Mailing List Archive

On Sun, 3 Aug 2003 01:50:16 -0700
"Ted Mittelstaedt" <tedm@toybox.placo.com> wrote:

> Hi All,
>
> We need to start thinking of Quagga's copyright.
>
> As we all know, the exisiting Zebra, zebra-pj, and Quagga code is
> copyright
>
> Copyright (C) 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
>
>
>
> I recommend that this get changed to the following:
>
> Copyright (C) 2003 The Quagga Project
> Portions Copyright 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
>
> on ALL files!

Well, that sounds sensible.

> I also recommend that we place on the Quagga website the following
> statement:
>
> "All patches submitted for inclusion into Quagga become copyright
> The Quagga Project"

Hm, I am not so sure about this one. I would say some people feel uncomfortable
with giving simply away all rights (no matter what their value really is). I am
no lawyer, but I'd rather say it is important to stress that all code
contributed must match the GPL thought, which should protect the project from
copyrighted code injected and afterwards "sued out". I guess this is what you
really want to reach as a goal, right?

If you take over the copyright to the "project" you probably have to face the
project being sued, whereas if you leave the copyright at the author, he is
more or less obviously responsible in case of infringement.

I watched the zebra list quite some time, and have long awaited your steps btw.
I always found it a not-working-setup that someone involved in a commercial
company doing-the-same-thing can be head of such a project. It is dangerous for
both the project and him. I feel a bit sorry for Kunihiro because he should
have stepped back long time ago seeing that it will not work out. It has a
certain tragedy in it.

I always thought about a rewrite of the core code to avoid some of the
outstanding topics around timing issues, threading and the like, but did not
really see any chance for a discussion inside the zebra project.
But this is something for the dev-list... :-)

Regards,
Stephan

PS: congrats to John & family ...

On Sun, 3 Aug 2003, Ted Mittelstaedt wrote:

> Hi All,
>
> We need to start thinking of Quagga's copyright.
>

Agreed.

> As we all know, the exisiting Zebra, zebra-pj, and Quagga code is
> copyright
>
> Copyright (C) 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
>
>
>
> I recommend that this get changed to the following:
>
> Copyright (C) 2003 The Quagga Project
> Portions Copyright 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
>
> on ALL files!
>

I like it with one modification: When you telnet to any Quagga port, it
should show that copyright notice but then, three lines below it, it
should display (time delayed to appear as if it were being typed
manually):

"ALL YOUR BASE ARE BELONG TO US!" ;-)


I'm only joking of course. It should really say, "ALL YOUR BASE ARE
BELONG TO JOHN". ;-)


> I also recommend that we place on the Quagga website the following
> statement:
>
> "All patches submitted for inclusion into Quagga become copyright
> The Quagga Project"

Agreed.

>
> If we get a developer who wishes to retain copyright on his submission,
> we can easily issue a special dispensation to him, and add his name into
> the Copyright as a separate "portions of" line. Obviously everyone will
> be credited in the changelog, etc.

THANK YOU, THANK YOU, THANK YOU!

>
> It is important that we start the process now of establishing that there
> is a separate and distinct copyright on the "old" Zebra code, and all future
> code that is added to Quagga.

Definately.

--
John Fraizer | High-Security Datacenter Services |
President | Dedicated circuits 64k - 155M OC3 |
EnterZone, Inc | Virtual, Dedicated, Colocation |
http://www.enterzone.net/ | Network Consulting Services |

On Sun, 3 Aug 2003, Stephan von Krawczynski wrote:

> > I also recommend that we place on the Quagga website the following
> > statement:
> >
> > "All patches submitted for inclusion into Quagga become copyright
> > The Quagga Project"
>
> Hm, I am not so sure about this one. I would say some people feel uncomfortable
> with giving simply away all rights (no matter what their value really is). I am
> no lawyer, but I'd rather say it is important to stress that all code
> contributed must match the GPL thought, which should protect the project from
> copyrighted code injected and afterwards "sued out". I guess this is what you
> really want to reach as a goal, right?

That is what I had in mind.

> If you take over the copyright to the "project" you probably have to face the
> project being sued, whereas if you leave the copyright at the author, he is
> more or less obviously responsible in case of infringement.

