[License-review] Approval request for GG License 1.0

[Engel Nyst]

Hello Valentino,

In this message I'll pass through the text and your answers. Not
everything below affects approval, there are also other questions or
comments.

On 12/30/2014 01:39 PM, Valentino Giudice wrote:
>
>
> c) "Derivative work": Any work obtained by applying "transformation
> processes", either manual or automatic, to this work. Those
> transformation processes includes, but are not limited to, editing,
> adaptations, obfuscations, compiling, joining or aggregation with
> other works, inclusion in other works and removing components that
> are parts of this work. Any work obtained by applying transformation
> processes to a derivative work shall be considered a derivative work
> itself


Derivative work is a term that the law defines, and typically not a
license. So I'd suggest to use another term here, for this concept
of covered material you're trying to call derivative work.

You can also search the web for the ongoing discussions on rewriting
some copyright related laws in European Union and add your suggestions
there. For example, do you think the law should define it for software
in software terms? Or do you think there is something you're trying to
express which is the job of a license?

> joining or aggregation with other works [...] removing components
> that are parts of this work


Removing components can result in no copyrightable component left over.
Aggregation with other works can refer to different works on a CD.

None of these two are derivative works of the original, and the attempt
to extend your license to them doesn't comply with OSD 9.

> a) Attribution (or "Citation")


Why do you call attribution "citation"? I didn't see anything in the
following definition that corresponds to the common meaning of citation.

> If you distribute this work or any derivative work you must provide,
> with the distributed work, any supplied piece of information among
> the following ones: the author's name (or pseudonym, if applicable),
> copyright notice of the author on this work, the title of this work
> and, if you are distributing a derived work, information about how
> to obtain this work. You must also make clear this work is released
> under this license, providing a full copy of this text or via a
> hyperlink (or URI).


All these informations are informations you can write in a notice and
then the license can require that notice to be preserved.

For example there's no need to say in the license "title of this work",
when you can just add title of the work in your notice. Do you disagree,
do you try to achieve more than that, with this paragraph?

> if the distributed work is available in the binary form, attribution
> in the source code is not enough


Check out Apache License 2.0 (AL2.0) [1]. Do you think your license
requires more than AL2.0 does, on this point?

> b) Removing (or changing) the attribution: If required by the author
> of this work, you must remove the information the section 3.a
> ("Attribution") require you to provide. Note that this removal is not
> allowed if not expressly required or approved by the author of this
> work. Also, if required by the author, you must replace, in the
> attribution, the name of the author with a pseudonym or vice versa.


I don't remember seeing this requirement in a software license. It's
possible to have it, though, without running afoul of the OSD.

> c) Stating changes: If you distribute a derivative work, it must be
> clear (and, so, expressly specified) that it is a derivative work and
> not the original work.


See Apache License 2.0. Same question, do you think it does less than
what you're trying to achieve by this?

> 5) Exceptions: This license is just an agreement between you and the
> author of this work. By means of this license, the author allows you
> to use this work under the terms of this license. See the section 2
> ("Grants"). This work could be provided also under other licenses,
> more or less restrictive. The author of this work could (upon request
> or not) allow a use of this work contrary to this license, especially
> to the section 3 ("Obligations and restrictions"). Such a permission
> could be provided only to a few individuals and shall be considered
> an exception.


This seems a no-op, for a license based on copyright. CC licenses
contain similar paragraphs, but I don't see what effect they really
have, other than educative. What do you think they add?

> Also, this license states the following exceptions: a) The section
> 1.c ("derivative work") establishes that the result of a compilation
> process shall be considered a derivative work and the section 3.c
> establishes that when you distribute a derivative work, you must
> specify that it is a derivative work and not this work. This is not
> required if the derivative work has been made only by compiling this
> work (and without any other transformation process). See the section
> 1.b ("This work").


Why is this an "exception"? It doesn't seem to say anything else than
the license already said.

> b) Any removal of parts of the attribution (or of the full
> attribution) required by the section 3.b ("Removing or changing the
> attribution") shall not be considered a violation of the section 3.a
> ("Attribution").


Same, this repeats what the license already said. If you think it didn't
say it clearly enough before, then you can fix that.

> It is inspired to the Creative Commons Attribution 3.0 Unported,
> which, unfortunately, is not available for software (quote from the
> Creative Commons website: "We recommend against using Creative
> Commons licenses for software"). The GG1 license is intended to bring
> the principles of the CC-BY 3.0 license in the world of software.


Well, I used to think CC's advice on this would better be followed, but
I'm not so sure anymore. Why do you think you can't use CC-BY 3.0, what
points exactly don't fit your wishes? If there is something, then again,
why doesn't AL2.0 do it?

> The most similar license is, of course, the Creative Commons
> Attribution license itself (version 3.0), but it is not an OSI
> license.


Your license differs in several points from CC-BY 3.0, as noted, apart
from the wording, and those points may make your license not compliant.

> The most similar OSI-approved license is probably the MIT license.
> The GG license is different because it requires a different type of
> attribution (and it provides more information about it). Also, its
> clauses are more precise.


See above about AL2.0.

> The GNU Lesser General Public License shall also be compared to GG1.
> They are quite similar and LGPL is not a copyleft license. But the
> GG license requires a stronger type of attribution.



You're saying LGPL is not a copyleft license. Is this a typo, or how do
you define copyleft to conclude that LGPL is not a copyleft license?

> Unfortunately, since I am not a lawyer, no legal analysis is
> available


I think not making a legal analysis is not a problem in this case.
(Others' opinion may, and will, vary.)

However, there are related problems, such as the attempt to define
derivative work.

> *Earlier public discussions* Unluckily, there is no public discussion
> about this license (as far as I know).


This is a problem. Do you have software or intend to write, to place
under this license?

The point of this question is that when you write a license alone or
mainly you, it's very likely the license will be designed to serve only
you and not a community.

You could propose it to your friends, colleagues, or the internet. You
can then fix it, before submitting another version for approval, or you
can find out whether there is any interest (maybe there is, maybe there
isn't any), or you might find out that there are already licenses who
fulfill these wishes fairly well.

> I do not know whether I have been able to present this license well
> enough or not


What I didn't quite understand, is what exactly is important to you and
other (OSI approved) licenses don't do.

Except removing attribution per request, but why is that important for
software? You already have the requirement to mark a derivative as
distinct from the original.


[1] http://opensource.org/licenses/Apache-2.0


-- 
~ "We like to think of our forums as a Free-Speech Zone. And freedom
works best at the point of a bayonet." (Amazon, Inc.)