On Sat, May 30, 2009 at 08:19:17PM +0200, P Kishor wrote:
> On Sat, May 30, 2009 at 7:03 PM, Christopher Schmidt
> <[email protected]> wrote:
> > On Sat, May 30, 2009 at 01:47:12PM +0200, Marc Wick wrote:
> >>
> >> >Do you think that ODbL is a suitable license for geospatial data? Does
> >> >it meet your needs?
> >>
> >> Definitely not. Data is either open or share-alike. It cannot be both.
> >> Share-alike data is NOT open. It is a serious restriction that prevents
> >> innovation and most interesting use cases from the beginning. It also
> >> makes users vulnerable to any kind of lawsuits.
> >>
> >> Don't get me wrong. Everybody has the right to protect their data, but
> >> they should not call it open if they want to prevent others from using it.
> >>
> >
> > This sounds like a classic "Viral" vs "Non-viral" discussion -- the same
> > one in Open Source code would be BSD vs. GPL.
> >
> > Is it your reasoning that the GPL is not 'open'? If so, then your
> > point here seems to be contrary to the point of the discussion.
> >
> > SA enforces certain restrictions -- as does "Attribution". Saying that
> > a license isn't open unless a user can do whatever they want with it --
> > essentially, that only Public Domain is 'open' -- seems silly and a
> > counterproductive point in this discussion.
> >
> 
> First, no discussion is counterproductive (well, ok, some other, but
> this discussion is certainly not yet counterproductive).

Being deliberately argumentative, or ignoring a point of discussion, in
order to espouse that Your Personal View is The Personal View is
counterproductive. The tone of the previous email struck me as having
the potential to be counterproductive, though obviously it depends which
way the discussion goes from there. Perhaps I was simply being overly
sensitive to something that tends towards a religious flamewar rather
than a rational discussion. 

> I may be misunderstanding him, but I don't think Marc suggested that
> _only_ PD is 'open.' I certainly don't think that only PD is 'open.'

I agree, but his statement that "Share alike ... is a serious
restriction that prevents innovation" could be applied to *any*
restriction on usage. Share Alike is not intended to prevent users from
using the data -- instead, it is intended to ensure that *all users of
the data* get the same rights as those who got the data directly. 

To say that share alike prevents people from using data -- with no
description of how or why -- is not useful. Why is it restrictive? What
aspects of share alike on data are you unwilling or unable to
participate in? 

The OdbL makes specific statements towards combinations which are not
subject to share alike in an effort to ensure that the data can be
proteceted -- see below for why this might be important -- while being
as open to innovation and usage as possible. For the use cases I find
most interesting, I've found the odbl to be a significant step forward
from the CC-By-SA license because it more explicitly states what you can
do to th data without invoking the SA clause, and many more of the
things I find interesting are clealry in the 'allowed' category than
with CC-SA licenses.

> Of course, everyone is free to license their data under whatever
> license their jurisdiction allows them to license it under, and
> whatever makes sense to them. However, in the context of scientific
> data, I do believe that SA's viral-ness can create problems. Data, in
> my view, are the non-IP building blocks of IP. By making data
> available under a PD-like data mark, unrestricted innovation is made
> possible.

I don't understand how this is significantly different from code. The
point of share alike is to ensure that all users of the data have the
same rights -- so that improvements to the data are shared, and so that
Eve, Inc. doesn't take the data, improve it, release a product that is
better than the source, and kill the source.

If OpenStreetMap data was public domain-ish -- or anything without a
sharealike clause -- the data could have easily been integrated into
commercial mapping providers. These providers -- specifically Google --
have already shown a desire to create community-centric data collection
efforts, from the methods used (many users editing via a web interface)
to the naming of the events for collecting data (mapping parties). can
anyone seriously say that OSM would be able to be as successful as it
is, if GOogle were able to take the data, integrate it into their map,
adn devote their significant resources to further collection of data --
*without* making it freely available? At this point, I would say tha
tthe primary thing that keeps the OSM community going is that they do
continue to have a unique and open dataset that is extended by the
community: if the data could be easily sucked into a larger framework
and made "free enough" (as in beer, as in tiles, as in Not The Whole
Dataset) for most users, th OSM project would likely suffer, in my
opinion.  

Now, I'm not saying that Google -- or any other mapping company -- would
do such a thing. It is possible that they would instead work with the
OSM community to build it up. But given what appears -- to me at least
-- to be direct competition with OSM even *without* the ability to use
their data without share alike restrictions, I can't imagine that
reomving those restrictions would have caused Google to be more
supportive and less directly competitive.

> My view is -- think of what is served by the license you are about to
> apply. Do you gain anything? If you collect a few GPS points, what do
> you gain by licensing it under BY or SA or ND or NC? Are you going to
> become rich? famous? Are you going to ensure future innovation?

I believe that the continued efforts behind the OSM project have led to
a significant number of innovations both inside and outside the project,
in code and in data. I think that if the data were not to be sharealike
licensed, it seems likely that it would have been trivial to remove all
need for OSM for the vast majority of people by simply 'taking over'
the role of the project, if you are a large map-providing organization.

> I certainly don't believe that I will become rich or famous by
> licensing my data under a license that actually ends up throttling
> innovation even remotely. By putting my data out in a PD-like mark, I
> actually ensure that my data remains permanently open. 

I don't deny that -- so long as the 'you' here is an individual entity,
this is likely true, and makes good sense.

> In fact, PD is the most concrete assurance of forever-openness. Sure,
> someone else can take my data, add value to it, and make that
> value-added version restricted. So what? In fact, I believe that is a
> good thing, and because my data are in PD, everyone else in this world
> has the same opportunity.

The "So what?" comes not from what happens when this happens to your
data, but when this happens to a dataset gathered by a large community.

The effect on the data is nil. The effect on the community could be
large, and to not protect the investment of the community from being
co-opted by an organization can have significant value. 

Regards,
-- 
Christopher Schmidt
MetaCarta

_______________________________________________
Geowanking mailing list
[email protected]
http://geowanking.org/mailman/listinfo/geowanking_geowanking.org

Reply via email to