Chronological Thread 

< Chronological >      < Thread >

---

Thank you for all your answers and clarifications.

Regarding the model I wanted to publish, I have chosen CC-BY-3.0 for the sake of the license compatibility, although I would prefer the GNU GPLv3+ to ensure that the access and libre use of the source code is guaranteed [3], I was not completely sure if the re-licensing of modified version under GNU GPLv3 is possible.

After some more reading now it seams that it would actually be possible, as CC-BY-3.0 is one-way compatible with GPLv3 [1]. However, this would exclude the publication back in the CellML repository (if CC-BY is a prerequisite for that), as GPLv3 cannot be re-licensed as CC-BY-3.0 (so double licensing would be required anyways).

> As for why this [CC-BY-3.0] license was picked (instead of the specific GNU ones) was to allow maximum reach with attribution, but the exact reasoning was probably recorded in discussions many, many years ago.

Unfortunately, I couldn't find the reference to neither "license", "copyright" or "CC-BY" in any of the archives.

It is true that GPLv3 does not require attribution by default, but it has a provision to require the preservation of the attribution notices in the article 7b.

Also I wanted to point out, that the Creative Commons licenses are primarily used for data and even Creative Commons suggests to refrain from using them for software [2]. I understand the CellML models as kind of source code to generate the model functions, hence I think that a software license would be more appropriate.

I believe that every author should have the right to publish the code under the license of their choice. The repository mission should be to ensure that the uploaded models are easily usable by others, thus only publishing works under open source [4] or free software licenses [5].

Best regards,

Tomas


[1] https://protect-au.mimecast.com/s/9YlRCr81GMHEjVRNU7Bvks?domain=creativecommons.org

[2] https://protect-au.mimecast.com/s/lCAvCvl1M9SxgNnBcXkMQQ?domain=creativecommons.org

[3] https://protect-au.mimecast.com/s/NjqfCwV1NRfKQ8EOC908hM?domain=gnu.org

[4] https://protect-au.mimecast.com/s/Fg8HCxng07Ck0M5GiwjXAA?domain=gnu.org

[5]
https://protect-au.mimecast.com/s/BRO2Cyoj4JiZWXPjUN0B8_?domain=opensource.org




On 25/10/2023 09:41, David Nickerson wrote:

Thanks Matthias - differences in repository behaviours was also something that came out in that work with the FAIR indicators for models effort. PMR will respect the license if you encode it in the particular RDF that is expected or the user sets it during the exposure creation. Things get a bit blurry if you want to apply different licenses to different files in a single workspace, but it is possible for CellML models when building an exposure - see https://protect-au.mimecast.com/s/4xQoCzvk8xt6BgYPuwHBYY?domain=aucklandphysiomerepository.readthedocs.io. Of course, the exposures are like a release/publication, whereas the workspace (i.e., the underlying git repo) doesn't know anything about licenses applied to exposures.

Cheers,
David.

On Wed, 25 Oct 2023 at 19:37, Matthias König <konigmatt AT googlemail.com> wrote:

Hi all,

@David. I would not sure that those fallback would work because
the repository is probably overwriting the license of the model.
I.e. you agree to the repository license and no matter what you
write in the model does not matter.
As least that is how that works in biomodels. They force a CC0
license on every model uploaded even if I explicitly state in the
meta-data the model is CC-BY.
So you will need in addition the information of the
repository/source license which applies (especially if the license
is more open as in the case of biomodels).

Best Matthias


On Tue, Oct 24, 2023 at 10:21 PM David Nickerson

<david.nickerson AT gmail.com>
wrote:

Hi Tomas,

Just to add that this separation of the license from the
repository content is an issue that really came to the fore
recently in working through some FAIR indicators for modelling
(see
https://protect-au.mimecast.com/s/Bab2CANpL9H4o6KmI2WWGe?domain=fair-ca-indicators.github.io).
As Tommy points
out, the license information is available on the
website/repository (if you know where to look) but once you
download a model from the repository, that knowledge does not
persist.

We originally envisioned modellers capturing the license
information in the model metadata and had the catch-all
fallback for PMR "to be safe". In practice, that approach
hasn't worked out and the default fallback has become the
standard practice. We have at various times talked about
harmonizing licensing information across the COMBINE standards
as part of the OMEX Metadata
specification
(https://protect-au.mimecast.com/s/xE_QCBNqMRHjm6XZurwTR7?domain=identifiers.org),
but without actually including it in the spec yet. Again,
something that has been pushed as we look to evaluate
FAIR-ness of models and models-in-repositories.

