not enforcable at law. This is why
Microsoft's licences are not enforcable in a court of law (at least in
Australia anyway, I am not sure if they have been tested in court in
America.)
> 2. What if MV mixes GPL and proprietary code in the directories
to make
> it difficult to post the GPL portion?
Then they must honour the GPL licence terms with respect to the GPL'd
code, and it is their responsibility to comply with this.
--
James
----- Original Message -----
From: "jeff millar" <[email protected]>
To: "Christoph Hellwig" <[email protected]>
Cc: <[email protected]>
Sent: Thursday, February 06, 2003 12:11 PM
Subject: Re: Monta Vista software license terms
> So...could some kind soul with the source simply post it somewhere?
> MontaVista doesn't have to provide it on their site. But they can't
prevent
> anyone with a copy from posting it. I've got some hardware from an
old
> project running hardhat linux and would like to upgrade.
>
> They used to post it and it helped. I had a "frank exchange of
views" with
> the MontaVista marketing folks about their change in policy a while
back.
>
> Some questions about posting the source, though...
> 1. Can MontaVista add separate terms to their purchase agreement
that
> forbid/discourage/impede publishing the source?
> 2. What if MV mixes GPL and proprietary code in the directories
to make
> it difficult to post the GPL portion?
>
> jeff
>
> > > dj> I'm told that the preview kits do include kernel source,
> > > dj> although I haven't checked for myself in a couple of
months.
> > >
> > So that you got the binaries now could you check where their
written offer
> > is and post it to lkml? Afterwards anyone on this list could ask
them
> > for their source (and at least I want to).
> >
> > Btw, what do people think of a regular list where anyone who got
such
> > an offer can post it so anyone else can get source aswell?
>
> -
> To unsubscribe from this list: send the line "unsubscribe
linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
One of the key arguements against the weakness of GPL!
Switch the to OSL and then there is legal teeth!
Later,
On Wed, 5 Feb 2003, Christoph Hellwig wrote:
> On Wed, Feb 05, 2003 at 12:57:49PM -0500, Nicolas Pitre wrote:
> > I find it quite saddening that a lot of people only care about making up shit
> > while the company in question is devoting a lot of money in salary to hire
> > kernel developers full time whose work has always been merged to community
> > trees so far.
>
> Blah. Why can't mvista put up the source to their products on some ftp
> server like suse/redhat/caldera/mandrake/etc?? That's the whole point of
> this discussion. mvista doesn't do anything illegal, but they certainly
> don't play nice.
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
Andre Hedrick
LAD Storage Consulting Group
Jeff,
Since you have not signed their contract for product, you have no idea if
there are restrictions. Nor do I.
Remember "Contract Law" may restrict what their customers can do.
Andre Hedrick
LAD Storage Consulting Group
On Wed, 5 Feb 2003, jeff millar wrote:
> So...could some kind soul with the source simply post it somewhere?
> MontaVista doesn't have to provide it on their site. But they can't prevent
> anyone with a copy from posting it. I've got some hardware from an old
> project running hardhat linux and would like to upgrade.
>
> They used to post it and it helped. I had a "frank exchange of views" with
> the MontaVista marketing folks about their change in policy a while back.
>
> Some questions about posting the source, though...
> 1. Can MontaVista add separate terms to their purchase agreement that
> forbid/discourage/impede publishing the source?
> 2. What if MV mixes GPL and proprietary code in the directories to make
> it difficult to post the GPL portion?
>
> jeff
>
> > > dj> I'm told that the preview kits do include kernel source,
> > > dj> although I haven't checked for myself in a couple of months.
> > >
> > So that you got the binaries now could you check where their written offer
> > is and post it to lkml? Afterwards anyone on this list could ask them
> > for their source (and at least I want to).
> >
> > Btw, what do people think of a regular list where anyone who got such
> > an offer can post it so anyone else can get source aswell?
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
On Wed, 2003-02-05 at 18:20, Christoph Hellwig wrote:
> Just put up your damn source tree so people can look at it like every one else
> does. That is the whole point!
I'm sure if you get given the binaries by mvista (register, download,
pay whatever), request the GPL source you will be free to host it on
your own site and pay the requisite bandwidth and hosting fee's.
Problem solved.
> --
> Alex Bennee <[email protected]>
> Hackers Inc
On Thu, 2003-02-06 at 09:12, Andre Hedrick wrote:
> Jeff,
>
> Since you have not signed their contract for product, you have no idea if
> there are restrictions. Nor do I.
>
> Remember "Contract Law" may restrict what their customers can do.
But not to the GPL code part of it
On Thu, 06 Feb 2003 14:37:26 GMT, Alan Cox said:
> On Thu, 2003-02-06 at 09:12, Andre Hedrick wrote:
> > Jeff,
> >
> > Since you have not signed their contract for product, you have no idea if
> > there are restrictions. Nor do I.
> >
> > Remember "Contract Law" may restrict what their customers can do.
>
> But not to the GPL code part of it
Although a US court recently held a "thou shalt not badmouth our product"
clause in a contract illegal, I don't see where the GPL prohibits a contract
that contains a clause "the recipient agrees not to engage in actions
A, B, or C" where C is "distribute the code".
The GPL prohibits such a restriction *IN THE DISTRIBUTION LICENSE*. It
doesn't say squat about whether I agree to some restriction in some OTHER
agreement.
--
Valdis Kletnieks
Computer Systems Senior Engineer
Virginia Tech
[email protected] <[email protected]> wrote:
>
> The GPL prohibits such a restriction *IN THE DISTRIBUTION LICENSE*. It
> doesn't say squat about whether I agree to some restriction in some OTHER
> agreement.
>From the GPL version 2:
6. Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.
[...]
The critical part being, of course, "You may not impose any further
restrictions on the recipients' exercise of the rights granted herein.". But
this is offtopic. MFT and Reply-To: set.
Charles
--
-----------------------------------------------------------------------
Charles Cazabon <[email protected]>
GPL'ed software available at: http://www.qcc.ca/~charlesc/software/
-----------------------------------------------------------------------
My two bits:
1. I am eternally grateful for the contributions Montavista has
made to the gnu tools. Their recent contribution of a
multithread cabable gdbserver to gdb-5.3 is invaluable.
2. I contantly see Montavista people feeding back improvements
here and there, and they are quick to respond when I have
questions about issues like the ppc405 erratum 77 workarounds.
And I've never given them any money!
Montavista *is* playing nice. As far as I can
tell, they obey the rules, and actively feed
their improvements back to the community.
Yes, they could do more -- but they don't have to, and if
they did, it might endanger their cash flow and thus the
stream of improvements they're feeding back to the community.
Please let them be. I like what they're doing.
- Dan
--
Dan Kegel
http://www.kegel.com
http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045
On Thu, 6 Feb 2003 [email protected] wrote:
> On Thu, 06 Feb 2003 14:37:26 GMT, Alan Cox said:
> > On Thu, 2003-02-06 at 09:12, Andre Hedrick wrote:
> > > Jeff,
> > >
> > > Since you have not signed their contract for product, you have no idea if
> > > there are restrictions. Nor do I.
> > >
> > > Remember "Contract Law" may restrict what their customers can do.
> >
> > But not to the GPL code part of it
>
> Although a US court recently held a "thou shalt not badmouth our product"
> clause in a contract illegal, I don't see where the GPL prohibits a contract
> that contains a clause "the recipient agrees not to engage in actions
> A, B, or C" where C is "distribute the code".
>
> The GPL prohibits such a restriction *IN THE DISTRIBUTION LICENSE*. It
> doesn't say squat about whether I agree to some restriction in some OTHER
> agreement.
I agree, but the terms of use of their licensed or re-licensed product can
be revoked, and their product may be the enitre kit. Pieces are parts,
the sum is the whole, and where this rime is going ... blah blah ...
Their customers may be bound to some contract law we have never seen.
The more interesting part is looking into the money trail of who is the
financial backing and how much was contributed. This is where I stop.
Naming names will get somebody toasting warm and I like it cool and quiet
now.
Cheers,
Andre Hedrick
LAD Storage Consulting Group
HIGH FIVE !
