[open-linguistics] META-NET Data Liberation Campaign

Wed Dec 5 20:59:56 UTC 2012

On 5 December 2012 19:28, Nancy Ide <ide at cs.vassar.edu> wrote:
> PS I should have added to my previous note:
>
> On Dec 5, 2012, at 1:22 PM, "Jimmy O'Regan" <joregan at gmail.com> wrote:
>
> CC-BY, would make a better option for your aims.
>
>
> Absolutely--CC-BY is what we use throughout the ANC project for data, and
> Apache 2.0 for software. I didn't realize that this discussion concerned
> data only--I thought that software was included as well.
>

There's generally less confusion around software licensing. The same
research-only/non-commercial use only terms for software can simply be
dismissed as 'not open' with much less discussion.

> People use GPL for data though, so (just out of curiosity) why not Apache
> 2.0?
>

The GPL is a worse choice of licence for data. Licences specifically
written for software, generally, are poor choices for data. In
particular, clauses relating to warranty and patents, which are
irrelevant to data, can be a cause for anxiety for data consumers who
have not realised that the licence was not designed for data.

People who use the GPL for data typically have been doing so since
before an acceptable alternative was available (versions of CC-BY-SA
(and CC-BY) before version 3.0 were regarded in open source circles to
be fatally flawed), or do so in cases where the data is tightly
coupled to software.

-- 
<Sefam> Are any of the mentors around?
<jimregan> yes, they're the ones trolling you