[Openstreetmap] Wikimania, August 4-8

Lars Aronsson lars at aronsson.se
Mon Jul 4 23:32:54 BST 2005


Arnulf Christl wrote:

> SteveC wrote:
> > I'll be talking about OSM at another conf
> > (http://www.ukuug.org/events/linux2005/) on the same weekend.
> > 
> > I spoke to Jimmy Wales http://en.wikipedia.org/wiki/Jimmy_wales about
> > OSM some time ago. He was uninterested in doing anything unless it was
> > part of the wikimedia foundation.
> 
> Yes we also tasted some of this attitude when we started off 
> last year all enthusiastic about a Wikimaps architecture (still 
> to be found on meta).

I would also be very skeptic if I were Jimbo.  (I'm user:LA2 both 
on Wikipedia and OSM.)  There are dozens of project ideas on 
meta.wikipedia.org, many of them with no more substance than 
"let's do something with maps".  Most of them are linked from 
http://meta.wikimedia.org/wiki/Wikimaps

I think a successful approach will have to do two things:

1) make clear that it has a grasp of all other existing ideas, 
because competing ideas must be killed off to create a focus,

2) show a practical road ahead that can produce useful results 
here and now, and not just a summary or critique of other ideas.

Your announced workshop, Arnulf, is the most promising answer to 
(1) and we need something like OSM for (2).

But just suppose we could get Jimbo to say (in a press 
announcement) that OSM is "the thing" and that all wikipedians 
should buy a GPS and submit their track logs and draw maps using 
the OSM applet.  Could the current OSM implmementation handle this 
influx of data, traffic and contributors?  Even if, say, Yahoo or 
Oracle supplied the necessary hardware?  Is the application "ready 
for prime time"?  Perhaps not.  I think OSM needs at least six 
more months of development and rethinking before this.  Is there 
any alternative to OSM that is ready to roll?

I think that even after a complete redesign of the OSM 
application, we will still want to reuse the track log data that 
has been uploaded.  The track logs are the constant, while the 
application is a moving target.  So a repository of 
open/free-licensed track logs could be a first step.  But are we 
done with the licensing and privacy issues yet?  I don't think so.


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




More information about the talk mailing list