CDLA permissive and sharing licenses


#1

https://cdla.io/sharing-1-0/

https://cdla.io/permissive-1-0/

Haven’t looked closely. Don’t see any info about relation to existing open data licenses.


#2

A little more background https://schd.ws/hosted_files/osseu17/31/CDLA%20OSS%20EU%20v1.pdf#page=8


#3

Hi @mlinksva

I just read the text of the permissive licence (HT @pwalsh) , as I understand it should be compatible with CC and ODC attribution licences (even though they are not explicitely named)

See:

“If You Publish Data You Receive or Enhanced Data:
(a) You may do so under a license of Your choice provided that You give anyone who Receives the Data from You the text of this Agreement, the name of this Agreement and/or a hyperlink or other method reasonably likely to provide a copy of the text of this Agreement;”

If I understand correctly the background paper you sent, Linux Foundation argues the CDLA is their attempt to address licence proliferation and incompatibility across licences (in their words “difficulties to combine data”)? If this is correct, then I wonder whether this clause does not do the exact opposite. See:

“You may provide additional or different license terms and conditions for use, reproduction, or distribution of that Enhanced Data, or for any combination of Data and Enhanced Data as a whole, provided that Your Use and Publication of that combined Data otherwise complies with the conditions stated in this License.“_

But maybe Linux Foundation did not address licence incompatibility but wanted to avoid problems of reuse, whenever people could see a case of shared authorship of data derivatives? See:

“3.3 You and each Data Provider agree that Enhanced Data shall not be considered a work of joint authorship by virtue of its relationship to Data licensed under this Agreement and shall not require either any obligation of accounting to or the consent of any Data Provider.”

What are other people’s thoughts on this?


#4

Hi,

I reached out to the Linux Foundation yesterday to try and arrange a call
to find out a little more. I know that people are Creative Commons have
been looking at the licences now that they’ve published.

I’m keen for the Foundation to:

  • submit the licences for review through our process here
  • state something about compatibility with CC licences, ideally by working
    with them directly on that to ensure that cross-licensing can happen

I did wonder whether Section 4 in the CDLA licences might be more
problematic. CC don’t ask for these types of assurances around personal
data – as you’re bound by local data protection and other laws regardless.
But their inclusion here might make compatibility problematic? IANAL so
interested in feedback from others.

I would also be interested in seeing what others think of this context
document, which distinguishes between inbound and outbound licences.

https://cdla.io/context/

The wording around separate outbound licences suggests that data
contributed under an open “inbound” licence might be otherwise restricted.
Which reads more like a data sharing / pooling arrangement.

At the ODI we’re going to continue to recommend CC-BY 4.0 and CC-BY-SA 4.0
as these already have traction, are recognised internationally, and offer
the same options.

Cheers,

L.


#5

Skeptical discussion at https://lwn.net/Articles/737212/


#6

BTW did I ever mention that we (OSMF-LWG) concluded our assessment on compatibility of CC-BY 4.0 as an inbound licence for data for OpenStreetMap https://blog.openstreetmap.org/2017/03/17/use-of-cc-by-data/ ?

Definitely we would prefer that ODI would not recommend CC-BY 4.0 as a data licence. And, my personal view, I doubt that many of the licensees actually understand the consequences of 2.a 5 (B), even though CC prominently displays the additional requirements over attribution on the human-readable summary page https://creativecommons.org/licenses/by/4.0/

In day to day business we find lots of uses in which it is assumed that CC-BY actually only means that attribution of the original source is required, which is see above, not the case.

Simon


#7

Rings a bell https://xkcd.com/927/
Good we have a new option on one hand (even when simplicity is not going to be their biggest strength I guess), specially coming from such relevant organisation. A pity that license’s creators are apparently working on silos and not caring about possible compatibility of their new licenses (sometimes, even on purpose) on the other.
Agree with Leigh and would love to see this one coming into the open definition approval and some more discussion on compatibility issues also happening, ideally with the participation from the respective the organisations. Nevertheless, remember the open definition license approval process does not require the license creator/steward to make the submission http://opendefinition.org/licenses/process/ so maybe we should just move the this thread there and make that conversation start more formally.


#8

There’s a mention of 50+ organisations being involved in the drafting: https://lwn.net/Articles/737534/ . I suppose they are the first expected users (possibly software companies, members of the Linux Foundation?). So, do we have examples of datasets released under this license yet?


#9

Hi,

Its under consideration here:

https://whosonfirst.org/blog/2018/08/21/license-change/

And used for some of Microsoft’s open data releases here

https://msropendata.com/datasets?sort=2&page=1&license=51259f77-12fd-47b4-8efe-a5b3344767ae

These are the first I’ve seen.

Moving from CC-BY for whose one first data seems like a mistake to me.

Cheers,

L.


#10

These are the first I’ve seen.

Thanks. So, only Microsoft-sponsored things so far?

Several of those examples include geocoded data. “Fair” is mentioned
as a reason, but I don’t see a contribution/community model.

This dataset is mostly a table of excerpts from the English Wikipedia:
https://msropendata.com/datasets/21032bb1-88bd-4656-9570-3172ae1757f0

Leaving aside that it’s probably a violation of CC-BY-SA 3.0 (that
could be fixed), it’s not clear to me how the CDLA would work if e.g.
all those IDs were imported into Wikipedia to generate links.

  1. Is an article which contains an extract from the dataset considered
    “enhanced data”? And a set of articles? And an XML dump on
    https://dumps.wikimedia.org/ ?
  2. If not, are they a violation of the sui generis right given the
    limits in §2.2 (I don’t see sublicensing) and §3.2 (different license
    would not be allowed)?
  3. If they are, how does one comply with §3.1(a)? Should the XML dumps
    be modified to carry such a statement?
  4. If an article is considered enhanced data, and I as an author bring
    legal action against Microsoft e.g. for GDPR purposes, in what cases
    would §5.2 mean that it would be a violation for me to
    use/edit/redistribute an article or Wikipedia dump including such
    Microsoft data?

#11

I also talked to the VP of Strategic Programs at the Linux Foundation recently regarding interoperability with other common data licenses but failed to obtain an answer. One would imagine interoperability would have been a design criteria?