On Mon, 25 Nov 2024, Simon Josefsson wrote:
The DCO v1.1 published on https://developercertificate.org/ says:
Copyright (C) 2004, 2006 The Linux Foundation and its contributors.
Everyone is permitted to copy and distribute verbatim copies of this
license document, but changing it is not allowed.
The license appears non-free to me, does anyone disagree?
The text is widely included in Debian main packages.
Is there any policy exception to allow this?
The DCO is not a license text, which seems to be the only exception I
know of, so having the text appear under a non-modifiable license
doesn't seem appropriate for main.
Does anyone know a way to talk to the Linux Foundation and ask them to release the text into the public domain? Or at least release it under
some license that allows modifications.
On Mon, 25 Nov 2024, Simon Josefsson wrote:
The DCO v1.1 published on https://developercertificate.org/ says:
Copyright (C) 2004, 2006 The Linux Foundation and its
contributors.
Everyone is permitted to copy and distribute verbatim copies of
this
license document, but changing it is not allowed.
The license appears non-free to me, does anyone disagree?
The text is widely included in Debian main packages.
Is there any policy exception to allow this?
The DCO is not a license text, which seems to be the only exception
I
know of, so having the text appear under a non-modifiable license
doesn't seem appropriate for main.
Does anyone know a way to talk to the Linux Foundation and ask them
to
release the text into the public domain? Or at least release it
under
some license that allows modifications.
It was released under CC-BY-SA-2.5 in 2005:
© 2005 Open Source Development Labs, Inc.
The Developer's Certificate of Origin 1.1 is licensed under a
Creative Commons Attribution-ShareAlike 2.5 License.
If you modify you must use a name or title distinguishable from "Developer's Certificate of Origin" or "DCO" or any confusingly
similar name.
https://web.archive.org/web/20060524185355/http://www.osdlab.org/newsroom/press_releases/2004/2004_05_24_dco.html
Hi.
The DCO v1.1 published on https://developercertificate.org/ says:
Copyright (C) 2004, 2006 The Linux Foundation and its contributors.
Everyone is permitted to copy and distribute verbatim copies of this
license document, but changing it is not allowed.
The license appears non-free to me, does anyone disagree?
"Simon" == Simon Josefsson <[email protected]> writes:
On Monday, November 25, 2024 12:35:17 AM MST Simon Josefsson wrote:
Hi.
The DCO v1.1 published on https://developercertificate.org/ says:
Copyright (C) 2004, 2006 The Linux Foundation and its contributors.
Everyone is permitted to copy and distribute verbatim copies of this
license document, but changing it is not allowed.
The license appears non-free to me, does anyone disagree?
I am curious as to what aspect of this license you consider non-free? I am not concerned about "Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.” Those exact
words appear at the top of the GPL, which is considered a free license.
https://www.gnu.org/licenses/gpl-3.0.html
"Simon" == Simon Josefsson <[email protected]> writes:
Simon> Hi. The DCO v1.1 published on
Simon> https://developercertificate.org/ says:
Simon> Copyright (C) 2004, 2006 The Linux Foundation and its
Simon> contributors. Everyone is permitted to copy and distribute
Simon> verbatim copies of this license document, but changing it is
Simon> not allowed.
Simon> The license appears non-free to me, does anyone disagree?
Seems fine under DFSG 4 (author's integrity).
I can say something like
I assert DCO with the following modifications.
"Simon" == Simon Josefsson <[email protected]> writes:
"Simon" == Simon Josefsson <[email protected]> writes:
The DCO is not technically a license, but it is a legal document that
usually comes along with a license from the contributor to the project
(that license usually being the project license). It also serves to replace
a Contributor License Agreement, offering only a base level of assurances that the contributor has permission to make the contributions. Without a
DCO, a contribution is legally suspect; it carries no inherent guarantee
that the contributor wrote the code or owns the code, for all you know the contribution was just copied off stack overflow, or from the contributor's company's private code base.
In practice, although it is not a license, the DCO should be accepted as is for the same reasons license text is accepted; free software does not
depend on the freedom to modify the DCO, and is in fact better served by a non-modifiable DCO. There is still no real reason why the DCO itself needs
to be licensed under free terms, and plenty of reasons why it shouldn't.
This is a non-issue.
"Simon" == Simon Josefsson <[email protected]> writes:
Simon> Interesting -- am I understanding you correctly that you
Simon> would like to treat the DCO as a license text?
