[OSM-talk] Doing God's Work / Oliver Kuehn interviews Jack Dangermond

Frederik Ramm frederik at remote.org
Sun Oct 10 18:57:28 BST 2010


Hi,

Alexrk wrote:
> Frederik, you held a presentation at the "Amtliche Daten vs. 
> OpenStreetMap" track at the Intergeo? Could you tell us something about it?

There was a session with three talks of 20 minutes and then a combined 
questions block of another 30 minutes; the whole session was placed at 
the end of the last day.

The three speakers were Mr Ludwig from the Bavarian ministry of finance 
(they are the keepers of Bavarian geodata), Mr McCutcheon of DDS, a 
large commercial geodata Navteq reseller and provider of other 
commercial geodata, and myself. The mood was amicable; I think every one 
of us had some respect for what the others did.

Ludwig presented an impressive array of geodata held by the state of 
Bavaria - stuff that we could never dream of collecting (and lots of 
stuff that we wouldn't take even if offered for free, lots of stuff that 
could never be run in a crowdsourcd fashion). They have the kind of 
geodata that allows farmers to plan which crops to plant where, the have 
geological data, a precise DGM, and much more. Compared to what they 
have, OSM is more of a "mass market dataset". They must be spending an 
immense amount of manpower on developing and complying with standards, 
not only between the different federal states in Germany but also 
internationally. It is important to note that whatever they collect, 
they usually collect that for the whole state of Bavaria - not only for 
a few hotspots; and with 70,000 sq km Bavaria is not exactly small. Last 
not least, their data is often an ingredient in legal disputes of all 
sorts, so it needs to be "official". But between the lines you could 
also hear that Ludwig seemed a bit unhappy that all this data they have 
was put to relatively little use, and was generating relatively little 
headlines - obviously (to me) because it was not freely available. He 
said that they're already giving full access to their data to schools 
and he was very open to cooperating with OSM, however it was clear that 
he was hoping for OSM to settle down a bit and become more 
reliable/established, ideally use a fixed data model and so on.

After that, I gave a brief introduction into OSM (my slides are 
available at http://www.geofabrik.de/media/2010-10-07-intergeo-osm.pdf), 
highlighting how quite a number of government agencies in Germany are 
now using OSM data and asking the question why (could it be that 
official data is difficult to get your hands on even if you are the 
German Federal Ministry of Economics and Technology - who have recently 
launched a map displaying broadband availability based on OSM?). I 
openly acknowledged that OSM has its disadvantages - mainly that we do 
not offer any quality assurances and that coverage is often deep not broad.

In the last talk, McCutcheon provided some in-depth comparisons between 
commercial data and OSM data (and some government sources too), 
acknowledging that OSM was often superior to commercial data for leisure 
uses, but, as expected, lacking basic data in rurual areas. He claimed 
that roads in OSM were often lacking attribution (I didn't have a chance 
to discuss - I assume that for him, attribution would mean a minimum 
basic set of attributes including stuff like maxspeed because name and 
highway type are present in the majority of OSM roads). He named the 
share-alike license as a major drawback, but concluded that any GIS 
project needs to carefully analyze the needs and then decide which - 
OSM, commercial, or governmental - set of geodata might be best suited.

In the questions session I told the story of a German fire department 
who are now using OSM data in addition to their other stuff after they 
recently were unable to fight a fire that had broken out in a building 
under a power line and they couldn't find out the operator of the power 
line in time. After the fact, they were told that OSM properly had the 
operator of the power line noted which got them interested. I used that 
example to make the point that while the government may have all sorts 
of cool data, even the government doesn't know where to find it. Ludwig 
said that they're working on consolidating all the metadata they have 
(what with EU projects like INSPIRE etc.) so "in a year or so" nobody 
should have an excuse not to find the right dataset for the information 
they're after; to which McCutcheon replied that finding the right data 
for people was his core business model and he guessed that it might take 
a while longer than a year before you can easily find out where to get 
which data from.

We parted as friends; I'm sure that there will be further cooperation. 
One of the problems Ludwig had with OSM was that for him and his 
administration, OSM didn't have "a face" - he would really like to have 
one person to talk to and to build relationships with. Consequently, we 
have decided to make Joachim Kast (who already attended the recent 
Geodata summit at the Ministry for the Interior) our (German) official 
contact person for talking to government and administration. We still 
expect local mappers to take the lead and won't overrule them but 
whenever they reach a point where their counterpart wants to talk to 
someone "higher up" we'll send in Joachim.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"



More information about the talk mailing list