[OSM-talk] The long tail - lowest common denominator

[Lars Aronsson]

Petter Reinholdtsen wrote:

> And I believe we should focus on the freedom of the users of the 
> OSM map.  To do that in the long term, we need to restrict the 
> same users ability to not share their improvements.

Language has a lot of traps, and is poorly suited for logic.  
Here you're talking about the "users" of OSM, even "the same 
users", as if the users were a constant in this equation, when in 
fact they are a variable.  Today OSM has a few hundred users.(*)  
If we shut down the project now, only these few hundred will be 
sorry.  But if the project becomes sucessful, it might have many 
thousand users in a year or two that will be able to benefit from 
the project's continued existence.  The current users are not at 
all the same group as the possible future users.

(*)Yes, I know the stats say OSM has 2486 registered user 
accounts, but I guess many accounts are empty and should not 
really be counted.  131 users edited nodes in the last month.

You can put any restrictions you want on me.  Put me in a basement 
and feed me bread and water and I'll continue to edit maps of 
southern Norway.  But if you put the same requirements on any 
newcomer to the project, I might still be the only one who edits 
maps of southern Norway a year from now.

This has implications on our privacy policy.  I'm already 
publishing all of my tracklogs in the open.  But if we require 
everybody to do this, it might limit how many are attracted to the 
project. Therefore, we the current members (most of us are geeks) 
shouldn't vote on the privacy policy based on our own preferences.  
It's rather the future users that should vote. I think people in 
general are more concerned about their privacy than the average 
geek is.  But don't ask me, ask them.

But it also has implications on how we license our data, albeit in 
another way.  I think we should be expansive: to allow and 
encourage as many uses as possible, commercial or not, as this 
increases the amount of people who benefit from our work directly 
(knowing about OSM, downloading planet.osm) or indirectly (by 
using mash-ups that are based on our data, without the end-users 
even knowing where the data came from).  It would be nice if our 
licensing could have a viral effect on closed map producers, 
especially if we could force Ordnance Survey maps to be released 
as copyleft.  But I think that effect is unrealistic.  It would 
also force us to spend a lot of our energy on legal action against 
any violations of our license.  As the opposite of expansionism, 
at the other extreme of the scale, we have the purist stance where 
only people of the true spirit of chairman Mao are allowed to 
touch our holy data.  That would be a sure way to keep our project 
small and unimportant to the world.

Developers of the Linux kernel can be elitist in a whole different 
way than administrators of Wikipedia.  Linux can be developed by a 
few dozen specialists, but Wikipedia *needs* thousands of editors.
And so does OSM.


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