[OSM-talk] Import guidelines & OSMF/DWG governance

[Lester Caine]

Pierre Béland wrote:
>  > Pierre - I'm not arguing against imports. Only unmanaged ones and ones we do
> not have easy access to the source data.
>  > As I understand it you can view the canvec data, but is it available as an
> overlay in an editor? That is the part of the jigsaw
>  > that I'd like to see handled better, so we can compare data against the
> existing map prior to any import, and are ABLE
>  > to analyze just what of the data can be imported directly and what needs to
> be merged in some way? Certainly a large section
>  > of the OS data is only useful as reference material and any import is only
> going to obliterate more accurate data, so having it
>  > available as an overlay works well.
>
> Lester - The National ressources department is collaborating and produce OSM
> files from his topographic data. The community has established guidelines. In
> general, contributors edit this file into JOSM, comparing with what already
> exists.  It is not an easy job.  But these contributors have made fantastic
> efforts.  We see too ofteen dogmatic declarations against imports without any
> nuance.
I would certainly argument against a formal 'demand' for a raw import of some of 
the OS layers into OSM and we have the tools to explain why we don't want that 
data. Having worked through large sections of my local area cleaning the 
licensing issues I was remapping things with 'source=OS' which are just stylised 
versions of situation on the ground, I can support that statement. I totally 
understand the 'It is not an easy job' so if you are happy that data available 
IS accurate enough to use directly and have the tools to show that then I have 
no objections.

> What we need as an organization is to establish governance practices that are
> efficient.  I am jealous of all the tools developped by the France community.
> The Talk-fr is very active and they are doing a great job. If you are not
> convinced, just look at the map of France.
The areas I have looked at are as 'complete' as those around here. The next step 
in both countries is to more accurately map the finer details. Something which 
is certainly not available from OS mapping so are details such as the exact 
configuration of a road junction with lane detail and pedestrian pathways 
available from third party data in France?

> And about governance,  if this community cannot manage his contributors, who
> can?  We continually have new mappers, some working more or less intensively. We
> should adapt or organization to this Wikipedia like structure and try to better
> structure local communities.
I certainly agree with the statement, but would strongly lobby against the 
'wikipedia' approach to solving the problem. New mappers NEED to be directed to 
proper guidance on how to provide new data, and I have proposed in the past that 
new data is ring fenced until a more established mapper can review it, much like 
we have in hg and git code management. At the very least a 'Do you wish to save 
this to the main database' warning would be appropriate at times until a new 
account has established some 'kama' in the data submitted? Importing data from 
third party sources should be something that does require 'kama' in 
understanding what one is doing and oversight by others should be added before 
some automatic processes are applied to the main database.

Some better involvement of local groups would be useful here I think?

-- 
Lester Caine - G8HFL
-----------------------------
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk