Basic legal question: Publication of a fix to psgml

[Florian v. Savigny]

Dear listmates,

although this is a legal question, I am confident enough that most
people know the answer. It is this:

Some months ago, I fixed the "old-style backquotes" problem in psgml
(which was fairly straightforward, even if I did not even know what
backquotes are about), and I had inquired about this on this list
beforehand. Back then, Stefan Monnier suggested that we should track
down the other contributors to find out about copyright to see if the
fix can be put on ELPA. We have tried this, but many of them do not
seem to be available.

Now, my question is this: Given that Lennart released it under the GPL
(there is no separate LICENSE file, but all the .el files contain the
standard sentences to that effect), do I actually need any previous
contributors' permission to make the fix publicly available? (The
problem is not that I do not want to ask them, but rather that it has
turned out to be more time-consuming than I can afford.) Or,
alternatively, could I simply publish a patch?

The reason I am asking is that at least three other people have
expressed their interest in continuing to use psgml, two of them in
private emails to me. It seems a shame to me to let this (in my
opinion) great package go unused, and on the other hand, I do not want
to infringe on the rights of those who have written it.

Many thanks in advance,

Florian

[Dmitry Gutov]

flo...@fsavigny.de (Florian v. Savigny) writes:

> although this is a legal question, I am confident enough that most
> people know the answer. It is this:
>
> Some months ago, I fixed the "old-style backquotes" problem in psgml
> (which was fairly straightforward, even if I did not even know what
> backquotes are about), and I had inquired about this on this list
> beforehand. Back then, Stefan Monnier suggested that we should track
> down the other contributors to find out about copyright to see if the
> fix can be put on ELPA. We have tried this, but many of them do not
> seem to be available.
>
> Now, my question is this: Given that Lennart released it under the GPL
> (there is no separate LICENSE file, but all the .el files contain the
> standard sentences to that effect), do I actually need any previous
> contributors' permission to make the fix publicly available? (The
> problem is not that I do not want to ask them, but rather that it has
> turned out to be more time-consuming than I can afford.) Or,
> alternatively, could I simply publish a patch?

IANAL, but as long as the original code is licensed under GPL, you can
absolutely publish the patch, and the modified code.

You won't be able to distribute it via GNU ELPA or together with GNU
Emacs because the project's policy requires copyright assignment.

-- Dmitry

[Andreas Röhler]

Am 31.03.2013 13:49, schrieb Florian v. Savigny:
>
> Dear listmates,

>
> although this is a legal question, I am confident enough that most
> people know the answer. It is this:
>
> Some months ago, I fixed the "old-style backquotes" problem in psgml
> (which was fairly straightforward, even if I did not even know what
> backquotes are about), and I had inquired about this on this list
> beforehand. Back then, Stefan Monnier suggested that we should track
> down the other contributors to find out about copyright to see if the
> fix can be put on ELPA. We have tried this, but many of them do not
> seem to be available.
>
> Now, my question is this: Given that Lennart released it under the GPL
> (there is no separate LICENSE file, but all the .el files contain the
> standard sentences to that effect), do I actually need any previous
> contributors' permission to make the fix publicly available? (The
> problem is not that I do not want to ask them, but rather that it has
> turned out to be more time-consuming than I can afford.) Or,
> alternatively, could I simply publish a patch?
>

When using GPLed code, you may simply publish your changed code GPLed again.

Thanks giving another example wrt to the noxious results of copyright assigment policy, which undermines goals of GPL.

Andreas

[Bastien]

Hi,

Andreas R�hler <andreas...@easy-emacs.de> writes:

> When using GPLed code, you may simply publish your changed code
> GPLed again.

More precisely, GPLed code gives you the right to reuse and publish
the code in your code if it's published under the same license.

So yes, you can publish your changes to this code.

> Thanks giving another example wrt to the noxious results of
> copyright assigment policy, which undermines goals of GPL.

Why pushing your agenda against copyright assignment into this thread?

I don't think it is irrevelant.

The question is "Can I publish the changes?" and the answer is "Yes".

As for including psgml into GNU ELPA, the answer is "Not until all
authors have sign the FSF copyright assignment".

