[Imports] edit import guideline wiki page

Andy Allan gravitystorm at gmail.com
Thu Nov 7 09:27:59 UTC 2013

On 6 November 2013 23:56, Daniel O'Connor <daniel.oconnor at gmail.com> wrote:

> I understand the problem we are trying to solve is preventing harmful
> imports, so the map data is of high quality.


> I think it's fair to say
> everyone on this list is here to help and uplift a user to the point they
> can import

Well, this is where I disagree. My belief is that we're still
approaching this from the wrong angle, and in fact with our "import
guidelines" and "imports" mailing list, we're pre-supposing that
"importing"[1] is the correct answer to the given problem.

So taking a step back, I believe the given problem is "I have an
external dataset, and I want to use it to improve OpenStreetMap". But
usually that rapidly becomes "I have an external dataset, and I want
to copy that information into the OpenStreetMap database" when there
are, in fact, many other possible approaches.

* Use the dataset to power QA tools, c.f. Ordnance Survey Locator Gazetteer[2]
* Use the dataset as a visual reference, c.f. GPX tracks which we
thankfully have never imported as nodes
* Use the dataset as a vector reference, c.f. Snapshot Server[3]
* Use the dataset as an import source, c.f. Tiger, AND, etc.

So no, I don't agree that I want to uplift every user to the point
they can import - but I do want to uplift the user to the point that
they can make the best use of their external dataset. As time goes by
I expect few and fewer datasets will be imported under any sensible
meaning of the term, but those datasets will still of course be useful
to us.

I ask that you and others bear that in mind when rewriting import
guidelines - to us it might seem obvious that an import is only one
approach, but we should ensure that anyone who approaches OSM with an
external dataset is clearly guided to the different possible


[1] For everyone who approaches OSM from the outside, to "import" is
the wholesale copying of external data into our database through a
process driven by themselves. The other approaches are rarely obvious
to them.
[2] https://wiki.openstreetmap.org/wiki/Ordnance_Survey_Opendata#OS_Locator
[3] https://wiki.openstreetmap.org/wiki/Snapshot_Server

More information about the Imports mailing list