[Imports] Making the best use of available data - a usecase without a solution?

Andy Allan gravitystorm at gmail.com
Wed Aug 1 16:57:59 BST 2012

On 1 August 2012 02:11, Paul Norman <penorman at mac.com> wrote:

> Does anyone have suggestions for a way to make the data available to users
> in a way that they can select an area, bring in appropriate objects and
> upload the result?

Yes, we have that workflow available now, using

a) snapshot-server to host and serve compatible data.

One snapshot-server can serve an unlimited number of discrete
datasets. It means that people don't all need their own copies
(shapefile, osm, whatever) of the entire dataset - they can just
map-call the particular area that they are working on.

b) Potlatch2's vector background layer, with the snapshot layer loader.

This allows you to "bring through" third-party datasources, preserving
names and other attributes. There are a few different options for
background layers, including local osm and shapefiles, but with the
snapshot loader you can talk directly to a snapshot-server. Moreover,
the third-party datasets can be predefined in p2 so that users can see
all the available datasources for their area.

In addition to your requirements, the combo can allow osm users to
assess and track which features have been merged in osm (or rejected
etc), so it's a great way to work through a particular dataset rather
than one person having to do everything. We've been using it in the UK
for a third-party datasource of bicycle information, which has a high
level of both duplication within OSM, and features in the dataset that
are represented slightly differently within OSM.


Let me know if you have any questions!


More information about the Imports mailing list