This is a policy that is particular to GNU Emacs and some other GNU
projects. It shows respect for potential authors (by not including
their code without their permission) and protects actual authors (by
allowing them to rely on the fact that changes against Emacs code by
signed contributors can be part of Emacs.)

--
Bastien

[Bastien]

Bastien <b...@altern.org> writes:

> I don't think it is irrevelant.

I meant "relevant" of course :)

--
Bastien

[Andreas Röhler]

Am 31.03.2013 17:09, schrieb Bastien:
> Hi,
>
> Andreas R�hler <andreas...@easy-emacs.de> writes:
>
>> When using GPLed code, you may simply publish your changed code
>> GPLed again.
>
> More precisely, GPLed code gives you the right to reuse and publish
> the code in your code if it's published under the same license.
>
> So yes, you can publish your changes to this code.
>
>> Thanks giving another example wrt to the noxious results of
>> copyright assigment policy, which undermines goals of GPL.
>
> Why pushing your agenda against copyright assignment into this thread?

Copyright assignment policy stifles cooperation, encourages bad manners, spreads FUD.
It's not me putting this at the agenda, it's there by it's inventors.

>
> I don't think it is irrevelant.
>

> The question is "Can I publish the changes?" and the answer is "Yes".
>
> As for including psgml into GNU ELPA, the answer is "Not until all
> authors have sign the FSF copyright assignment".
>
> This is a policy that is particular to GNU Emacs and some other GNU
> projects. It shows respect for potential authors (by not including
> their code without their permission)

So GPL doesn't show this respect?(!)
Requiring (unpaid!) copyright assignment means perverting the GPL.

and protects actual authors (by
> allowing them to rely on the fact that changes against Emacs code by
> signed contributors can be part of Emacs.)
>

Emacs - a marvelous tool misused for bad policy here.

[Andreas Röhler]

Am 31.03.2013 17:33, schrieb Bastien:

> Bastien <b...@altern.org> writes:
>
>> I don't think it is irrevelant.
>

> I meant "relevant" of course :)
>

Maybe read what the OP asked.

[Bastien]

Andreas R�hler <andreas...@easy-emacs.de> writes:

I re-read it. The OP asked if he can publish the changes.

--
Bastien

[Eli Zaretskii]

> Date: Sun, 31 Mar 2013 17:35:41 +0200
> From: Andreas Röhler <andreas...@easy-emacs.de>
> Cc: help-gn...@gnu.org

>
> >> Thanks giving another example wrt to the noxious results of
> >> copyright assigment policy, which undermines goals of GPL.
> >
> > Why pushing your agenda against copyright assignment into this thread?
>
> Copyright assignment policy stifles cooperation, encourages bad manners, spreads FUD.

But saying that in GNU forums is rude and off-topic. Please do this
somewhere else.

[Andreas Röhler]

Am 31.03.2013 18:06, schrieb Eli Zaretskii:
>> Date: Sun, 31 Mar 2013 17:35:41 +0200

>> From: Andreas R�hler <andreas...@easy-emacs.de>

>> Cc: help-gn...@gnu.org
>>
>>>> Thanks giving another example wrt to the noxious results of
>>>> copyright assigment policy, which undermines goals of GPL.
>>>
>>> Why pushing your agenda against copyright assignment into this thread?
>>
>> Copyright assignment policy stifles cooperation, encourages bad manners, spreads FUD.
>
> But saying that in GNU forums is rude and off-topic. Please do this
> somewhere else.
>
>
>

Hi Eli,

please consider the case at stake: a great library like psgml.el, which delivered solutions hardly
found elsewhere, was kept nearly unknown for more than a decade.

Don't you accept there is a difference between not bundling code and not mentioning it?
Not to say: discourage it's use.

If you might consider the effect already visible in this thread, you will remark the notion of the explicit consentement of authors for a change or republishing as so
called matter of respect.

Don't you see the danger which raises for all free software from there?
All legal tracts and stances are interpreted in the light of habit.
If a court notices the habit of requesting personal consentements for otherwise GPLed code changes as a need,
the story of free software might be over.