No.
Simon> And that it
Simon> is license that applies to the work in Debian?
No.
Simon> As far as I understand, DCO's are about granting rights on
Simon> contributions. Not granting rights to users, which is what
Simon> the DFSG is about. So I'm not sure I follow why the DFSG is
Simon> relevant for the DCO text at all. The DCO appears to me like
Simon> any other text file in a source package.
I agree, the DCO is another file in a source package.
Debian does not require that people be granted the right to modify files
in Debian source packages.
DFSG 4: The license may restrict source-code from being distributed in modified form only if the license allows the distribution of "patch
files" with the source code for the purpose of modifying the program at
build time. The license must explicitly permit distribution of software
built from modified source code. The license may require derived works
to carry a different name or version number from the original software.
So, I think it would be problematic if the DCO were included in a built binary unless there were mechanisms to change the text of the DCO in a
built binary (and doing so were legal).
However, the DCO is generally run by humans rather than computers.
A project could include contribution guidelines including the text of
DCO 1.1 and explaining the project specific modifications.
The humans evaluating whether contributions could be accepted could
evaluate as appropriate.
Yes, I am looking for a way to make the DCO fit the DFSG. Yes, in other circumstances I might look for a way to say no rather for a way to say
yes.
People have explained why we want to find a yes answer in this
circumstance and why what the DCO is doing is reasonable.
I believe DFSG 4 gives us that yes if we choose to use it.
The DCO is very intentionally *not* a license of any kind. A Contributor License Agreement is a license. The DCO was invented because many contributors were uncomfortable with contributor license agreements for various reasons. All the DCO does is confirm that you have the right to
make the contribution, it does not grant any permission from anybody to anybody.
Interesting. How do you combine that view with the interpretation that
went into rejecting GFDL as a free license? I'm having trouble
following your line of thinking after reading the following, especially regarding 2.1 Invariant Sections:
https://www.debian.org/vote/2006/vote_001
To be specific, how is this situation any different from a GFDL document
with an Invariant Section including the text of the DCO or, say, the GPL license, or the GNU Manifesto?
/Simon
On Tuesday, November 26, 2024 12:44:18 PM MST Simon Josefsson wrote:
Interesting. How do you combine that view with the interpretation that
went into rejecting GFDL as a free license? I'm having trouble
following your line of thinking after reading the following, especially
regarding 2.1 Invariant Sections:
https://www.debian.org/vote/2006/vote_001
To be specific, how is this situation any different from a GFDL document
with an Invariant Section including the text of the DCO or, say, the GPL
license, or the GNU Manifesto?
/Simon
The GFDL with Invariant Section is not the right comparison. The correct comparison is to the GPL (which has the exact same wording).
The difference is that nothing that is delivered to the user in the form of code or documentation that they cannot modify according to the GPL and the DCO. With invariant sections of the GFDL, documentation is delivered to the end user that they cannot modify. The distinct is important. The very foundation of most DFSG-free licenses is that copyright and license information must be preserved in all copies (in other words, copyright and licensing information are invariant, often specifically required to be verbatim). Everything else can be changed by the users. In very few cases in
Debian are you free to remove copyright or licensing information from the top
of a file and redistribute it.
If you feel strongly that the DCO is not DFSG-compliant, I would encourage you
to make your best argument to the greater Debian community (it would need to be an argument that has not to this point been made). I would recommend that
instead of focusing on the DCO, you should focus on why you feel the GPL (and,
honestly, the Apache License 2.0, all the BSD variants, Expat, and most other
licenses distributed in Debian that require that the unchanged text of the license be included in all copies and derivative works) is not compatible with
the DFSG as more people are familiar with the GPL than the DCO. Any conclusion reached about the GPL will automatically apply to the DCO.
I disagree. I don't see any problem with the license on the GPL text
itself, when GPL is used as a license on a piece of work in Debian and documented in debian/copyright.
The rest of your argument assume that I
do. The DCO is not used in that way, and nobody has suggested they
ought to be treated the same way so far. So the situations aren't comparable.
On Tuesday, November 26, 2024 1:31:51 PM MST Simon Josefsson wrote:
I disagree. I don't see any problem with the license on the GPL text
itself, when GPL is used as a license on a piece of work in Debian and
documented in debian/copyright.
Good.
