[OSM-talk] about freedom in PD/BSD/MIT/Apache (was: The long tail - lowest common denominator)

Lars Aronsson lars at aronsson.se
Fri Jul 7 23:59:07 BST 2006


Immanuel Scholz wrote:
> I do not agree to any non-viral license.

On the other hand, I can agree to almost anything.  I do 
appreciate the recursive elegance of GPL and the power it has 
demonstrated in cases like Opentom.org.  But in my mind, it is a 
bigger lack to have a license that we cannot enforce than having 
one that is non-viral.  And what chances do we really have to 
*fight and win* against anybody who breaks our current license?

Selling the TomTom based on the Linux kernel without releasing the 
source code for the device drivers was clearly a breach of the 
GPL, and the company was forced to comply.  But are mash-ups of 
OSM and Google Maps a breach of our license?  If we would release 
planet.osm in the public domain, we wouldn't have to worry.  If we 
require a viral license, how do we define what's allowed and how 
do we realistically fight any violations?  And is our time, money 
and effort well spent in such fights?

If the license statement is there only for aesthetic reasons, we 
could just as well claim that all data must be submitted "in the 
spirit of chairman Mao", or any similar nonsense.

> For a public domain (or BSD, Apache...) license, I feel like 
> "just another data source for some big company". I have this 
> picture of some manager in mind, looking at OSM and says to

With this reasoning, I don't really understand how you draw the 
line to the "NC" component of the CC license.  What if the 
marketing guy says "wait, they have a CC-SA license" and then they 
find a way to comply with that, just like Red Hat Linux complies 
with GPL or like the publisher Directmedia Gmbh in Berlin complies 
with GFDL when they publish the German Wikipedia on DVD. What if a 
company complies with the viral license and still makes a lot of 
money from our unpaid work, is that ok?

I personally think that it will be a very long time before any 
company big enough to have a marketing guy would prefer to use map 
data from OSM over buying it from the likes of TeleAtlas.  In the 
foreseeable future, say 5 years, the only realistic uses for OSM 
data are web applications and amateur hacks that can't afford 
TeleAtlas map data.  That might seem limited, but I think it still 
is an important goal.  The commercial scenario you describe, that 
would perhaps make a difference between a viral license and public 
domain publishing, just doesn't seem very realistic to me.


-- 
  Lars Aronsson (lars at aronsson.se)
  Aronsson Datateknik - http://aronsson.se




More information about the talk mailing list