Best regards,

Andreas

[Eli Zaretskii]

> Date: Sun, 31 Mar 2013 18:57:25 +0200
> From: Andreas Röhler <andreas...@easy-emacs.de>
>

> please consider the case at stake: a great library like psgml.el, which delivered solutions hardly
> found elsewhere, was kept nearly unknown for more than a decade.
>
> Don't you accept there is a difference between not bundling code and not mentioning it?
> Not to say: discourage it's use.
>
> If you might consider the effect already visible in this thread, you will remark the notion of the explicit consentement of authors for a change or republishing as so
> called matter of respect.
>
> Don't you see the danger which raises for all free software from there?
> All legal tracts and stances are interpreted in the light of habit.
> If a court notices the habit of requesting personal consentements for otherwise GPLed code changes as a need,
> the story of free software might be over.

It doesn't matter what I think on this matter. I'm saying that
attacking the FSF on one of the technical GNU forums is rude and
unethical. There are specialized forums for this kind of discussions,
please take this subject there.

[Andreas Röhler]

Am 31.03.2013 20:57, schrieb Eli Zaretskii:
>> Date: Sun, 31 Mar 2013 18:57:25 +0200

>> From: Andreas R�hler <andreas...@easy-emacs.de>

>>
>> please consider the case at stake: a great library like psgml.el, which delivered solutions hardly
>> found elsewhere, was kept nearly unknown for more than a decade.
>>
>> Don't you accept there is a difference between not bundling code and not mentioning it?
>> Not to say: discourage it's use.
>>
>> If you might consider the effect already visible in this thread, you will remark the notion of the explicit consentement of authors for a change or republishing as so
>> called matter of respect.
>>
>> Don't you see the danger which raises for all free software from there?
>> All legal tracts and stances are interpreted in the light of habit.
>> If a court notices the habit of requesting personal consentements for otherwise GPLed code changes as a need,
>> the story of free software might be over.
>
> It doesn't matter what I think on this matter. I'm saying that
> attacking the FSF on one of the technical GNU forums is rude and
> unethical.

When pointing at some probably wrong decision, there is no attack at the FSF.
IMO FSF does a great job with its anti-DRM and many other campaigns.

As IPython got the FSF Free Software Award some days ago: Does the IPython-project require copyright assigments? Never heard from.
So that policy seems not so closely linked to FSF as you assume. It's just a decision effecting Emacs development and usage.

[Eli Zaretskii]

> Date: Sun, 31 Mar 2013 21:49:31 +0200
> From: Andreas Röhler <andreas...@easy-emacs.de>
>

> When pointing at some probably wrong decision, there is no attack at the FSF.
> IMO FSF does a great job with its anti-DRM and many other campaigns.
>
> As IPython got the FSF Free Software Award some days ago: Does the IPython-project require copyright assigments? Never heard from.
> So that policy seems not so closely linked to FSF as you assume. It's just a decision effecting Emacs development and usage.

Which part of "There are specialized forums for this kind of discussions,
please take this subject there" did you not understand?

[Florian v. Savigny]

Thanks to everybody for the clear answers!

Andreas, although I know I might be risking to lengthen this
needlessly, I would like to confirm what the others said: I did merely
ask whether I can publish it, and I am completely happy with the
answer.

Although, as may be guessed, I do understand, and even share, your
motive - keeping great software that was written for everybody
available to everybody -, I do not see how the ELPA copyright policy
hinders that in any way. Lots of elisp packages are not on ELPA, but
that does not preclude distributing and using them, or making them
known. The fact that psgml has not been available in a usable form for
some time now even though a fix has been in place for some time is
simply due to my not asking about how to publish it, not to the ELPA
copyright requirements. I don't think anything will keep me from
announcing this fix publicly enough.

FWIW, my gut feeling is that the Emacs people have a good reason for
their requirements. Emacs has always been a very conservative piece of
software, and to me, this is indeed part of what makes it so
exciting. There might be a downside to this, but there always is,
isn't there?

(Actually, I think the fundamental, well, problem with psgml is
actually that it does not seem to have a maintainer anymore. If it
had, the question would not have occurred at all, I presume.)