Eeeeew... I don't think we want the project to face potential legal
action but, in any event, if Joe Blow submits code to the project and then
Joe Blow gets sued by Jim Smith who holds the copyright on the code that
Joe "borrowed" and then submitted to the project, we still have to remove
Jims code from the project in the end. It just releaves the project from
the burdon of legal proceedings. Am I right?

> I always thought about a rewrite of the core code to avoid some of the
> outstanding topics around timing issues, threading and the like, but did not
> really see any chance for a discussion inside the zebra project.
> But this is something for the dev-list... :-)

Oh, PLEASE! PLEASE! PLEASE! I for one will quite welcome work in this
area.

>
> PS: congrats to John & family ...

Thanks!

--
John Fraizer | High-Security Datacenter Services |
President | Dedicated circuits 64k - 155M OC3 |
EnterZone, Inc | Virtual, Dedicated, Colocation |
http://www.enterzone.net/ | Network Consulting Services |

On Sun, 3 Aug 2003, Ted Mittelstaedt wrote:

> Hi All,
>
> We need to start thinking of Quagga's copyright.

Its Kunihiro's, and is likely to remain so for quite a while. You'd
have to rewrite the majority of his code - not possible. And even if
it were possible, it wouldnt be a good idea. The support code in lib/
is in most places very /very/ elegant, zebra in general is very very
elegant - at least from the more abstract design POV.

> As we all know, the exisiting Zebra, zebra-pj, and Quagga code is
> copyright
>
> Copyright (C) 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
>
>
>
> I recommend that this get changed to the following:
>
> Copyright (C) 2003 The Quagga Project
> Portions Copyright 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
>
> on ALL files!

No. Cant do that. Wouldnt be ethical - the vast majority of the code
was written by Kunihiro, that which was not was still built on top of
his code. Ted, didnt we have a discussion once on ZNOG about the
ethics of GPL projects stripping copyrights from BSD code? :)

> I also recommend that we place on the Quagga website the following
> statement:
>
> "All patches submitted for inclusion into Quagga become copyright
> The Quagga Project"

Ah, but then we need a legal entity for the Quagga project, how do we
do that?

> If we get a developer who wishes to retain copyright on his
> submission, we can easily issue a special dispensation to him, and
> add his name into the Copyright as a separate "portions of" line.
> Obviously everyone will be credited in the changelog, etc.
>
> It is important that we start the process now of establishing
> that there is a separate and distinct copyright on the "old" Zebra
> code, and all future code that is added to Quagga.

Future code yes. Its all GPL anyway. There's nothing bad about having
Kuniro listed as copyright holder - remember the ZNOG discussion? :)
Over time, as people develop quagga, we can extend the copyright
notice to suit.

> While it won't matter much in the beginning, this will matter
> very much later on. If we ever have to deal with an infringement
> case, (either as defendant or plantiff) trying to go backwards from
> some future date and establish who owns what is like trying to
> unbake a cake.

I have copyright on 2 (admittedly small) files in Quagga (not zebra),
hence from a legal POV i have a stake and a claim, if needs ever be.
Quagga also has a rather large amount of work on ripd and ripngd
in, from Vincent Jardin. The copyright says Kunihiro (iirc), but he
might have a stake too.

(i'm ignoring any other zebra.org codebase copyright holders who
hence also have a stake in Quagga and who might follow on with zebra
- but i cant think of any. Unless we hear from Yasu :) )

In time, things might change.

> Image for example if someone successfully sued Kunihiro for
> infringing on their code in Zebra. We definitely want to establish
> now what code is "ours" and thus wouldn't be required to be
> removed.

Well, its fairly clear, the vast bulk is Kunihiro's.

> Ted

regards,
--
Paul Jakma paul@clubi.ie paul@jakma.org Key ID: 64A2FF6A
warning: do not ever send email to spam@dishone.st
Fortune:
Real Users find the one combination of bizarre input values that shuts
down the system for days.