On Thu, 6 Feb 2003, Dan Kegel wrote:
> My two bits:
>
> 1. I am eternally grateful for the contributions Montavista has
> made to the gnu tools. Their recent contribution of a
> multithread cabable gdbserver to gdb-5.3 is invaluable.
>
> 2. I contantly see Montavista people feeding back improvements
> here and there, and they are quick to respond when I have
> questions about issues like the ppc405 erratum 77 workarounds.
> And I've never given them any money!
>
> Montavista *is* playing nice. As far as I can
> tell, they obey the rules, and actively feed
> their improvements back to the community.
>
> Yes, they could do more -- but they don't have to, and if
> they did, it might endanger their cash flow and thus the
> stream of improvements they're feeding back to the community.
>
> Please let them be. I like what they're doing.
> - Dan
>
> --
> Dan Kegel
> http://www.kegel.com
> http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
Andre Hedrick
LAD Storage Consulting Group
On Wed, 5 Feb 2003, Nilmoni Deb wrote:
> This is not really about anybody's feelings or politeness. Nothing
> personal is involved here. I reported the matter here because I wanted to
> clarify certain claims made by certain vendor. Since it involves the
> linux kernel, why not ask the folks at the kernel list itself ?
> And contacting the vendor may not get me an unbiased opinion.
> I am glad that one major issue has been clarified. As for rousing
> "public reaction", the "public" is free is get aroused (or not aroused)
> and that should not deter people from bringing issues to the forefront.
But it is not about the Linux kernel. It is purely about interpretation of
the GPL, which would be far better addressed in other forums dealing with
either FSF, GPL, or legal issues.
This whole thing is basically one person trying to discredit a company for
not doing things the GPL doesn't require. If they distributed source they
satisfied their responsibilities, and they have none to non-customers.
--
bill davidsen <[email protected]>
CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
On Thu, 6 Feb 2003, Bill Davidsen wrote:
> On Wed, 5 Feb 2003, Nilmoni Deb wrote:
>
> > This is not really about anybody's feelings or politeness. Nothing
> > personal is involved here. I reported the matter here because I wanted to
> > clarify certain claims made by certain vendor. Since it involves the
> > linux kernel, why not ask the folks at the kernel list itself ?
> > And contacting the vendor may not get me an unbiased opinion.
> > I am glad that one major issue has been clarified. As for rousing
> > "public reaction", the "public" is free is get aroused (or not aroused)
> > and that should not deter people from bringing issues to the forefront.
>
> But it is not about the Linux kernel. It is purely about interpretation of
> the GPL, which would be far better addressed in other forums dealing with
> either FSF, GPL, or legal issues.
It is about the linux kernel as the vendor distributes its customized
linux kernel. And it is also about GPL licensing. And this forum is a good
choice since many developers of the kernel have indirectly or directly
interacted with the vendor before (a few are even employees).
> This whole thing is basically one person trying to discredit a company for
> not doing things the GPL doesn't require.
Thats a very stupid comment to make. My first post says:
"Its the last sentence that is of concern. Does this mean no 3rd
party (who is not a customer) can get the GPL source code part of their
products ? Seems like a GPL violation of clause 3b in
http://www.gnu.org/licenses/gpl.html ."
It clearly asks a question about whether it is a violation or is not, as
regards a specific clause of the license. Next time please read the post
carefully before making comments.
> If they distributed source they
> satisfied their responsibilities, and they have none to non-customers.
That point was settled long ago as is obvious from the thread. Making the
same point repeatedly is just a waste of time.
>
> --
> bill davidsen <[email protected]>
> CTO, TMR Associates, Inc
> Doing interesting things with little computers since 1979.
There is a saying in English to describe
what harrassing Montavista amounts to:
cutting off your nose to spite your face
- Dan
--
Dan Kegel
http://www.kegel.com
http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045
On Thu, 6 Feb 2003, Nilmoni Deb wrote:
>
> On Thu, 6 Feb 2003, Bill Davidsen wrote:
> > But it is not about the Linux kernel. It is purely about interpretation of
> > the GPL, which would be far better addressed in other forums dealing with
> > either FSF, GPL, or legal issues.
>
> It is about the linux kernel as the vendor distributes its customized
> linux kernel. And it is also about GPL licensing. And this forum is a good
> choice since many developers of the kernel have indirectly or directly
> interacted with the vendor before (a few are even employees).
It is no more about the kernel than a question about liability is about
auto racing just because your particular interest is in cars. The GPL
applies to many kinds of software, and your question is totally general.
> > This whole thing is basically one person trying to discredit a company for
> > not doing things the GPL doesn't require.
>
> Thats a very stupid comment to make. My first post says:
>
> "Its the last sentence that is of concern. Does this mean no 3rd
> party (who is not a customer) can get the GPL source code part of their
> products ? Seems like a GPL violation of clause 3b in
> http://www.gnu.org/licenses/gpl.html ."
>
> It clearly asks a question about whether it is a violation or is not, as
> regards a specific clause of the license. Next time please read the post
> carefully before making comments.
>
> > If they distributed source they
> > satisfied their responsibilities, and they have none to non-customers.
Yes, it clearly asks a question about the GPL in general, since it applies
equally to all covered software. All the protesting on earth will not
change the point, you are asking a license question which is not kernel
specific.
I read the post carefully, you are a non-customer who didn't get what he
wanted for free and tried to make your complaint sound like a legitimate
question. You faild.
--
bill davidsen <[email protected]>
CTO, TMR Associates, Inc
Doing interesting things with little computers since 1979.
On Sun, 9 Feb 2003, Bill Davidsen wrote:
> On Thu, 6 Feb 2003, Nilmoni Deb wrote:
>
> >
> > On Thu, 6 Feb 2003, Bill Davidsen wrote:
>
> > > But it is not about the Linux kernel. It is purely about interpretation of
> > > the GPL, which would be far better addressed in other forums dealing with
> > > either FSF, GPL, or legal issues.
> >
> > It is about the linux kernel as the vendor distributes its customized
> > linux kernel. And it is also about GPL licensing. And this forum is a good
> > choice since many developers of the kernel have indirectly or directly
> > interacted with the vendor before (a few are even employees).
>
> It is no more about the kernel than a question about liability is about
> auto racing just because your particular interest is in cars. The GPL
> applies to many kinds of software, and your question is totally general.
Ur gripe is about why I posted it in lkml and I have answered that. U can
argue that I could _also_ have posted it in other forums but thats up to
me. U can't deny that it _is_ about the kernel (but not the kernel alone).
Whether it is about other GPL software as well is irrelevant beccause I
never claimed its about the kernel alone.
>
> > > This whole thing is basically one person trying to discredit a company for
> > > not doing things the GPL doesn't require.
> >
> > Thats a very stupid comment to make. My first post says:
> >
> > "Its the last sentence that is of concern. Does this mean no 3rd
> > party (who is not a customer) can get the GPL source code part of their
> > products ? Seems like a GPL violation of clause 3b in
> > http://www.gnu.org/licenses/gpl.html ."
> >
> > It clearly asks a question about whether it is a violation or is not, as
> > regards a specific clause of the license. Next time please read the post
> > carefully before making comments.
> >
> > > If they distributed source they
> > > satisfied their responsibilities, and they have none to non-customers.
>
> Yes, it clearly asks a question about the GPL in general, since it applies
> equally to all covered software. All the protesting on earth will not
> change the point, you are asking a license question which is not kernel
> specific.
U r beating about the same bush as usual. My question _does_ apply to
other GPL software too but thats irrelevant. Because the case I
highlighted is about the kernel. Remember that the GPL also says something
about the copyright holder of the _particular_ software (which is not the
FSF for sure in case of the kernel). So the idea of discussing it in lkml
is a good idea.
> I read the post carefully, you are a non-customer who didn't get what he
> wanted for free and tried to make your complaint sound like a legitimate
> question. You faild.
I don't give a damn what u consider as "failed" or "succeeded" because my
questions were answered by more than one people here and I am happy that
the issue has been resolved. So take ur whining elsewhere.
>
> --
> bill davidsen <[email protected]>
> CTO, TMR Associates, Inc
> Doing interesting things with little computers since 1979.
