[OSM-talk] OpenStreetMap Future Look

Peter Wendorff wendorff at uni-paderborn.de
Wed Jan 9 10:17:45 GMT 2013

Am 09.01.2013 10:42, schrieb Paweł Paprota:
> On 01/09/2013 03:05 AM, Frederik Ramm wrote:
>> My idea of dealing with this complex situation is to:
> This all sounds nice but has it been done on a scale that OSM needs it
> to work? Are there large projects that have this kind of organization
> that you describe?
> There are dozens of large open source communities (Apache Software
> Foundation, KDE, Ubuntu etc.) that we can learn from. I would say more -
> not only learn from but by looking at those projects we can see what's
> possible. Some of those communities have been around for a decade or
> more - my theory is that it is safe to assume that between them all they
> tried more or less everything in terms of organizing themselves.
> So what we can choose from is out there today and working. Everything
> else has failed the test of time. 
Or it has never been tried.
You cannot say, everything that does not yet exist is impossible - and 
even if you do, OSM exists, and until now it's working and possible.
I would be very careful to say, Apache, KDE, Ubuntu etc. work better 
currently, because I'm not sure.
OpenOffice, now part of Apache, is well known, sure - but I think, still 
more people know Microsoft Office than OpenOffice.
Ubuntu is well known, but with increasing critics due to the trials of 
Canonical to make money and adding advertisements, product search and so on.

OSM has some drawbacks: You cannot get the one voice of OSM as a 
journalist, as it may depend who you are talking to, and there may be 
completely different opinions on some topics, but what's the 
alternative? Openoffice has at least one fork (Libreoffice), Ubuntu is 
one of many linux distributions, KDE "competes" with Gnome and others.
In contrast under the "hood" of OSM many projects are free to evolve, 
and except the license change which lead to fosm (according to the 
website that project is nearly dead or the website is not up to date, I 
think) OSM still is one.
I think, the main reason for that is that as soon as you cross the 
technical hurdle (yes, that's not that easy I fear), you are free to do 
your own stuff in the osm universe without forking the project.
> I don't think we want to be pioneers
> and try something enitrely new that has not been done or, even worse,
> try something that was tried before and failed (why repeat mistakes of
> others?).
I don't see any example why the osm way failed before?
The examples you gave are projects that did not try to stay with the 
volunteers-with-less-leadership model. Sometimes that's necessary, and 
for some project types I'm sure it is (e.g. software development at some 
stage), but in some parts we have that already: Osmosis can be extended 
by plugins, but there are maintainers who decide what goes into the 
core. The rails ports code is free, and as mentioned above, while 
welcome, changes have to be explained and good arguments have to be 
given to settle them into the master on osm.org.
We have this structures, but on a lower level.
OSM in parts is more like a free community of people and of groups of 
people, who are doing their stuff in cooperative, communicative and 
collaborative way than a centralized moloch of decision makers who 
reject what other people do because it does not fit to their opinion 
what should be done by someone.


More information about the talk mailing list