The rest of your argument assume that I
do. The DCO is not used in that way, and nobody has suggested they
ought to be treated the same way so far. So the situations aren't
comparable.
The DCO is part of the chain of custody of the license, similar to the how the
copyright is part of the chain of custody of the license. It documents that the person providing the license has the right to provide you with the license, similar to how a copyright statement documents that the person providing the license has the right to provide you with the license.
I don't follow. Do you believe the DCO is a license text?
And that the DCO is part of the license that applies to the work in Debian?
On Wed, 27 Nov 2024, Soren Stoutner wrote:
On Wednesday, November 27, 2024 2:48:51 AM MST Simon Josefsson wrote:
I don't follow. Do you believe the DCO is a license text?
Yes. There is no question that the DCO is a license. It says right
in the text it is a license.
On Wednesday, November 27, 2024 2:48:51 AM MST Simon Josefsson wrote:On Wed, 27 Nov 2024, Soren Stoutner wrote:
I don't follow. Do you believe the DCO is a license text?
Yes. There is no question that the DCO is a license. It says right
in the text it is a license.
So if someone wants to include any non-free file in a package, they can
just claim in its header that it is a license, and it will be excepted
from the DFSG? That seems like a strange concept.
On Wednesday, November 27, 2024 4:48:30 AM MST Ulrich Müller wrote:
On Wednesday, November 27, 2024 2:48:51 AM MST Simon Josefsson wrote:On Wed, 27 Nov 2024, Soren Stoutner wrote:
I don't follow. Do you believe the DCO is a license text?
Yes. There is no question that the DCO is a license. It says right
in the text it is a license.
So if someone wants to include any non-free file in a package, they can
just claim in its header that it is a license, and it will be excepted
from the DFSG? That seems like a strange concept.
I can’t imagine anyone making that claim, and no one has in this thread.
However, if a project includes their Contributor License Agreement in their source code (the DSO in this case), doing so does not pose any DFSG problems (and, indeed, is a good idea, because it strengthens the chain of custody of the open-source license presented by the project to the end users).
Soren Stoutner <[email protected]> writes:their
On Wednesday, November 27, 2024 4:48:30 AM MST Ulrich Müller wrote:
On Wednesday, November 27, 2024 2:48:51 AM MST Simon Josefsson wrote:On Wed, 27 Nov 2024, Soren Stoutner wrote:
I don't follow. Do you believe the DCO is a license text?
Yes. There is no question that the DCO is a license. It says right
in the text it is a license.
So if someone wants to include any non-free file in a package, they can
just claim in its header that it is a license, and it will be excepted
from the DFSG? That seems like a strange concept.
I can’t imagine anyone making that claim, and no one has in this thread.
However, if a project includes their Contributor License Agreement in
problemssource code (the DSO in this case), doing so does not pose any DFSG
of(and, indeed, is a good idea, because it strengthens the chain of custody
the open-source license presented by the project to the end users).
I don't think so -- the fact that a DCO is included does not guarantee
that contributions were made under that DCO. It is just an auxilliary
text file with a non-free license, referred to by a README. It may be
that some contributions were given under the DCO, but there is no
guarantee, and this doesn't modify anything about the rights I'm given
as a user from the outgoing license.
Generally, I can't align your thinking with the decisions about the FDL:
https://www.debian.org/vote/2006/vote_001
I made the example earlier of a manual that has an FDL Invariant Section
that includes a copy of the GPL. I believe Debian consider that as
non-free, since a manual shouldn't contain things that cannot be
modified, but if I'm understanding you correctly you argue that would be acceptable.
The GFDL with Invariant Section is not the right comparison. The correct comparison is to the GPL (which has the exact same wording).
* Soren Stoutner:
The GFDL with Invariant Section is not the right comparison. The correct comparison is to the GPL (which has the exact same wording).
The FSF gives permission to make modified versions of the GPL,
though: <https://www.gnu.org/licenses/gpl-faq.html#ModifyGPL>
I don't think the unsigned DCO really supports the chain of custody/title
all that much, but it may still put minds at ease that the project is following best practices regarding IP.
| Sysop: | Keyop |
|---|---|
| Location: | Huddersfield, West Yorkshire, UK |
| Users: | 715 |
| Nodes: | 16 (2 / 14) |
| Uptime: | 04:06:22 |
| Calls: | 12,099 |
| Calls today: | 7 |
| Files: | 15,003 |
| Messages: | 6,517,888 |