On Sun, 3 Aug 2003 05:39:59 -0400 (EDT)
John Fraizer <syscow@EnterZone.Net> wrote:

> > If you take over the copyright to the "project" you probably have to face
> > the project being sued, whereas if you leave the copyright at the author,
> > he is more or less obviously responsible in case of infringement.
>
> Eeeeew... I don't think we want the project to face potential legal
> action but, in any event, if Joe Blow submits code to the project and then
> Joe Blow gets sued by Jim Smith who holds the copyright on the code that
> Joe "borrowed" and then submitted to the project, we still have to remove
> Jims code from the project in the end. It just releaves the project from
> the burdon of legal proceedings. Am I right?

Yes, this is exactly what I meant. I find it not amusing for
whoever-is-actually-head-of-the-project to deal with legal actions only because
someone he does not even know (very well) injected copyrighted code. I think it
should be clear responsibility of the contributor to deal with his "own"
contributions and their respective outcomings.
Think about the ongoings around SCO and linux. If Linus took over all copyright
from the contributors to the "project" he would probably face big problems
today. But now he can just sit back and have a look at what code SCO thinks is
its own, then name the contributor and thats about it. No problem for the
project, only a matter of replacement. And: no financial desaster lurking from
behind ...

Regards,
Stephan

>-----Original Message-----
>From: Paul Jakma [mailto:paul@dishone.st]
>Sent: Sunday, August 03, 2003 3:01 AM
>To: Ted Mittelstaedt
>Cc: Quagga Users
>Subject: Re: [Quagga-users 16] Quagga Copyright
>
>
>On Sun, 3 Aug 2003, Ted Mittelstaedt wrote:
>
>> Hi All,
>>
>> We need to start thinking of Quagga's copyright.
>
>Its Kunihiro's, and is likely to remain so for quite a while. You'd
>have to rewrite the majority of his code - not possible. And even if
>it were possible, it wouldnt be a good idea. The support code in lib/
>is in most places very /very/ elegant, zebra in general is very very
>elegant - at least from the more abstract design POV.
>

Well, you can do it the other way then:

Copyright 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
Portions Copyright (C) 2003 The Quagga Project

on the files that contain submitted modifications.

The point it that unless an establishment that Quagga is a separate and
distinct fork is made now, from a legal standpoint all submittals to
Zebra will continue to assumed to be handed over to Kunihiro's copyright.

>
>No. Cant do that. Wouldnt be ethical - the vast majority of the code
>was written by Kunihiro, that which was not was still built on top of
>his code. Ted, didnt we have a discussion once on ZNOG about the
>ethics of GPL projects stripping copyrights from BSD code? :)
>

Yes we did but the point was that GPL completely strips out the
BSD copyright verbage, replacing it with it's own. The GPL license
also is fundamentally different than BSD. I'm not advocating
stripping out Kunihiro's copyright, only supplementing it with
Quagga's, and besides that, both are GPL licenses and are identical.

>> I also recommend that we place on the Quagga website the following
>> statement:
>>
>> "All patches submitted for inclusion into Quagga become copyright
>> The Quagga Project"
>
>Ah, but then we need a legal entity for the Quagga project, how do we
>do that?
>

From Copyright law's POV, if you, Paul Jakma, begin to refer to Quagga
as separate from yourself, this creates a "legal entity" It is not
necessary to codify it by registering it under law, etc. Granted
"The Quagga Project's" copyrights are much harder to defend in a
court if a legal entity isn't registered, but they aren't dependent on
such registration.

>
>Future code yes. Its all GPL anyway. There's nothing bad about having
>Kuniro listed as copyright holder - remember the ZNOG discussion? :)
>Over time, as people develop quagga, we can extend the copyright
>notice to suit.
>

I never advocated NOT LISTING Kunihiro as a copyright holder. Please
re-read my post, I think you are missing something.

>> While it won't matter much in the beginning, this will matter
>> very much later on. If we ever have to deal with an infringement
>> case, (either as defendant or plantiff) trying to go backwards from
>> some future date and establish who owns what is like trying to
>> unbake a cake.
>
>I have copyright on 2 (admittedly small) files in Quagga (not zebra),
>hence from a legal POV i have a stake and a claim, if needs ever be.
>Quagga also has a rather large amount of work on ripd and ripngd
>in, from Vincent Jardin. The copyright says Kunihiro (iirc), but he
>might have a stake too.
>

Sorry, once the code was published carrying Kunihiro's copyright
at the top with Vincent Jardin's permission, he has lost copyright
on it. Now, you can e-mail Vincent and ask him if he formally ever
gave permission to assign copyright over to Kunihiro, if he says
no, then you can assume that you never had permission to publish
his code under zebra-pj without a copyright statement from him.
Thus, legally his copyright is preserved, and you can simply add
him as a copyright holder to the code and everything is fine.

