[Imports] [NUUG kart] kartverket imports to OpenStreetMap

[Gnonthgol]

Den 18. okt. 2013 00:02, skreiv Sverre Didriksen:
> I used to like working with OSM, but after my user was blocked I 
> don't. Hopefully it will come back. So for now I will leave the import 
> to you guys.
The block is just temporary, just talk to someone in the Data Working 
Group if you have questions. I would be pretty sad if you left the OSM 
community after this. I love your work and appreciates your input.

> I don't know if your example with the 50 000 nodes was mine, but just 
> in case; you of all should know that it failed because my user was 
> blocked by you. The nodes are always transfered first and the ways and 
> relations etc at the end. So a break like that will always make it 
> fail. And by the way, it was no problem reverting it.
Don't worry about that changeset, it is not important and it is no 
problems reverting it in the current state. The block were probably done 
to prevent further damage in case the user were not responsible enough 
to work together on this import (in all fairness most people uploading 
50 000 nodes without any discussion are pretty reckless).

I hate the import guidelines as much as anyone when it is directed 
towards me. If the data was realeased a couple of years back when the 
effects of undiscussed imports were little known I would probably have 
imported the data with little or no discussion. However I now see that 
it is pretty important to iron out all the small problems with an import 
before it is being implemented, failing to do so will cause a lot of 
pain later down the line.

As far as this import goes, I understand that you have been trying to 
import data from the N50 dataset. This is a vector dataset meant for 
rendering and therefore have a lot of small problems when importing it 
into a dataset like OSM that have to consider other uses. We have 
already started to discuss some of these problems but there are still 
things to look into. I have been reading the documentation for the data 
we have been given for almost two weeks now and have just begun 
understanding the way they handle some of the features.

Some people think that N50 is impossible to import into OSM in any form 
because the differences in the way the datasets are. It might be that we 
have to find alternative ways of using the data to assist with the 
improvement of OSM. I hope that we can figure out a way to use the data 
in a reasonable way so that we do not make OSM worse then the 
alternatives, but we need help from experienced mappers and data 
consumers to do this.

Gnonthgol