[Geodata] Fwd: LINZ Data upload for OpenStreetMap will go dark on April 1

[Hamish]

[cc from the NZOpenGIS mailing list for informational porpoises, as this is a world-wide issue]


Hi,

well we're going to have to deal with this at some stage. Under
current rules the LINZ Data Upload account is going to be locked out
of OpenStreetMap starting April 1st. (we don't own the Crown's copyright
and can't accept the licensing agreement or [effectively] reassign
copyright to the OSM Foundation on the Crown's behalf) And so some time
after April 1st all of our hard work would then be deleted from the main
database.

we are not alone, Australia is in the same boat:
 http://lwn.net/Articles/422498/

and this thread is required reading:
 http://lwn.net/Articles/422493/
it is worth it I think to read that in-full. Especially as it contains
a series of posts by Mike of the Creative Commons ("mlinksva")
clarifying their position and concerns, and well, they are/have
lawyers who know what they're talking about. [public domain for OSM is
not their recommendation after all]

To be honest I find it all highly demotivating. Reading through Mike's
comments does give me a bit of hope, though, I would suggest we all
make a few final CC-by-SA .osm planet file dumps as backup by mid-late
March, before any of our hard work gets deleted or non-license
compatible data gets added to it. I assume the extract servers will be
overloaded March 31st.


this has all been endlessly debated before, but we can't ignore it
anymore. fwiw my 3c:
 (putting my bias clearly on display :-), but hopefully my reasoning as well)
<opinion>
- OSM map data is not a "fact", it's an abstraction of facts by its
very definition, with a lot of human choices about what to include,
how to call things, how many nodes to put into that corner, etc. but
even if it were like an exact photograph instead of a painting, a
photo of reality still has valid copyright. see above LWN thread for
better comments on that by "epa", Mike, and myself.
- I've read the ODbL from end to end (a while back, may have been a
final draft), and my impression at that time was that it was a big
complicated pile of poo. the lawyers for EvilCorp would have a field day
with all its loopholes and contradictions.
- the new contributor terms are ugly too. (yo' momma)
- effectively the data is going from a GPL-like strong copyleft to a
LGPL-like weak copyleft. (I say that from a reading of the new
license/CT; not from what has been said about it). While not my
preference that isn't a deal breaker for me, as long as it's clearly
understood & out in the open. (but it's not, this has been sold to the
community as a way to strengthen the copyleft)
- I don't know enough about Contract versus Copyright law to comment
fully, but the "breaks open" vs. "breaks closed" and Privity of
Contract arguments by landley in the above LWN thread gives me much
concern.
- how do they enforce a click-through agreement for the Xapi wget
interface and still have it be useful?
- EULAs are the legal equivalent of a junk bond. like it or not, copyright
law is the strongest of the (c)/contract/trademark/patent family of IP
laws, so the best we've got to work with.
- just because you've done all the hard work to load the gun and find
a target, if you find it is pointed at your hunting partner's head you
don't actually have to pull the trigger, even if it does mean you've
got to go to the effort of finding another target, or not firing at
all. (emotive enough for you? ;-)  but seriously, fait accompli
rhetoric isn't worth much.

and finally..
there's a region by region ranking of how many users have accepted the
new terms:
 http://www.odbl.de
I'm proud to say NZ/Aust ranks second (only Albania beats us) in not
saying yes to it. but I think the methodology on that site is highly
highly flawed, as they only look at who made the *last* revision,
ignoring the "derived" part of derived work. So the non-version-1
numbers there are actually useless and highly misleading. Even if they
follow the history tree back fully to weed out what are map features
derived from non-agreeing sources, when Ways are split in osm, only
half the Way retains the history, the other half becomes something
new. And that's a one-way street. The new half appears to the system
to be an original work while it is actually not. So even if they
followed the modification tree fully for that analysis, they'd still
missing many many CC-by-SA-only contributions. A deeper algorithm
might be able to sort it out, in theory, perhaps, but I'm not holding
my breath for that.
</opinion>


best,
Hamish