[Foundation-l] New project proposal : wikimaps

Daniel Mayer maveric149 at yahoo.com
Thu Jan 20 19:55:32 UTC 2005


--- Magnus Manske <magnus.manske at web.de> wrote:
> I hereby request this as a new project (we already do have wikimaps.org, 
> right?).

I think this is a bit premature. Wikimaps should be as human editable as
possible and I can't really figure out how to do that with the implementation
you presented (and I work as a full time GIS specialist).  

Much more time is needed to develop something workable, IMO. Did you contact
the person I mentioned in one of previous posts on this? 

There is also a lot that goes into forming a new project. One cannot just ask
and have it happen anymore. 

> IMHO a separate, single server would be best suited for this. With a lot 
> of data, map generation will slow down, and it should not slow down 
> other projects in turn. Also, due to the speed of the wikimedia projects 
> (or the lack thereof), I propose waiting for the next batch of servers 
> to be installed before doing this.

We have very serious server/performance issues to deal with first. 
 
> There's the question of the license for the data we will collect. PD? 
> GFDL? CC-BY-SA? If we say geo-data should be public domain, can we 
> really slap a "GFDL" on it? Or maybe we should dual-license out to-be 
> data CC-BY-SA and GFDL, to give maximum freedom while ensuring the data 
> stays free?

My experience working in the field has convinced me that PD is not at all the
way to do. Way too many people take PD GIS data (such as from the U.S.
Government), make a few improvements and then place their whole version under a
proprietary license. And this is done literally *hundreds* of times, which
heavily divides effort by making a great many incompatible forks. 

In fact, needless replication of incompatible effort like this is perhaps the
biggest problem faced in my industry. For example, street centerline files are
created for the whole U.S. by the U.S. Government. These files are often fairly
buggy. So each mid to large-sized city has their own GIS staff correct their
part of the file - but these corrections hardly ever find their way back into
the country-wide files (I've never heard of a single case). Multiply that
several times to get the number of non-local government forks of the same GIS
data (who then sell their versions at a high price). 

A copyleft license would encourage and much better enable positive feedback
into improving the GIS files. If not, then people will just treat our work the
same as they do PD U.S. government work with all the associated forking. 

This situation will not change until/unless the U.S. federal court system rules
on whether or not all GIS data should be covered by copyright at all. IMO,
slavish tracing of streets and natural features has no creative merit and is
thus not eligible for copyright. Yet copyright claims and lawsuits to protect
those claims are very common. 

So we must live within the world that is presented to us and wait for the day
when that world attains some sanity. Until then we need to protect our work
through the use of a copyleft license. 

-- mav


	
		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail



More information about the foundation-l mailing list