>
>
On Wed, Feb 05, 2003 at 08:11:55PM -0500, jeff millar wrote:
> So...could some kind soul with the source simply post it somewhere?
> MontaVista doesn't have to provide it on their site. But they can't prevent
> anyone with a copy from posting it. I've got some hardware from an old
> project running hardhat linux and would like to upgrade.
>
> They used to post it and it helped. I had a "frank exchange of views" with
> the MontaVista marketing folks about their change in policy a while back.
>
> Some questions about posting the source, though...
> 1. Can MontaVista add separate terms to their purchase agreement that
> forbid/discourage/impede publishing the source?
Can they or do they? I know that people I've worked with have signed
broad NDAs on code they've gotten from MontaVista, though such
agreements almost certainly qualify as 'additional restrictions'
against the GPL.
--
"Love the dolphins," she advised him. "Write by W.A.S.T.E.."
On Mon, Feb 10, 2003 at 01:18:54AM -0600, Oliver Xymoron wrote:
> Can they or do they? I know that people I've worked with have signed
> broad NDAs on code they've gotten from MontaVista, though such
> agreements almost certainly qualify as 'additional restrictions'
> against the GPL.
The agreement for the trial download contains clauses that seem to
qualify as such additional restrictions and would make redistribution
illegal at least. Fortunately that agreement is rather void in Europe
anyway and I'll post the diff for the kernel tree of that demo soon.
On Mon, 2003-02-10 at 07:18, Oliver Xymoron wrote:
> > 1. Can MontaVista add separate terms to their purchase agreement that
> > forbid/discourage/impede publishing the source?
>
> Can they or do they? I know that people I've worked with have signed
> broad NDAs on code they've gotten from MontaVista, though such
> agreements almost certainly qualify as 'additional restrictions'
> against the GPL.
The GPL doesn't permit such restrictions (its an additional restriction)
but as I understand it the MontaVista packages include a lot more than
just free software, so the GPL rights probably only apply to the GPL
packages in their setup
--
Alan Cox <[email protected]>
On Mon, Feb 10, 2003 at 01:24:38PM +0000, Alan Cox wrote:
> On Mon, 2003-02-10 at 07:18, Oliver Xymoron wrote:
> > > 1. Can MontaVista add separate terms to their purchase agreement that
> > > forbid/discourage/impede publishing the source?
> >
> > Can they or do they? I know that people I've worked with have signed
> > broad NDAs on code they've gotten from MontaVista, though such
> > agreements almost certainly qualify as 'additional restrictions'
> > against the GPL.
>
> The GPL doesn't permit such restrictions (its an additional restriction)
> but as I understand it the MontaVista packages include a lot more than
> just free software, so the GPL rights probably only apply to the GPL
> packages in their setup
I certainly agree, but the problem is the NDA puts the shoe on the
other foot and now it's the customer that has to consult a lawyer or
risk a nuisance suit before proceeding. So while it may not forbid, it
certainly discourages and impedes. Let me point out that I never saw
the NDA in question but said coworker was sufficiently intimidated by
it that he was unwilling to give me a copy of the kernel and gcc
sources because of it.
Not that it ended up mattering - I managed to get stock Debian running
on the board with all the board support I needed in a matter of a
couple of days.
--
"Love the dolphins," she advised him. "Write by W.A.S.T.E.."
On Mon, 10 Feb 2003 11:42:45 -0600, Oliver Xymoron wrote:
>I certainly agree, but the problem is the NDA puts the shoe on the
>other foot and now it's the customer that has to consult a lawyer or
>risk a nuisance suit before proceeding. So while it may not forbid,
>it
>certainly discourages and impedes. Let me point out that I never saw
>the NDA in question but said coworker was sufficiently intimidated
>by
>it that he was unwilling to give me a copy of the kernel and gcc
>sources because of it.
I believe such a provision would, unfortunately, by considered
legally enforceable. The rationale would be that the rights you (the
recipient of the derived work) have under the GPL would only apply if
the distributor were bound by the GPL. The only way the distributor
could be bound by the GPL was if he or she did something that he
didn't have the right to do without the GPL to give him or her such a
right.
However, without the GPL, you already had the right to possess and
use the original work. Without the GPL, the distributor already had
the right to possess and use the original work and to create derived
works. There is no issue of distribution rights to the original work
because everyone involved started with the right to use and possess
the original work.
You don't need to assent to the GPL to receive GPL'd works. You
don't need to assent to the GPL to modify GPL'd works. So the only
question would be, do you need to assent to the GPL to distribute a
modified work even if both you and the recipient of the work already
have the right to the unmodified work and the right to create such
modifications and you own the difference between the two works.
I've researched this question, and the evidence seems to suggest
that, no, it is not an additional right. Creating a derived work is a
right. Distributing the rights to the original work is a right. But
distributing a derived work when you can already create the derived
work, do not need to distribute any rights to the original work, and
own the rights to the difference between the two, does not seem to be
an additional right to the original work. It is the simple sum of the
rights both parties already have.
Or, to put it more simply, if you can use the linux kernel and
modify the linux kernel, you have pretty much all the rights to the
linux kernel that there are, and so does everyone else. Being able to
distribute a derived work you had the right to make to someone who
already had the right to possess the original work is not an
additional right to the original work.
IANAL, but I'm fairly familiar with copyright law. I'd be quite
interested in legal citations or court precedent to the contrary.
--
David Schwartz
<[email protected]>
> On Mon, 10 Feb 2003 11:42:45 -0600, Oliver Xymoron wrote:
> >I certainly agree, but the problem is the NDA puts the shoe on the
> >other foot and now it's the customer that has to consult a lawyer or
> >risk a nuisance suit before proceeding. So while it may not forbid, >it
> >certainly discourages and impedes. Let me point out that I never saw
> >the NDA in question but said coworker was sufficiently intimidated >by
> >it that he was unwilling to give me a copy of the kernel and gcc
> >sources because of it.
> I believe such a provision would, unfortunately, by considered
> legally enforceable. The rationale would be that the rights you (the
> recipient of the derived work) have under the GPL would only apply if
> the distributor were bound by the GPL. The only way the distributor
> could be bound by the GPL was if he or she did something that he
> didn't have the right to do without the GPL to give him or her such a
> right.
The GPL gives me the right to distribute modified versions _only if I
comply with the GPL_. And GPL forbids further restrictions when
distributing.
If your bizarre interpretation was right, no software licence at all would
have any validity. In particular, I'd be more than very surprised if the
GPL was so sloppily written. It was written with the input of eminent
lawyers, after all.
IANAL, etc.
--
Dr. Horst H. von Brand User #22616 counter.li.org
Departamento de Informatica Fono: +56 32 654431
Universidad Tecnica Federico Santa Maria +56 32 654239
Casilla 110-V, Valparaiso, Chile Fax: +56 32 797513
On Tue, 11 Feb 2003 08:42:26 +0100, Horst von Brand wrote:
>>On Mon, 10 Feb 2003 11:42:45 -0600, Oliver Xymoron wrote:
>>>I certainly agree, but the problem is the NDA puts the shoe on the
>>>other foot and now it's the customer that has to consult a lawyer
>>>or
>>>risk a nuisance suit before proceeding. So while it may not
>>>forbid, >it
>>>certainly discourages and impedes. Let me point out that I never
>>>saw
>>>the NDA in question but said coworker was sufficiently intimidated
>>>>by
>>>it that he was unwilling to give me a copy of the kernel and gcc
>>>sources because of it.
>> I believe such a provision would, unfortunately, by considered
>>legally enforceable. The rationale would be that the rights you
>>(the
>>recipient of the derived work) have under the GPL would only apply
>>if
>>the distributor were bound by the GPL. The only way the distributor
>>
>>could be bound by the GPL was if he or she did something that he
>>didn't have the right to do without the GPL to give him or her such
>>a
>>right.
>The GPL gives me the right to distribute modified versions _only if
>I
>comply with the GPL_. And GPL forbids further restrictions when
>distributing.
I realize that. But that has nothing to do with what I said, which
analyzes only those rights you have without agreeing to the GPL by
virtue of the fact that you possess the work and were not subject to
any restrictions in the process of acquiring and using it.
