[talk-au] NSW GNB licensing

[Andrew Harvey]

On Sun, 30 Apr 2017, at 11:10 AM, cleary wrote:
> That's good to get the copyright issue fixed.
> 
> However, as I noted in a previous post, I have reservations about GNB
> data. Co-ordinates are often incorrect, sometimes by kilometres (even
> showing a couple of NSW locations as being across state border in
> Victoria) whereas the NSW LPI Map seems to show locations accurately.
> 
> Also some (possibly many) GNB names and places are historic- if added to
> OSM, I think they should have tags such as "abandoned" added. In some
> cases there are not even ruins that can be identified. While continued
> use of the name for an unpopulated area can be labelled as
> "place=locality" in OSM, some GNB place names are not in use and not
> known to local residents, so I question whether they should be mapped.
> 
> So I welcome the ability to use the data as a reference but caution
> about adding features to the map that have not been surveyed or
> otherwise verified.

I agree with this. I wasn't advocating any kind of import or blind use
of the GNR in OSM, rather ensuing current and historical use was
permitted and can continue.

I think the GNR data can still be useful when combined with other
sources like local knowledge or ground survey, but like any other non
ground source, it should be used with caution and accounting for issues
like you've raised.

I'm always interested in comparing OSM to authoritative government data
so I've been working on https://github.com/andrewharvey/nsw-gnr-osm to
try to convert GNR into the OSM schema, it may be of interest to others.