Best regards,

Florian

[Bastien]

Hi FLorian,

note that "ELPA" is sometimes confusing: "GNU ELPA" is the official
GNU Emacs Lisp Package Archive, which lives under the same copyright
conditions than code for GNU Emacs (i.e. authors need to sign the FSF
copyright assignement for their code to be accepted in GNU ELPA.)

But other "ELPA" repositories like Marmalade do not require this,
see for example: http://marmalade-repo.org

I guess many Emacs users expect useful code to live in at least one
of those repositories: maybe psgml could use one, it would boost its
discoverability?

Best,

--
Bastien

[Florian v. Savigny]

> But other "ELPA" repositories like Marmalade do not require this,
> see for example: http://marmalade-repo.org
>

> Maybe psgml could use one [of these], it would boost its
> discoverability?

Thanks very much, Bastien, I did not know this. I'll spread the word!
;-)

Kind regards,

Florian

[Nicolas Neuss]

Bastien <b...@altern.org> writes:

> Hi,

I think there is a more important reason for those copyright
assignments: The FSF considers its projects as so central for free
software that their copyright status has to be beyond any doubt.

For example, there could arise severe damage if some malicious company
(you name it) suddenly claimed that central pieces of GNU software were
contributed (maybe years ago) by some of their employees without
employer's permission.

Nicolas

[I will read answers to this, but (probably) not respond, because I also
think that this is the wrong newsgroup for this discussion.]

[Andreas Röhler]

In which way a CA-paper could avoid that?
What about the opposite danger: a company signs, provides crucial parts and withdraws CA?
See current discussions at emacs-devel on a lower scale.

Emacs got a lot of re-write already for pur CA-reasons.
Isn't that spoiled human labor in the light of GPL?

Andreas

[Dmitry Gutov]

Andreas Röhler <andreas...@easy-emacs.de> writes:
> In which way a CA-paper could avoid that?
> What about the opposite danger: a company signs, provides crucial parts and withdraws CA?
> See current discussions at emacs-devel on a lower scale.

You cannot withdraw the assignment for code that's already accepted.
That's the whole reason why that discussion on emacs-devel still hasn't
stopped.

[Andreas Röhler]

IMHO it hasn't stopped, because CA opens a completely new legal area above the GPL. Legal consequence from GPL
has been widely discussed, seem accepted and working.

CA, as we learned, is partly privately negotiated. While GPL is open, individual CA agreements are a not known.
To evaluate the legal state of different parts in Emacs wrt CA will need one or more lawyers.

Andreas

[Bastien]

Hi Andreas,

Andreas R�hler <andreas...@easy-emacs.de> writes:

> CA, as we learned, is partly privately negotiated.

Not to my knowledge.

The copyright assignment is signed by contributors. I'm not aware of
any negociations, and I'd be surprised to learn that there are private
negociations.

> To evaluate the legal state of different parts in Emacs wrt CA will
> need one or more lawyers.

This suggests that the copyright status of Emacs is uncertain.

I hope you understand it is quite a serious statement and requires
some evidence.

The copyright status for code in Org's code is very clear, and I think
this is also the case for the copyright status of the rest of Emacs.
I sometimes get impatient when I have to wait for the FSF Copyright
Clerk to handle new copyright assignments for a new contributor, but
I know he has a lot of work and he's doing a great job.

--
Bastien

[Phillip Lord]

Andreas Röhler <andreas...@easy-emacs.de> writes:

>> For example, there could arise severe damage if some malicious company
>> (you name it) suddenly claimed that central pieces of GNU software were
>> contributed (maybe years ago) by some of their employees without
>> employer's permission.
>>
>

> In which way a CA-paper could avoid that?
> What about the opposite danger: a company signs, provides crucial parts and withdraws CA?
> See current discussions at emacs-devel on a lower scale.

It doesn't. It's the employer disclaimer of rights that avoids this.

Of course, this doesn't avoid the problem entirely; I mean, the
disclaimer could be a fake. The employer could claim the person who
signed it didn't have the right to. The law could change entirely,
meaning that the entire world is now owned by Facebook, Google or
Microsoft.