>If your bizarre interpretation was right, no software licence at all
>would
>have any validity. In particular, I'd be more than very surprised if
>the
>GPL was so sloppily written. It was written with the input of
>eminent lawyers, after all.
Your generalization doesn't apply because of several major
differences between most software licenses and the GPL:
1) Most software licenses do not grant everyone the right to use the
work covered.
2) Most software licenses do not grant anyone the right to create
derived works.
3) Most software licenses require your assent before you can use the
covered work, in fact, most require your assent before you have the
right to possess the covered work.
However, one sticky point is that the GPL talks about 'modifying' a
work. You can create derived works without modifying the original
work and the GPL is unclear in this respect. However, I would argue
that linking to a library file is using it and including a header
file in your C code is using it. After all, there is nothing else you
can do with such files.
DS
David Schwartz <[email protected]> said:
> On Tue, 11 Feb 2003 08:42:26 +0100, Horst von Brand wrote:
> >>On Mon, 10 Feb 2003 11:42:45 -0600, Oliver Xymoron wrote:
> >>>I certainly agree, but the problem is the NDA puts the shoe on the
> >>>other foot and now it's the customer that has to consult a lawyer
> >>>or risk a nuisance suit before proceeding. So while it may not
> >>>forbid, it certainly discourages and impedes. Let me point out that
> >>>I never saw the NDA in question but said coworker was sufficiently
> >>>intimidated by it that he was unwilling to give me a copy of the
> >>>kernel and gcc sources because of it.
> >> I believe such a provision would, unfortunately, by considered
> >>legally enforceable. The rationale would be that the rights you (the
> >>recipient of the derived work) have under the GPL would only apply if
> >>the distributor were bound by the GPL. The only way the distributor
> >>could be bound by the GPL was if he or she did something that he
> >>didn't have the right to do without the GPL to give him or her such a
> >>right.
> >The GPL gives me the right to distribute modified versions _only if >I
> >comply with the GPL_. And GPL forbids further restrictions when
> >distributing.
> I realize that. But that has nothing to do with what I said, which
> analyzes only those rights you have without agreeing to the GPL by
> virtue of the fact that you possess the work and were not subject to
> any restrictions in the process of acquiring and using it.
I just don't get it. If I get sources to foo under the GPL, I can spindle
and mutilate them to my heart's content at home. But as soon as I do
distribute it, the GPL is in force. There is no "not bound by the GPL
because I'm not doing ..." and then distributing "and I wasn't bound by
GPL, so..."
> >If your bizarre interpretation was right, no software licence at all
> >would have any validity. In particular, I'd be more than very surprised
> >if the GPL was so sloppily written. It was written with the input of
> >eminent lawyers, after all.
> Your generalization doesn't apply because of several major
> differences between most software licenses and the GPL:
>
> 1) Most software licenses do not grant everyone the right to use the
> work covered.
Irrelevant.
> 2) Most software licenses do not grant anyone the right to create
> derived works.
Irrelevant.
> 3) Most software licenses require your assent before you can use the
> covered work, in fact, most require your assent before you have the
> right to possess the covered work.
Don't see the relevance here. Besides, you never "possess" anything, you
are just given permision to use.
> However, one sticky point is that the GPL talks about 'modifying' a
> work. You can create derived works without modifying the original
> work and the GPL is unclear in this respect.
Right. If I take gcc and make a C# compiler based on it, it is also GPL as
far as it is derived (i.e., a substantial ammount of code was pilfered from
gcc). No change to gcc needed, just what constitutes a derivative work.
--
Dr. Horst H. von Brand User #22616 counter.li.org
Departamento de Informatica Fono: +56 32 654431
Universidad Tecnica Federico Santa Maria +56 32 654239
Casilla 110-V, Valparaiso, Chile Fax: +56 32 797513
On Tue, 11 Feb 2003 21:42:29 +0100, Horst von Brand wrote:
>David Schwartz <[email protected]> said:
>>On Tue, 11 Feb 2003 08:42:26 +0100, Horst von Brand wrote:
>>>The GPL gives me the right to distribute modified versions _only
>>>if >I
>>>comply with the GPL_. And GPL forbids further restrictions when
>>>distributing.
>>I realize that. But that has nothing to do with what I said,
>>which
>>analyzes only those rights you have without agreeing to the GPL by
>>virtue of the fact that you possess the work and were not subject
>>to
>>any restrictions in the process of acquiring and using it.
>I just don't get it. If I get sources to foo under the GPL, I can
>spindle
>and mutilate them to my heart's content at home. But as soon as I do
>distribute it, the GPL is in force. There is no "not bound by the
>GPL
>because I'm not doing ..." and then distributing "and I wasn't bound
>by
>GPL, so..."
But you don't need to distribute the rights to the Linux kernel
because already already has the rights to the Linux kernel. And when
you say you "get sources .. under the GPL", you're implying that you
have to assent to the GPL to get sources, but you don't. The GPL is
not a shrink-wrap agreement and you don't have to assent to it just
to receive or use sources. There is no distribution issue because
everyone already has the right to possess and use every GPL'd work,
so there are no rights to distribute.
I'm puzzled whether you're actually reading and thinking about what
I'm saying, because your responses seem to indicate that either
you're not taking the time to understand me or I'm not being clear
enough.
>>>If your bizarre interpretation was right, no software licence at
>>>all
>>>would have any validity. In particular, I'd be more than very
>>>surprised
>>>if the GPL was so sloppily written. It was written with the input
>>>of
>>>eminent lawyers, after all.
>> Your generalization doesn't apply because of several major
>>differences between most software licenses and the GPL:
>>
>> 1) Most software licenses do not grant everyone the right to
use
>>the
>>work covered.
>
>Irrelevant.
>> 3) Most software licenses require your assent before you can
use
>>the
>>covered work, in fact, most require your assent before you have the
>>right to possess the covered work.
>Don't see the relevance here. Besides, you never "possess" anything,
>you are just given permision to use.
I'll explain it again, repeating myself more slowly.
Nobody signs the GPL. So the only way you can determine whether or
not someone is bound by the GPL is if they did something that they
could not have obtained the right to do other than by the GPL.
You don't need to assent to the GPL to receive GPL'd works. You
don't need to assent to the GPL to distribute rights to use a GPL'd
work (because everyone is already given that right).
So what did you do that you couldn't do without the GPL? The answer
is that you distributed a derived work to people who already had the
right to possess the original work. I am saying that that is not an
*additional* right to the *original* work. It's the simple sum of
other rights. So you don't need to assent to the GPL to get it.
The difference with other software licenses is that you *do* need to
assent to them to get the rights you already have without the GPL.
For example, with Microsoft's EULA, you don't have the right to use
the work without assenting to the EUAL, so mere use of the work can
be used as proof of assent.
As an example, suppose you and I both have some greeting card
program. I produce a greeting card that includes some graphics
included with the greeting card program. That greeting card is a
derived work. I can't distribute it to anyone I want because it
contains embedded graphics and I would be distributing those graphics
to people who had no right to them.
On the other hand, if I distributed it only to people who also owned
the greeting card program, there would be no rights question. Every
recipient of my derived work already has the rights to use and
possess the work from which it is derived. The only additional right
they need is the right to my modifications, which I can give them.
Again, the right to possess and use a derivative work when you
already have the right to possess and use the original work and the
right to make the derivative work is not an *additional* right to the
*original* work. I can't say it any clearer than that, and I welcome
any citations to law or court precedent to the contrary.
--
David Schwartz
<[email protected]>
> I'll explain it again, repeating myself more slowly.
David has a good point here. It is worth emphasizing that what you believe
to be true may not be legally true. Everyone believes certain things about
the GPL but those beliefs are just opinions.
If David is correct, and what he is saying seems plausible, then because
there is no click through, shrink wrap, or written agreement to the GPL,
if you had the rights to the software *before* it was modified then
you had those rights, end of story. And the GPL grants those rights
to everyone. This is why almost 100% of the other licenses in the world
start out with "XYZ expressly reserves all rights in the Software
not specifically granted to Licensee."
