Re: word lists
Werner LEMBERG <wl@gnu.org> writes:
> >> let's assume that I take the first volume of `Harry Potter' and do
> >> the following: […]
> > […]
> > Unless you are discussing a proposed package for Debian based on
> > Harry Potter, please don't speculate on hypotheticals about parsing
> > copyright law here.
>
> Thanks for your grumpy answer.
Thanks for responding. It was not intended to be grumpy. I'd describe it
more as “stern” :-)
> Yes, this is about a potential package, namely a word list of
> hyphenated German words; see
> http://repo.or.cz/w/wortliste.git
[…]
> (1) What happens with the copyright of works if I extract
> information as explained in my previous mail?
One important consideration is whether the result would, under copyright
law, constitute a “derivative work” of the prior works. (If it is, then
copyright in the prior works would restrict the actions of anyone
dealing with the derivative work.)
That question – is it a derivative work from the prior works? – could
only be answered decisively by a court decision on the matter, in each
relevant jurisdiction.
I don't know of any such decisions. Do you know of any relevant court
rulings on the matter?
> (2) What copyright can a word list have at all, given that it gets
> mechanically extracted?
Purely mechanical extraction implies that there is no creative
transformation, which denies the party doing that transformation from
having any copyright in the resulting work.
But the copyright in the prior work still obtains, and still has exactly
as much force in restricting derivative works, regardless of how it was
extracted. The question is whether the resulting work constitutes a
derivative work.
> (3) The frequency of words is a statistical information, a byproduct
> of collecting the data mechanically, without any manual
> intervention. Can such information be copyrighted, too?
The question, again, is whether the resulting work constitutes a
derivative work of the work on which it is based.
I don't know, but I'd guess no sensible court would think copyright
resticts statistical collection of word-frequency in a work. Do you know
of any relevant court rulings on the matter?
> (3) What license do you recommend for a list of German words (as in
> our project) enriched with information on weighted hyphenation
> points?
In general I recommend the more well-understood and definitely-free
licenses, like GNU GPLv3 <URL:https://www.gnu.org/licenses/gpl.html>,
or the Expat license <URL:http://www.jclark.com/xml/copying.txt>.
Beyond that, I don't know what specific permissions and restrictions you
are looking for in a license. Also, this list isn't much good for
license advice, beyond what will be welcome and acceptable for inclusion
in Debian.
--
\ “… whoever claims any right that he is unwilling to accord to |
`\ his fellow-men is dishonest and infamous.” —Robert G. |
_o__) Ingersoll, _The Liberty of Man, Woman and Child_, 1877 |
Ben Finney
Reply to: