[License-review] ESA-PL Weak 2.3

Patrice-Emmanuel Schmitz pe.schmitz at googlemail.com
Mon May 7 15:35:32 UTC 2018

To Carsten: I do not challenge the willingness of ESA to produce its own
F/OSS license. This may add to license proliferation and be motivated by
some institutional ego but it may also address some needs and after all,
the NASA did so…

What is more disturbing and in my opinion should not be approved by OSI is
the lack of clarity regarding compatibility: this draft ESA license is
compatible with the GPL and with CeCILL, but not with the AGPL and the
EUPL. However the ESA license addresses SaaS (like the AGPL and the EUPL),
and CeCILL is compatible with both the AGPL and the EUPL. Therefore the
information regarding compatibility looks unclear/incomplete, since there
are serious doubts that all ESA stakeholders know the provisions of CeCILL,
which is mainly a French national license.

2018-05-06 18:50 GMT+02:00 Patrice-Emmanuel Schmitz <
pe.schmitz at googlemail.com>:

> On 06.05.2018, 15:48 Carsten Gerlach wrote:
> > since CeCILLv2.1 permits the license switch to
> > the EuPL, it was felt that no explicit permission was necessary.
> This is quite surprising and should probably be reconsidered. Indeed, one
> can assume that the provisions of CeCILL v2.1 are (or should be?)
> reasonably well known in France, where the licence is produced by three
> agencies (CEA, CNRS & INRIA ) and used. But for sure, the provisions and
> even the existence of CeCILL v2.1 are most probably completely unknown in
> all other countries that are member of ESA.
> Therefore the omission of the EUPL which is promoted by European
> institutions and which is the sole licence with a working value in 23
> language is astonishing. Doing this because "CeCILL v2.1 permits the
> license switch to the EUPL, so it was felt that no explicit permission was
> necessary" seems a very poor argument!
> Another critical point is the use of the term "license switch" (The FSF
> uses the term "relicensing" which has probably the same meaning). The use
> of such term is dangerous and no lawyer should use it in so far it provides
> the impression that a recipient can change the license of an existing
> project or software, or switch from license A to a license B.
> What is authorised by the EUPL (article 5, compatibility clause) is to
> reuse or merge all or parts of the covered code in ANOTHER project that
> needs to be distributed under a compatible copyleft license, because parts
> of this other work where covered by this compatible copyleft licences (i.e.
> the GPLv2 or v3). But this is not a "license switch" or a "relicensing" of
> the original code that - outside this other combined derivative - stays
> covered by the EUPL.
> Similarly, the provisions of CeCILL (article 5.3.4) that "The licensee can
> include the software in a code that is subject to the provisions of one of
> the versions of the GNU GPL, and/or GNU Affero GPL and/or EUPL, and
> distribute that entire code under the terms of the same version of the GNU
> GPL, and/or GNU Affero GPL and/or EUPL." should receive the same
> interpretation: you may reuse the code in another project, but the original
> work stays licensed under CeCILL.
> Therefore my recommendation is to ban dangerous terms like "relicensing"
> or "license switch" and recommend "reuse or merge in another software
> project".
> P-E Schmitz, lawyer, Joinup advisor.
> 2018-05-06 15:48 GMT+02:00 Carsten Gerlach <cgerlach at tcilaw.de>:
>> On 05.03.2018 12:22, Patrice-Emmanuel Schmitz wrote:
>> > Another question to Carsten Gerlach:
>> >
>> > The “weak” ("reasonable" or "moderated" sounds better) copyleft
>> > ESA-PL_Weak provides the right to (re-)distribute the Software and/or
>> > Modifications under a compatible licence.
>> > Annex A lists only two licences: the GNU GPL v2 and later and the CeCILL
>> > v2 or later.
>> >
>> > Questions:
>> > 1. Why not the GNU/AGPL, knowing that the ESA-PL_Strong covers SaaS
>> > (point 3.4 which differs from tha ESA-PL_Weak)?
>> The AGPL could be added to the list of compatible licenses, but was not
>> considered by the ESA licensing board due to its limited practical
>> relevance in the space industry. Generally, the list of compatible
>> licenses was to be kept as short as possible and based on the prior
>> experience of the agency.
>> > 2. Why not the EUPL v1.2, which should be considered since ESA is
>> > "quasi" a specific European Agency (even if some members differ),
>> > knowing that CeCILL v2.1 is compatible with the EUPL (and the
>> reciprocate)?
>> > NB: in addition, the EUPL V1.2 authorises specific agreements concerning
>> > applicable law and arbitration, where CeCILL refers to the French law
>> > only and to the court of Paris
>> See above. Additionally, since CeCILLv2.1 permits the license switch to
>> the EuPL, it was felt that no explicit permission was necessary.
>> Kind regards, Carsten
> --
> Patrice-Emmanuel Schmitz
> pe.schmitz at googlemail.com
> tel. + 32 478 50 40 65

Patrice-Emmanuel Schmitz
pe.schmitz at googlemail.com
tel. + 32 478 50 40 65
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20180507/05245578/attachment.html>

More information about the License-review mailing list