I think people should read his posting below carefully, which is why I
quoted it.
> Nobody signs the GPL. So the only way you can determine whether or
> not someone is bound by the GPL is if they did something that they
> could not have obtained the right to do other than by the GPL.
>
> You don't need to assent to the GPL to receive GPL'd works. You
> don't need to assent to the GPL to distribute rights to use a GPL'd
> work (because everyone is already given that right).
>
> So what did you do that you couldn't do without the GPL? The answer
> is that you distributed a derived work to people who already had the
> right to possess the original work. I am saying that that is not an
> *additional* right to the *original* work. It's the simple sum of
> other rights. So you don't need to assent to the GPL to get it.
>
> The difference with other software licenses is that you *do* need to
> assent to them to get the rights you already have without the GPL.
> For example, with Microsoft's EULA, you don't have the right to use
> the work without assenting to the EUAL, so mere use of the work can
> be used as proof of assent.
>
> As an example, suppose you and I both have some greeting card
> program. I produce a greeting card that includes some graphics
> included with the greeting card program. That greeting card is a
> derived work. I can't distribute it to anyone I want because it
> contains embedded graphics and I would be distributing those graphics
> to people who had no right to them.
>
> On the other hand, if I distributed it only to people who also owned
> the greeting card program, there would be no rights question. Every
> recipient of my derived work already has the rights to use and
> possess the work from which it is derived. The only additional right
> they need is the right to my modifications, which I can give them.
>
> Again, the right to possess and use a derivative work when you
> already have the right to possess and use the original work and the
> right to make the derivative work is not an *additional* right to the
> *original* work. I can't say it any clearer than that, and I welcome
> any citations to law or court precedent to the contrary.
>
> --
> David Schwartz
> <[email protected]>
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
--
---
Larry McVoy lm at bitmover.com http://www.bitmover.com/lm
On Mon, Feb 10, 2003 at 01:33:20PM -0800, David Schwartz wrote:
>
> However, without the GPL, you already had the right to possess and
> use the original work. Without the GPL, the distributor already had
> the right to possess and use the original work and to create derived
> works. There is no issue of distribution rights to the original work
> because everyone involved started with the right to use and possess
> the original work.
I don't know which legal system you're baing the above on (USA?), but
from my understanding of UK law (CDPA 1988, as ammended), it's different.
The acts of redistributing a work [ 16 (1)(b) ], or adapting a work
(i.e. derived work) [ 16 (1)(e)] are reserved, and as such lacking
some form of licence would constitute infringement.
So given the above, your argument would seem to fail wrt UK law.
Given that there are UK based authors for parts of the kernel, it would
seem that they would have a basis in law to sue for infringement.
Mind having the funds to pursue court action is another thing entirely.
DF
On Wed, 12 Feb 2003 03:25:30 +0000, Derek Fawcus wrote:
>On Mon, Feb 10, 2003 at 01:33:20PM -0800, David Schwartz wrote:
>>However, without the GPL, you already had the right to possess
>>and
>>use the original work. Without the GPL, the distributor already had
>>the right to possess and use the original work and to create
>>derived
>>works. There is no issue of distribution rights to the original
>>work
>>because everyone involved started with the right to use and possess
>>the original work.
>I don't know which legal system you're baing the above on (USA?),
>but
>from my understanding of UK law (CDPA 1988, as ammended), it's
>different.
Primarily USA, but I don't agree with your analysis of UK law.
>The acts of redistributing a work [ 16 (1)(b) ], or adapting a work
>(i.e. derived work) [ 16 (1)(e)] are reserved, and as such lacking
>some form of licence would constitute infringement.
>So given the above, your argument would seem to fail wrt UK law.
This neither says nor implies that the right to distribute the
original work and the right to distribute derived works are unique
rights to the original work. I claim they are not.
I would go further and argue that with source code, there is no
distinction between use and the the creation of derived works. The
primary way you would use a header file is to include it in a project
(thus creating a derived work). The primary way would use the GCC
source code is to compile it (this creating a derived work).
We *use* source code to create derived works and for no other
primary purpose. So I would argue that for source code, there isn't a
distinction between being able to use the work and being able to
create derived works.
I believe this is part of the reason the GPL talks about modifying
rather than creating derived works. However, the term "modifying" is,
as far as I know, not a legally precise one. Does compiling a program
"modify" it? Does linking to a library "modify" it? I don't think
anyone really knows. Section 2 tries to equate modification with
creating derived works, but then you'd have to figure out how you
could use the source code to the Linux kernel in the intended way
without compiling it.
The puzzling thing is the GPL refers to things like "the act of
running the program", though it's not clear how you could run most
programs without compiling them first. If compiling them is
unrestricted, then so is creating derived works.
IANAL, but I know that nobody knows the answer to these questions.
DS
David Schwartz <[email protected]> said:
[...]
> But you don't need to distribute the rights to the Linux kernel
> because already already has the rights to the Linux kernel.
Yes, because they got them from someone.
> And when
> you say you "get sources .. under the GPL", you're implying that you
> have to assent to the GPL to get sources, but you don't. The GPL is
> not a shrink-wrap agreement and you don't have to assent to it just
> to receive or use sources.
Right.
> There is no distribution issue because
> everyone already has the right to possess and use every GPL'd work,
> so there are no rights to distribute.
Maybe putting a binary up on FTP or giving it to you on a CD can't be taken
legally as agreeing to GPL, but I'd consider that rather far-fetched. AFAIU,
if I have a copyright on something I have the legal right to stop anybody
from using it against my will, a license is just an expression of what I
allow you to do. GPL is a license using copyright, it just gives everybody
the right to do as they please and only restricts redistribution.
In any case, IANAL. And I believe you aren't either, so this discussion is
irrelevant. Get a lawyer versed in this stuff to check out your theory.
--
Dr. Horst H. von Brand User #22616 counter.li.org
Departamento de Informatica Fono: +56 32 654431
Universidad Tecnica Federico Santa Maria +56 32 654239
Casilla 110-V, Valparaiso, Chile Fax: +56 32 797513
David Schwartz wrote:
>
> On the other hand, if I distributed it only to people who also owned
> the greeting card program, there would be no rights question. Every
> recipient of my derived work already has the rights to use and
> possess the work from which it is derived. The only additional right
> they need is the right to my modifications, which I can give them.
Don't they only have the right to use their own copy and not yours???
Mark
David Schwartz wrote the following with regards to whether a company can
restrict distribution of modified GPL'd code through an NDA:
> Nobody signs the GPL. So the only way you can determine whether or
> not someone is bound by the GPL is if they did something that they
> could not have obtained the right to do other than by the GPL.
Okay, good so far.
> You don't need to assent to the GPL to receive GPL'd works. You
> don't need to assent to the GPL to distribute rights to use a GPL'd
> work (because everyone is already given that right).
Again true. Everyone has the right to use a GPL'd work, so nobody has
to *distribute* that right.
> So what did you do that you couldn't do without the GPL? The answer
> is that you distributed a derived work to people who already had the
> right to possess the original work. I am saying that that is not an
> *additional* right to the *original* work. It's the simple sum of
> other rights. So you don't need to assent to the GPL to get it.
The problem that I see here is that the distributor is shipping the
modified work as a whole.
If someone were to distribute a proprietary patch to a GPL'd piece of
software along with instructions on how to apply it, then that patch and
those instructions could be covered under an NDA. However, as soon as
they distribute binaries compiled from patched code, then they are
liable under the GPL because they are distributing work derived from a
GPL'd work. The GPL tries to address this in section 2, but I don't
think that it actually has the power to restrict the distribution of
modifications (ie patches) as long as the derived work is not
distributed since all I'm doing is giving you instructions on what
changes to make to something that you have the right to modify.
In the case of linux, I could have a private patch and keep it secret
and ship it to customers under NDA with instructions on how to apply it.
As long as I do not ship it with the linux source or ship precompiled
linux binaries then the GPL does not apply. It would be a pretty
disgusting thing to do, but I think it would be technically legal.
> As an example, suppose you and I both have some greeting card
> program. I produce a greeting card that includes some graphics
> included with the greeting card program. That greeting card is a
> derived work. I can't distribute it to anyone I want because it
> contains embedded graphics and I would be distributing those graphics
> to people who had no right to them.
