[talk-au] ODbL data.gov.au permission granted
mike at ayeltd.biz
Tue Nov 15 14:09:03 GMT 2011
On 15/11/2011 11:52, Andrew Laughton wrote:
> I fail to see a contradiction. If you are not sure about
> something, you ask explicitly and get an explicit answer. That is
> what we got. That is what is written on the wiki with the kind
> assistance of data.gov.au <http://data.gov.au/>.
> If it helps, me formally affirm and represent what I have said
> before: I have had a series of correspondance with data.gov.au
> <http://data.gov.au> where: 1) I have explictly pointed out we are
> moving to another license specifically written for open data, that
> it might not jive with CC-BY and so they may not be happy with the
> provisions for downstream attributions, and asked them if they
> could explictly give us permission to continue use or if we should
> remove it; 2) The conclusion being yes, we can "incorporate and
> publish such CC-BY licensed geographic coordinate datasets under a
> free and open license, including the Open Database License,
> provided that primary attribution is made here
> and that each dataset used is also listed here in the format
> /Dataset Name, Date Published, License, Agency Name, originally
> retrieved from http://data.australia.gov.au"/; 3) For public
> transparency, the operative version of the statement is not in the
> correspondance but directly drafted at
> and actively reviewed by data.gov.au <http://data.gov.au> to their
> Hi Mike
> I might be able to help a little.
> The words "... provided that primary attribution is made ..."
> Would seem at first glance the exclude any license that does not
> require attribution.
> Perhaps you could explain to us what happens if a third party takes
> OSM data, and publishes it without any attribution at all.
> Would they be in violation of the Open Database License ? If not,
> the problem is that you are now distributing government data in
> violation of copyright law.
Richard has just answered the direct question, and better than I can.
May I make a couple of other observations that may also help.
First, The new OSM contributor terms were deliberately written so that
if anyone wants attribution, (you, me, but chiefly aimed at goverment
organisations making data available), then they can get it and it
survives any license that OSM uses, (even one without attribution!).
This was the key question I put to them, was this acceptable by itself?
and the answer was yes.
Second, a personal perspective of one thing I've learnt slowly over the
last three years. Open IP data in particular, as compared to software
and general creative works such as photos and writings, needs
attribution breaking up into a series of levels and the original
publisher of data should state clearly what they want. I think
(personal view) that this underlies a lot of the travails we have had
and that reviews of Creative Commons and OpenDataCommons (ODbL) licenses
need to pay attention to it ... i.e. important no matter what project
you happen to be in. I started a paper
http://docs.google.com/View?id=dd9g3qjp_103fdxjk3qt which went to
Creative Commons as a way encapsulating some of our experiences ... it
is still very rough, but I am continuing to work it as formal input to
the CC 4 process. In summary, what I think is eventually required is
are license(s) whose attribution clauses clearly allow or do not allow this:
o OSM incorporates data.gov.au data into the OSM database. OSM
attributes data.gov.au on its web page, "primary" or "Level 1"
attribution. I think this is very important whether or not a legal
o Someone extracts some OSM data (one node or the whole thing, with or
without data.gov.au), and no matter how they do it, they get a either a
copy of OSM's web page or a link to it. This is "Level 2" attribution. I
feel that this is useful, practical but not essential.
o That person or organisation mixes OSM data with their own and
publishes it. They Level 1 attribute OSM on their website. There is now
an attribution chain back to data.gov.au ... I think that is the key
thing. Level 3 attribution, providing data.gov.au attribution, should
not be required.
o A new person then makes a map from the last set of data, they Level 1
attribute the last set. Here, OSM's attribution has also now directly
gone but remains in the chain and the Level 4 requirement of having to
attribute thousands and thousands on a single map is not an issue.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Talk-au