This is the kind of thing I am wanting to see Quagga avoid.

>(i'm ignoring any other zebra.org codebase copyright holders who
>hence also have a stake in Quagga and who might follow on with zebra
>- but i cant think of any. Unless we hear from Yasu :) )
>
>In time, things might change.
>
>> Image for example if someone successfully sued Kunihiro for
>> infringing on their code in Zebra. We definitely want to establish
>> now what code is "ours" and thus wouldn't be required to be
>> removed.
>
>Well, its fairly clear, the vast bulk is Kunihiro's.
>

To us, not to a court. Lawyers make a living off such ambiguity.

Ted

>-----Original Message-----
>From: quagga-users-bounces@lists.quagga.net
>[mailto:quagga-users-bounces@lists.quagga.net]On Behalf Of Stephan von
>Krawczynski
>
>> I also recommend that we place on the Quagga website the following
>> statement:
>>
>> "All patches submitted for inclusion into Quagga become copyright
>> The Quagga Project"
>
>Hm, I am not so sure about this one. I would say some people feel
>uncomfortable
>with giving simply away all rights (no matter what their value
>really is). I am
>no lawyer, but I'd rather say it is important to stress that all code
>contributed must match the GPL thought, which should protect the
>project from
>copyrighted code injected and afterwards "sued out". I guess this
>is what you
>really want to reach as a goal, right?
>
>If you take over the copyright to the "project" you probably have
>to face the
>project being sued, whereas if you leave the copyright at the author, he is
>more or less obviously responsible in case of infringement.
>

This is correct, I probably should have clarified this somewhat.

I assumed that anyone submitting significant code to Quagga would want
their copyright on it. Thus, "taking over" the copyright would be regularly
done only on very small patches, such as #ifdef's for a particular platform,
syntax error corrections, cast corrections, etc.

Ted

(NB to vincent: see below for some discussion which relates to
your patch which hasso integrated into zebra-pj^Wquagga).

On Sun, 3 Aug 2003, Ted Mittelstaedt wrote:

> Well, you can do it the other way then:
>
> Copyright 1996, 97, 98, 99, 2000 Kunihiro Ishiguro
> Portions Copyright (C) 2003 The Quagga Project
>
> on the files that contain submitted modifications.

yes, thats possible.

but who exactly is 'The Quagga Project'? :)

> The point it that unless an establishment that Quagga is a separate
> and distinct fork is made now, from a legal standpoint all
> submittals to Zebra will continue to assumed to be handed over to
> Kunihiro's copyright.

IANAL, but that legal standing is indepedent of whether the banner
for Quagga says "Portions Copyright ... Quagga" or not.

Patches to code written kunihiro will be copyright kunihiro, /unless/
they are significant enough that the author would retain his
copyright.

In practical terms, i /suspect/ (IANAL) that this means that a patch
that adds new concepts and processes will be copyright author.
Further, such a patch would probably be implemented in new files,
which the author would have his copyright notice on.

2 good examples are Opaque LSAs and NSSA support in ospfd:

- Opaque LSA support was added with all opaque LSA related functions
in a new file (opaque_lsa.{c,h} - 2 files, but you know what i mean).
And it was then hooked into place into the rest of ospfd. The
copyright is KDD R&D Laboratories, Inc.

- NSSA support. Was added in by extending all the existing files. (eg
translation into ospf_abr, Type-7 LSA generation into ospf_lsa,
etc.). So while NSSA support is a sufficiently new and distinct
concept that was added to ospfd, by being implemented as numerous
extensions to existing code its author(s), i suspect, ceded copyright
to the existing copyright holder of the files they extended - but
i'm not at all sure on this point.