Usually those programs do give you the right to redistribute works
created using the program, so your example is somewhat spurious.
> Again, the right to possess and use a derivative work when you
> already have the right to possess and use the original work and the
> right to make the derivative work is not an *additional* right to the
> *original* work. I can't say it any clearer than that, and I welcome
> any citations to law or court precedent to the contrary.
That statement seems to be true as far as it goes. However, having the
right to posess and use the original work, and the right to make a
derivative work does *not* give you the right to distribute the derived
work. For that you fall under the GPL. As I said earlier however, you
certainly have the right to distribute modifications under any license
you want, *as long as you do not distribute the modified work itself*.
As soon as you distribute the derived work, you fall under the GPL.
This means that the ultimate end-user would have to be the one applying
the patches and compiling the derived work.
Like I said earlier, this would totally bypass the purpose of the GPL
and I would be repulsed by a company that did this.
Chris
--
Chris Friesen | MailStop: 043/33/F10
Nortel Networks | work: (613) 765-0557
3500 Carling Avenue | fax: (613) 765-2986
Nepean, ON K2H 8E9 Canada | email: [email protected]
On Wed, 12 Feb 2003 10:32:39 -0500, Chris Friesen wrote:
>David Schwartz wrote the following with regards to whether a company
>can restrict distribution of modified GPL'd code through an NDA:
I apologize for continuing this. This will be my last response on
the issue unless anyone cites actual case law.
>The problem that I see here is that the distributor is shipping the
>modified work as a whole.
Yes, but only to people who already have the rights to use and
possess the original work. And only by people who already have the
right to distribute the unmodified work.
What I'm saying is that if I want to send a derived work to you, and
the following conditions all apply, I already have all the rights I
need:
1) I have all the rights to the difference between the original work
and the derived work. (Since I made the derived work, I must.)
2) You have the right to possess and use the original work. (The GPL
grants everyone the right to possess and use GPL'd works without
agreeing to the GPL's requirements.)
3) I have the right to distribute the original work. (The GPL
permits verbatim distribution.)
4) I have the right to make the derived work. (The GPL must permit
such things as they are the normal way source code is used.)
Since all four of the conditions above apply without agreeing to the
GPL's source disclosure terms, you can't infer agreement with the GPL
based solely on the distribution of the modified work.
>If someone were to distribute a proprietary patch to a GPL'd piece
>of
>software along with instructions on how to apply it, then that patch
>and
>those instructions could be covered under an NDA. However, as soon
>as
>they distribute binaries compiled from patched code, then they are
>liable under the GPL because they are distributing work derived from
>a GPL'd work.
Under what theory? Under the theory that they didn't have the right
to distribute the original work? The GPL gives them that right. Under
the theory that they didn't have the right to make the derived work?
>In the case of linux, I could have a private patch and keep it
>secret
>and ship it to customers under NDA with instructions on how to apply
>it.
>As long as I do not ship it with the linux source or ship
>precompiled
>linux binaries then the GPL does not apply. It would be a pretty
>disgusting thing to do, but I think it would be technically legal.
You could certainly ship it with the Linux source. The GPL gives you
the right to distribute the covered work unmodified without source
disclosure requirements. It can't give you the right to distribute
modified versions because those rights are not its to give.
I'll try it again. To distribute a derived work, you need the
following rights:
1) The right to make and possess the derived work.
2) The right to distribute the original work.
3) The right to distribute the 'difference' between original and
derived works.
The right to distribute derived works is not an *additional* right
to the *original* work. It is subsumed under the rights to create
derived works and the right to distribute the original work. Of
course, you also need the rights to the new content in the derived
work. If anyone can cite a law or case precedent to the contrary,
please do. Otherwise, I maintain that there is no special right to
distribute derived works. This does not exist as a distinct right
under copyright law in any country I know of.
>>As an example, suppose you and I both have some greeting card
>>program. I produce a greeting card that includes some graphics
>>included with the greeting card program. That greeting card is a
>>derived work. I can't distribute it to anyone I want because it
>>contains embedded graphics and I would be distributing those
>>graphics
>>to people who had no right to them.
>Usually those programs do give you the right to redistribute works
>created using the program, so your example is somewhat spurious.
My example did not include any such right, so an example where they
do would be your example, not mine.
>>Again, the right to possess and use a derivative work when you
>>already have the right to possess and use the original work and the
>>right to make the derivative work is not an *additional* right to
>>the
>>*original* work. I can't say it any clearer than that, and I
>>welcome
>>any citations to law or court precedent to the contrary.
>That statement seems to be true as far as it goes. However, having
>the
>right to posess and use the original work, and the right to make a
>derivative work does *not* give you the right to distribute the
>derived
>work.
Correct, you need the additional right to distribute the original
work, which the GPL does give you.
>For that you fall under the GPL. As I said earlier however,
>you
>certainly have the right to distribute modifications under any
>license
>you want, *as long as you do not distribute the modified work
>itself*
I'm saying that you can distribute the modified work itself. The
only rights to the *original* work that you need to distribute a
*modified* work is the right to make the derived work and the right
to distribute the original work.
>As soon as you distribute the derived work, you fall under the GPL.
You just agreed with my argument to the contrary!
>This means that the ultimate end-user would have to be the one
>applying
>the patches and compiling the derived work.
I don't agree and my analysis above shows my reasoning.
>Like I said earlier, this would totally bypass the purpose of the
>GPL
>and I would be repulsed by a company that did this.
I'll keep my non-legal opinions about the GPL out of it.
DS
On Wed, 12 Feb 2003 12:18:39 PST, David Schwartz said:
> The right to distribute derived works is not an *additional* right
> to the *original* work. It is subsumed under the rights to create
> derived works and the right to distribute the original work. Of
> course, you also need the rights to the new content in the derived
> work. If anyone can cite a law or case precedent to the contrary,
> please do. Otherwise, I maintain that there is no special right to
> distribute derived works. This does not exist as a distinct right
> under copyright law in any country I know of.
At least in the US:
17 USC 106 says:
Sec. 106. - Exclusive rights in copyrighted works
Subject to sections 107 through 121, the owner of copyright under this title
has the exclusive rights to do and to authorize any of the following:
(1) to reproduce the copyrighted work in copies or phonorecords;
(2) to prepare derivative works based upon the copyrighted work;
So you don't even get the right to *CREATE* a derivative work unless it's
explicitly given to you. So you're back to the GPL (clause 2) -
2. You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:
You can slice-and-dice *YOUR* copy to your heart's content, but if you
distribute it, you have to abide by the conditions. Note that at least
in the US, the submission of patches is probably covered by 17 USC 107,
which is the "fair use" clause - patches are obviously "criticism or comment" ;)
--
Valdis Kletnieks
Computer Systems Senior Engineer
Virginia Tech
On Wed, 12 Feb 2003 15:46:33 -0500, [email protected] wrote:
>On Wed, 12 Feb 2003 12:18:39 PST, David Schwartz said:
[snip]
>So you don't even get the right to *CREATE* a derivative work unless
>it's
>explicitly given to you. So you're back to the GPL (clause 2) -
I already addressed this. There is no way to use source code other
than to create derivative works.
Please, explain to me how you use a header file other than to create
derivative works. Explain to me how you use the gcc source code other
than to compile it into a derivative work. For source code, there is
no distinction between use and the creation of derivative works.
Source code is a recipe. You use it to create some form of end
product, which is a derivative work. You can't have the right to use
a recipe without the right to produce and eat whatever the recipe is
for.
>2. You may modify your copy or copies of the Program or any portion
>of it, thus forming a work based on the Program, and copy and
>distribute such modifications or work under the terms of Section 1
>above, provided that you also meet all of these conditions:
>You can slice-and-dice *YOUR* copy to your heart's content, but if
>you
>distribute it, you have to abide by the conditions. Note that at
>least
>in the US, the submission of patches is probably covered by 17 USC
>107,
>which is the "fair use" clause - patches are obviously "criticism or
>comment" ;)
As I already explained, you don't need this clause to give you the
right to distribute modified works because this right is the simple
sum of rights you already have.