Cheers,
Andre.

On Wed, 25 Oct 2023 at 00:30, Tommy Yu

<tommy.yu AT auckland.ac.nz>
wrote:

Hi Tomas,

Generally speaking, the models hosted on the Physiome
Model Repository are licensed under the Creative Commons
Attribution 3.0 Unported License

<https://protect-au.mimecast.com/s/hAlYCD1vOJHqmvL6i8ddEw?domain=creativecommons.org>.
For the
particular workspace that you've linked, it has an
exposure (navigable via the `Latest Exposure` button in
that page) and it leads to the top level exposure
container, and from there there is a link to the exposure
view of the workspace, where the license for the work is
visible.  As for why this license was picked (instead of
the specific GNU ones) was to allow maximum reach with
attribution, but the exact reasoning was probably recorded
in discussions many, many years ago.

The workspace view being shown is a general file listing
and are managed by collaborator(s) that may be working on
the model, though it is up to them to add whatever files
they deem necessary (including the license file) to this
workspace that underlies the exposure presentation layer
for the repository itself.

If you'd like to publish a modified version of the model
to the repository, please create an account, create a fork
of that workspace, and push your changes to that fork to
get the changes into your fork as a start using Git.  The
full details of this whole process may be found in the
documentation, which is linked via the top tool bar.

Regards,
Tommy.

On 24/10/23 23:40, Tomas Stary wrote:
> Hello,
>
> I would like to publish a modified version of a model
from the CellML
> repository [1].
>
> I wanted to make sure, that I have the necessary
permission to do that
> and find out what are the conditions. However, the
licensing information
> of the published files is missing, which could mean that
all the
> copyrights may be reserved by the copyright holders [2],[3].
>
> I wonder if the general licensing information is
specified in any other
> place I might have failed to find?
>
> I have also noticed that this problem is not specific
for the model [1],
> but might be relevant to other CellML repositories as
well. Would it be
> possible for you to include the licensing information
into the
> repositories, perhaps in the form of LICENSE file with
the terms?
>
> I believe that some kind of copyleft license [4] would
work best to keep
> the project as open as possible.
>
> Best regards,
>
> Tomas
>
> [1]
https://protect-au.mimecast.com/s/LH2sCE8wg6HLG98Au4y088?domain=models.physiomeproject.org
>
> [2]
https://protect-au.mimecast.com/s/gNU7CGv0kQtklYNyiXyZ0J?domain=gnu.org
>
> [3]
https://protect-au.mimecast.com/s/IViiCJyBn7IZzrW4UmqeKj?domain=choosealicense.com
>
> [4]
https://protect-au.mimecast.com/s/fAJaCK1Do5HOwEyAHyoJrM?domain=gnu.org
>
> --
> Karlsruhe Institute of Technology (KIT)
> Institute of Biomedical Engineering
>
> Tomas Stary, PhD
> Research Software Engineer
>
> Fritz-Haber-Weg 1 (Building 30.33, Room 507)
> 76131 Karlsruhe, Germany
>
> Tel: +49 721 608-41578
>
> E-Mail:
tomas.stary AT kit.edu
> Web:
https://protect-au.mimecast.com/s/LO4mCL7Ep5FKOjDoCGaBmL?domain=ibt.kit.edu
<https://protect-au.mimecast.com/s/LO4mCL7Ep5FKOjDoCGaBmL?domain=ibt.kit.edu>

--
Karlsruhe Institute of Technology (KIT)
Institute of Biomedical Engineering

Tomas Stary, PhD
Research Software Engineer

Fritz-Haber-Weg 1 (Building 30.33, Room 507)
76131 Karlsruhe, Germany

Tel: +49 721 608-41578

E-Mail:tomas.stary AT kit.edu
Web:https://protect-au.mimecast.com/s/LO4mCL7Ep5FKOjDoCGaBmL?domain=ibt.kit.edu

---

• Re: [cellml-discussion] copyrights and licensing issues, Tomas Stary, 11/01/2023

---

Archive powered by MHonArc 2.6.18.