IANAL, but thats the kind of difference that lawyers/judges will
understand. Indeed, one could argue that the level of complexity of
these 2 extensions to ospfd were the factor which decided whether
they were implemented by "standalone portion + hooks to existing
code" (the former case) or "extend existing code" (the latter).

Something to bear in mind when contributing to quagga anyway. :)

> Yes we did but the point was that GPL completely strips out the BSD
> copyright verbage, replacing it with it's own.

We had this argument already :) And one school of thought is that the
BSD licence /allows/ this, as long as copyright is maintained (but
lets not get into that). I thought we were both agreed that changing
copyright is not good.

> advocating stripping out Kunihiro's copyright, only supplementing
> it with Quagga's, and besides that, both are GPL licenses and are
> identical.

But copyright is distinct from licence.

> From Copyright law's POV, if you, Paul Jakma, begin to refer to
> Quagga as separate from yourself, this creates a "legal entity"

Ah. really? interesting.

> It is not necessary to codify it by registering it under law, etc.
> Granted "The Quagga Project's" copyrights are much harder to defend
> in a court if a legal entity isn't registered, but they aren't
> dependent on such registration.

ok.

> I never advocated NOT LISTING Kunihiro as a copyright holder.
> Please re-read my post, I think you are missing something.

Possibly.

AIUI, you advocated removing his notice from the banner. (not from
files, etc..). I think that, for now, we shouldnt do this. Lets wait
for a while. If or when quagga has progressed, then we'll think about
it.

> Sorry, once the code was published carrying Kunihiro's copyright at
> the top with Vincent Jardin's permission, he has lost copyright on
> it.

Ah, eek.

> Now, you can e-mail Vincent and ask him if he formally ever gave
> permission to assign copyright over to Kunihiro, if he says no,
> then you can assume that you never had permission to publish his
> code under zebra-pj without a copyright statement from him. Thus,
> legally his copyright is preserved, and you can simply add him as a
> copyright holder to the code and everything is fine.

Is vincent subscribed? (i dont think he is). Cc'ing him.

> This is the kind of thing I am wanting to see Quagga avoid.

Well, copyright != licence though.

> To us, not to a court. Lawyers make a living off such ambiguity.

Indeed.

> Ted

regards,
--
Paul Jakma paul@clubi.ie paul@jakma.org Key ID: 64A2FF6A
warning: do not ever send email to spam@dishone.st
Fortune:
Loan-department manager: "There isn't any fine print. At these
interest rates, we don't need it."

>-----Original Message-----
>From: Paul Jakma [mailto:paul@clubi.ie]
>Sent: Monday, August 04, 2003 12:54 AM
>To: Ted Mittelstaedt

>Patches to code written kunihiro will be copyright kunihiro, /unless/
>they are significant enough that the author would retain his
>copyright.
>

You see, this is one of the things I would like us to avoid.

What your basically doing by putting patches under the Kunihiro
copyright, is using Kunihiro as a convenient legal entity to assign
copyright over to, much like the BSD people use the University
of California, Berkeley, as a convenient legal entity to assign
copyright of BSD code to. (keep in mind many people copy the
BSD license but still use their own name on the copyright, in
this case the code really isn't considered "BSD" code)

My problem with this is I think that Kunihiro is an objectionable
person to give this copyright too. We do not know what his plans
are nor what he will do in the future.

Suppose we fix all the bugs in Zebra-now-Quagga and the resultant
Quagga code becomes so attractive that a commercial entity wants
to license it. If you put all the patches in under Kunihiro's
copyright, he can license Quagga to this entity over our objections.
(ie: it would NOT be under GPL for this company)


>> advocating stripping out Kunihiro's copyright, only supplementing
>> it with Quagga's, and besides that, both are GPL licenses and are
>> identical.
>
>But copyright is distinct from licence.
>

Yes, but under the GPL, you lose just about all rights that a copyright
holder normally is given by the law. Except, that the GPL doesen't
let you off the hook on going after infringers, you have to assign
copyright over to the FSF for that, or fund it yourself.