You already have the right to produce derivative works. You already
have the right to distribute the original work. You already have the
right to distribute the modifications. Your recipient already has the
right to use and possess the original. There is no additional right
to the original work for this section to give you. The right it seeks
to give you is the simple sum of rights you already have.
"Distributing derived works" is not a specific right under any
copyright law I know of. It's the sum of other rights. You need some
rights to the original and some rights to the derivation. I'm saying
you already have all the rights to the original you need without
section 2.
--
David Schwartz
<[email protected]>
On Wed, Feb 12, 2003 at 01:30:21PM -0800, David Schwartz wrote:
> On Wed, 12 Feb 2003 15:46:33 -0500, [email protected] wrote:
> >On Wed, 12 Feb 2003 12:18:39 PST, David Schwartz said:
>
> [snip]
>
> >So you don't even get the right to *CREATE* a derivative work unless
> >it's
> >explicitly given to you. So you're back to the GPL (clause 2) -
>
> I already addressed this. There is no way to use source code other
> than to create derivative works.
[ ... ]
> As I already explained, you don't need this clause to give you the
> right to distribute modified works because this right is the simple
> sum of rights you already have.
>
> You already have the right to produce derivative works. You already
> have the right to distribute the original work. You already have the
> right to distribute the modifications. Your recipient already has the
> right to use and possess the original. There is no additional right
> to the original work for this section to give you. The right it seeks
> to give you is the simple sum of rights you already have.
and I gave you (off list) details of why this fails wrt UK law.
16(1)(a) reserves copying, 16(1)(b) reserves redistribution,
and 16(1)(e) reserves adaption, together with copying and
redistributing adaptions.
So taken together they have removed all of the above, and require that
one has a licence to do so.
Then 50C gives back some limited rights wrt computer programs. Namely
that one can make such copies and adaptations as one needs in order to
use the program. However, importantly it does _not_ give back any
redistribution rights, so they still require a licence.
This should meet your requirement for a jurisdiction who's law limits
what your are suggesting.
Hence to redistribute a program with patches applied needs a licence.
DF
On Wed, 12 Feb 2003 13:30:21 PST, David Schwartz said:
> You already have the right to produce derivative works.
No. At least in the US, 17 USC 106 says producing a derivative right is reserved to the
copyright holder, except for the cases enumerated in 17 USC 107-121.
So if you're producing a derivative work without having gotten the rights
to do so, you're screwed in the legal sense.
Clause 2 of the GPL gives you the right to produce derivative works *IF*
you accept the conditions. Having accepted that clause, you're bound by
it - that's what makes the GPL work.
Please enumerate what *OTHER* way you are getting the right to produce a
derivative work, rather than via the GPL clause 2. (Note that this *could*
happen, if for instance code is dual-licensed and you are getting the right
via the other license).
> You already
> have the right to distribute the original work. You already have the
> right to distribute the modifications. Your recipient already has the
> right to use and possess the original. There is no additional right
> to the original work for this section to give you. The right it seeks
> to give you is the simple sum of rights you already have.
Note again that two of these rights (distribute the original, distribute
the modifications) are *NOT* ones you inherently have - you are getting them
*WITH RESTRICTIONS* on what you can and can't do (see clause 2 again).
> "Distributing derived works" is not a specific right under any
> copyright law I know of. It's the sum of other rights. You need some
It's the sum of several rights, one of which is "creating a derived work".
If you can't legally create a derived work, you can't legally distribute same.
--
Valdis Kletnieks
Computer Systems Senior Engineer
Virginia Tech
On Wed, 12 Feb 2003 16:43:10 -0500, [email protected] wrote:
>On Wed, 12 Feb 2003 13:30:21 PST, David Schwartz said:
>>You already have the right to produce derivative works.
>No. At least in the US, 17 USC 106 says producing a derivative right
>is reserved to the
>copyright holder, except for the cases enumerated in 17 USC 107-121.
>So if you're producing a derivative work without having gotten the
>rights to do so, you're screwed in the legal sense.
You have been given that right. The GPL, without clause 2, gives you
the right to use the work (see citation below). For source code, the
only way to use it is to produce derivative works.
>Clause 2 of the GPL gives you the right to produce derivative works
>*IF*
>you accept the conditions. Having accepted that clause, you're
>bound by
>it - that's what makes the GPL work.
>
>Please enumerate what *OTHER* way you are getting the right to
>produce a
>derivative work, rather than via the GPL clause 2. (Note that this
>*could*
>happen, if for instance code is dual-licensed and you are getting
>the right via the other license).
The GPL says:
"Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope. The act of
running the Program is not restricted, and the output from the
Program is covered only if its contents constitute a work based on
the Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does."
Please explain to me, say for the case of the Linux kernel source
code, how you can run it without first producing a derivative work.
Granting a person the right to *use* source code means granting them
the right to produce derivative works of that source code because
that is how source code is used.
How can I use the Linux kernel, say on an FTP site or a CD that I
ordered, without copying it onto my computer? How can I compile it
without copying it into memory?
You cannot use a C source file without modifying it. In order to
compile it, you must pass it through a preprocessor. This produces a
modified copy of the original.
Source code is like a recipe. The right to use it implicitly
includes the right to follow the recipe and eat the results because
that is how one uses a recipe. Along the way, one makes copies and
derived works. It's simply unavoidable.
This doesn't mean that all copying and modifying is automatically
allowed for all cases where you have the right to use source code.
However, it does mean that absent a restrictive agreement to the
contrary that limits your *usage* rights, you can create derived
works and you can make copies because that's how you use source code.
You can't give the derived works or copies to those who have no
right to the original work (because you can't give others rights to
code that is not yours). However, no special right to the original
work is needed to distribute derived works among those who already
have the right to use and possess the original work and make the
derived works.
In the case of the GPL, you even have the additional right to
distribute the original work. I would argue that you can distribute
derived works even without the right to distribute the original
provided all recipients have the right to use and possess the
original. But this argument isn't even needed.
>Note again that two of these rights (distribute the original,
>distribute
>the modifications) are *NOT* ones you inherently have - you are
>getting them
>*WITH RESTRICTIONS* on what you can and can't do (see clause 2
>again).
No, that is not true. The GPL says:
"1. You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any
warranty; and give any other recipients of the Program a copy of this
License along with the Program."
There are no significant restrictions placed upon the distribution
of the original work. Even if there were, it's not clear that it
would be enforceable considering that you are only distributing the
work to people who already have the right to possess and use it.
There are only a small number of cases on this point and they're
split both ways. In all the cases where the copyright holder
prevailed (such as the lawsuit against mp3.com), it was based on a
showing that such distribution foreclosed on their sales, an argument
that could not be made in the case of the GPL. So this could be
argued either way even without clause 1.
As for distributing modifications, as I have already argued at least
four times now, this is not a special right to the *original* work so
there is no need to invoke clause 2 to get that right. It is subsumed
under the right to make derivative works and the right to distribute
the original work. It is the simple sum of those two rights (except
the additional rights you need to the modifications themselves).
To make and distribute a derived work, you need certain rights to
the original work. Specifically, you need the right to make the
derived work in the first place and you need the right to distribute
the original work. I am saying that you have both of these rights
without clause 2. It is even arguable that you have them without
clause 1.
This will be the last time I repeat myself. I promise. I will not
respond to this thread unless something genuinely new comes up.
DS
David,
>> To make and distribute a derived work, you need certain rights
>> to the original work. Specifically, you need the right to make the
>> derived work in the first place and you need the right to
>> distribute the original work. I am saying that you have both of
>> these rights without clause 2. It is even arguable that you have
>> them without clause 1.
What about clause 5:
5. You are not required to accept this License, since you have not
signed it. However, nothing else grants you permission to modify
or distribute the Program or its derivative works. These actions
are prohibited by law if you do not accept this License.
Therefore, by modifying or distributing the Program (or any work
based on the Program), you indicate your acceptance of this
License to do so, and all its terms and conditions for copying,
distributing or modifying the Program or works based on it.
This clearly states that you have no rights to distribute anything
unless you accept all the terms of the GPL.