It's about due diligence. The FSF is taking reasonable steps to ensure
that it can carry out it's mission. This all sounds sensible to me.
If unfortunate; I don't think anyone would argue that its not a pain in
the ass.

Phil

[Andreas Röhler]

Am 03.04.2013 10:29, schrieb Bastien:
> Hi Andreas,

>
> Andreas R�hler <andreas...@easy-emacs.de> writes:
>

>> CA, as we learned, is partly privately negotiated.
>
> Not to my knowledge.
>

http://lists.gnu.org/archive/html/emacs-devel/2013-04/msg00066.html

BTW experienced that proceeding already years ago, which is just polite and correct as such.

> The copyright assignment is signed by contributors. I'm not aware of
> any negociations, and I'd be surprised to learn that there are private
> negociations.
>
>> To evaluate the legal state of different parts in Emacs wrt CA will
>> need one or more lawyers.
>
> This suggests that the copyright status of Emacs is uncertain.

Law and legal stats differ in theory and reality.
In theory more laws define more cases and should provide more security.
In reality the realm of possible interpretation is expanded too.

>
> I hope you understand it is quite a serious statement and requires
> some evidence.
>
> The copyright status for code in Org's code is very clear, and I think
> this is also the case for the copyright status of the rest of Emacs.
> I sometimes get impatient when I have to wait for the FSF Copyright
> Clerk to handle new copyright assignments for a new contributor, but
> I know he has a lot of work and he's doing a great job.
>

To know the copyright status as such --as assignments may differ-- you must read assignment by assignment, evaluate and understand the diffs correctly -
which is absurd as IMHO the whole CA arts.

BTW German law --Urheberrechtsgesetz-- for example declares non-valid such copyright assignments - even if signed.
Will a US-court accept CA from a German, when german law declares CA possible?

Read Article 32
Equitable remuneration

http://www.gesetze-im-internet.de/englisch_urhg/englisch_urhg.html#p0157

I.e. GPL is not affected, as "anyone" matches. Will the FSF count as "anyone"?

Best,

Andreas

[B. T. Raven]

Die Wed Apr 03 2013 07:03:40 GMT-0500 (Central Daylight Time) Andreas
Röhler <andreas...@easy-emacs.de> scripsit:

> Am 03.04.2013 10:29, schrieb Bastien:
>> Hi Andreas,
>>

Your English is so good that I can't tell whether you are a native
speaker or not. Did you mean to write "if (wenn) German law declares CA
IMpossible"? I'm sure the FSF's lawyers know about the status of German
law, don't they? In any case, donative assignments are so legally thorny
that any discussion of them should probably be limited to private
communication with RMS or between/among the FSF's lawyers in camera.

>
> Read Article 32
> Equitable remuneration
>
> http://www.gesetze-im-internet.de/englisch_urhg/englisch_urhg.html#p0157
>
> I.e. GPL is not affected, as "anyone" matches. Will the FSF count as
> "anyone"?

It will certainly count under U.S. law now that Microsoft enjoys the use
of a virtual mind and soul (i.e. is even more of a person than any
corporation was under the personification doctrine established after the
Civil War). It can now have political opinions an can spend virtually
unlimited amounts of money to make those opinions heard.

Sorry, off topic but interesting.

Ed

>
> Best,
>
> Andreas
>
>
>

[Andreas Röhler]

Thanks! FWIW ;)

Did you mean to write "if (wenn) German law declares CA
> IMpossible"?

yep, sorry, a typo.

> I'm sure the FSF's lawyers know about the status of German
> law, don't they?

Don't know. It was just an example. What about japanese or chinese law when negotiating with it citizens?
Just wanted to point at the costs and risks when entering that procedure.

> In any case, donative assignments are so legally thorny
> that any discussion of them should probably be limited to private
> communication with RMS or between/among the FSF's lawyers in camera.
>

AS the GPL works, making a success story of free software, assignment-stubborness has a kind of tragic-comic.

Not being surprised that nastiness of legal stuff corresponds with a respective style of communication:

http://lists.gnu.org/archive/html/bug-gnu-emacs/2013-03/msg00495.html

Anyway, as always in error too :)

Andreas

[ ... ]