>> From Copyright law's POV, if you, Paul Jakma, begin to refer to
>> Quagga as separate from yourself, this creates a "legal entity"
>
>Ah. really? interesting.
>

Yes, it does, although just about everywhere, that legal entity has
to be codified by registering it, in order for it to be able to
participate in contract law. But copyrights aren't contracts.
However you do need the entity to be able to enter into a contract
in order to be able to sell it's copyright.

>
>AIUI, you advocated removing his notice from the banner. (not from
>files, etc..).

No, that was John, not me.

I think that, for now, we shouldnt do this. Lets wait
>for a while. If or when quagga has progressed, then we'll think about
>it.
>
>> Sorry, once the code was published carrying Kunihiro's copyright at
>> the top with Vincent Jardin's permission, he has lost copyright on
>> it.
>
>Ah, eek.
>
>> Now, you can e-mail Vincent and ask him if he formally ever gave
>> permission to assign copyright over to Kunihiro, if he says no,
>> then you can assume that you never had permission to publish his
>> code under zebra-pj without a copyright statement from him. Thus,
>> legally his copyright is preserved, and you can simply add him as a
>> copyright holder to the code and everything is fine.
>
>Is vincent subscribed? (i dont think he is). Cc'ing him.
>
>> This is the kind of thing I am wanting to see Quagga avoid.
>
>Well, copyright != licence though.
>

Some cynical persons might say that it is interesting how the GPL
stays away from the issue of who owns what copyright on what,
precisely because the drafters wanted to avoid having The GNU
Project involved in copyright infringement actions against GPL
infringers, unless they chose to become involved. That cynical
person might comment that this allows the FSF to choose infringement
battles they know they can win easily.

Ted

On Mon, 4 Aug 2003, Ted Mittelstaedt wrote:

> You see, this is one of the things I would like us to avoid.

It would be nice to, yes.

> What your basically doing by putting patches under the Kunihiro
> copyright, is using Kunihiro as a convenient legal entity to assign
> copyright over to,

> My problem with this is I think that Kunihiro is an objectionable
> person to give this copyright too. We do not know what his plans
> are nor what he will do in the future.

Ok, fine - i agree with you! we shouldnt auto-copyright work on
quagga.

but lets first get some work done on quagga :)

> Suppose we fix all the bugs in Zebra-now-Quagga and the resultant
> Quagga code becomes so attractive that a commercial entity wants to
> license it. If you put all the patches in under Kunihiro's
> copyright, he can license Quagga to this entity over our
> objections. (ie: it would NOT be under GPL for this company)

well... thats questionable. IANAL. but yes, lets avoid that.

> Yes, but under the GPL, you lose just about all rights that a
> copyright holder normally is given by the law.

such as?

> Except, that the GPL doesen't let you off the hook on going after
> infringers,

?

> you have to assign copyright over to the FSF for that,
> or fund it yourself.

yes, of course :)

> Yes, it does, although just about everywhere, that legal entity has
> to be codified by registering it, in order for it to be able to
> participate in contract law. But copyrights aren't contracts.
> However you do need the entity to be able to enter into a contract
> in order to be able to sell it's copyright.

right. anyway, dont think we quite need that yet.

> No, that was John, not me.

ah, sorry.

> Some cynical persons might say that it is interesting how the GPL
> stays away from the issue of who owns what copyright on what,

well, those cynical people might want to consider that the GPL is a
licence granted by the copyright holder.

why should the licence discuss the issue of who owns the copyright??
(anyway, way OT now.)

[snip far out ramblings on the GPL]

> Ted

regards,
--
Paul Jakma paul@clubi.ie paul@jakma.org Key ID: 64A2FF6A
warning: do not ever send email to spam@dishone.st
Fortune:
It's fabulous! We haven't seen anything like it in the last half an hour!
-- Macy's

On Mon, 4 Aug 2003, Paul Jakma wrote:

> Ok, fine - i agree with you! we shouldnt auto-copyright work on
> quagga.

missing the appropriate ending, should read:

"we shouldnt auto-copyright work on quagga to kunihiro, absolutely."

regards,
--
Paul Jakma paul@clubi.ie paul@jakma.org Key ID: 64A2FF6A
warning: do not ever send email to spam@dishone.st
Fortune:
Old programmers never die, they just hit account block limit.