--
Dan
On Wed, 12 Feb 2003 17:04:40 -0600, Daniel Forrest wrote:
>David,
>>>To make and distribute a derived work, you need certain rights
>>>to the original work. Specifically, you need the right to make the
>>>derived work in the first place and you need the right to
>>>distribute the original work. I am saying that you have both of
>>>these rights without clause 2. It is even arguable that you have
>>>them without clause 1.
>What about clause 5:
>5. You are not required to accept this License, since you have not
>signed it. However, nothing else grants you permission to modify
>or distribute the Program or its derivative works. These actions
>are prohibited by law if you do not accept this License.
>Therefore, by modifying or distributing the Program (or any work
>based on the Program), you indicate your acceptance of this
>License to do so, and all its terms and conditions for copying,
>distributing or modifying the Program or works based on it.
>This clearly states that you have no rights to distribute anything
>unless you accept all the terms of the GPL.
I'll repeat section 1 of the GPL again:
"1. You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any
warranty; and give any other recipients of the Program a copy of this
License along with the Program."
This clearly says that you can distribute the program without
accepting the section 2 restrictions.
I previously addressed "modification". This is not a precise legal
term, and I presume it was intended to mean the production of derived
works. The GPL says:
"Activities other than copying, distribution and modification are not
covered by this License; they are outside its scope. The act of
running the Program is not restricted, and the output from the
Program is covered only if its contents constitute a work based on
the Program (independent of having been made by running the Program).
Whether that is true depends on what the Program does."
This grants the right to use the covered work. As I've argued, using
source code includes the right to produce derived works because there
is no other way to use source code. (Well, I suppose you could look
at it or paper your wall with it. But its intended purpose is to be a
recipe for the production of derived works.) It is quite clear that
the GPL was never intended to restrict anyone's ability to use the
covered works and RMS has clearly argued that the GPL is not a
shrink-wrap agreement, that is, one you must agree to in order to use
a copyrighted work.
Section 5 does say, "However, nothing else grants you permission to
modify or distribute the Program or its derivative works. These
actions are prohibited by law if you do not accept this License." I
agree with this statement. However, my argument is that you can
distribute derived works under the rights given only in the two
sections I cite above, neither of which alone or in combination
invokes the section 2 (source disclosure) requirements.
-
To preserve the linux-kernel list from having to see more of this, I
would ask anyone interested in responding to please send any further
comments directly to me. I'll respond to them to you privately, and
then if you still feel they must go to the linux-kernel list, you can
send them with my responses in one shot. This will spare the kernel
list from having to see two or three messages where zero or one would
suffice.
I'm not trying to silence anyone. You can still send your message to
the kernel list. I'm just saying once you see my response, you may
not wish to send it to the list or you may wish to send it with a
response to my response. I feel obligated to respond to most public
comments that challenge my view for fear that silence will be equated
with an inability to rebut.
--
David Schwartz
<[email protected]>
David Schwartz wrote:
> 1) I have all the rights to the difference between the original work
> and the derived work. (Since I made the derived work, I must.)
The logic of derivation does not work like that.
You don't have all the rights to the difference.
It is not a case of B = C - A. People do not own individual lines or
bytes of the source code. (Sometimes they do, but often they don't).
The "difference" between the original work and the derived work is
_itself_ a derived work, because you can't possibly *create* a diff
without deriving from the original work.
In the act of you creating the diff you must have worked from the
original GPL'd work, so the diff is derived from that, as well as from
your original work. When you write "diff -ur orig-kernel my-kernel
>file", the output from that command is a derived work of the
orig-kernel directory.
This is true even if you didn't copy any context lines from the
original tree.
That means your difference is covered by the GPL, even though you
wrote the changes.
(If you did manage to write the same diff working from other sources,
or if the diff only makes "fair use" of the original, that would be
different. But any large, complex diff such as a patch to the Linux
VM simply cannot be written without starting from the original Linux VM
code. It is a question of how well the glove fits the hand, as it were).
(If the difference was just whole files using a well defined
interface, such as a device driver or filesystem, your point seems
quite applicable though.)
-- Jamie
On Thu, 13 Feb 2003 02:21:27 +0000, Jamie Lokier wrote:
>David Schwartz wrote:
>> 1) I have all the rights to the difference between the original
>>work
>>and the derived work. (Since I made the derived work, I must.)
>The logic of derivation does not work like that.
>You don't have all the rights to the difference.
>It is not a case of B = C - A. People do not own individual lines
>or
>bytes of the source code. (Sometimes they do, but often they
>don't).
>The "difference" between the original work and the derived work is
>_itself_ a derived work, because you can't possibly *create* a diff
>without deriving from the original work.
I don't think you understand what I mean by "difference". It's not a
precise mathematical term and it doesn't refer to 'diff' output. It
means the rights to the derived work apart from the rights to the
original work.
>In the act of you creating the diff you must have worked from the
>original GPL'd work, so the diff is derived from that, as well as
>from
>your original work. When you write "diff -ur orig-kernel my-kernel
>>file", the output from that command is a derived work of the
>orig-kernel directory.
Certainly, but that's not what I'm talking about.
>This is true even if you didn't copy any context lines from the
>original tree.
>
>That means your difference is covered by the GPL, even though you
>wrote the changes.
Of course the original work is covered by the GPL and of course the
derived work as a whole is covered by the GPL to the extent that the
original work is covered by the GPL.
>(If you did manage to write the same diff working from other
>sources,
>or if the diff only makes "fair use" of the original, that would be
>different. But any large, complex diff such as a patch to the Linux
>VM simply cannot be written without starting from the original Linux
>VM
>code. It is a question of how well the glove fits the hand, as it
>were).
When I say 'difference', I don't mean the output of a 'diff'. I mean
the literal subtraction. I mean the derived work less the original
work. Or, to put it another way, the rights to the output of the
additional creative effort added by the person who made the derived
work.
>(If the difference was just whole files using a well defined
>interface, such as a device driver or filesystem, your point seems
>quite applicable though.)
Suppose you write a play and then I write an adaptation of that play
in the form of a screenplay for a motion picture. To produce a motion
picture based upon my screenplay, you would need certain specific
rights to your play and certain specific rights to my adaptations. By
"difference" I mean the rights to the adaptations themselves. I do
agree that it is nonsensical to have, say, distribution rights to the
adaptations without rights to the original.
In other words, to have distribution rights to the derived work, you
must have distribution rights to the original. You must also have
distribution rights to the "difference". This is a notional legal
thing, there does not have to exist such a "difference" as an actual
object.
Consider if I translate a book you wrote. To distribute the
translation, you need the right to distribute the original work, the
right to make the translation, and other rights to the "difference",
that is, the right to distribute the translation apart for the right
to distribute the original work.
I think you're hung up on my use of terminology. If you don't like
the word "difference" substitute "derived work apart from the
original work".
For example, if I translate your book into another language, the
thing you need to get from me to distribute the translation is what I
call the "right to distribute the difference", but of course there is
no actual "difference". Call it the "right to distribute the
translation apart from the right to distribute the original work" or
the "right to distribute the additional creative work added in the
derivation process".
It's a complex and technical legal issue, but I'm not doing anything
unusual with it. See, for example, section 103 which states:
"The copyright in a compilation or derivative work extends only to
the material contributed by the author of such work, as distinguished
from the preexisting material employed in the work, and does not
imply any exclusive right in the preexisting material."
DS
David Schwartz wrote:
> It's a complex and technical legal issue,
I agree.
> but I'm not doing anything
> unusual with it. See, for example, section 103 which states:
>
> "The copyright in a compilation or derivative work extends only to
> the material contributed by the author of such work, as distinguished
> from the preexisting material employed in the work, and does not
> imply any exclusive right in the preexisting material."
This seems to be the realm of the "user does the link" and "user does
the patch" discussions from many years ago on gnu.misc.discuss.
Clarity wasn't forthcoming then, either.
(Personally I suspect it would be a long and complicated argument in
court, questioning awkward things like intent of the law, intent of
the authors, quid pro quo and other considerations if the parties were
willing to pursue it all the way. But I know very little about law!)
-- Jamie