[Talk-GB] UK government postcode/geolocation/nhs information leaked

[Tim Waters]

2009/9/16 John Robert Peterson <jrp.crs at gmail.com>:
> We may not be able to "use it" but mabee we can use it to highlight
> inacuaracies in our data, how do the legalities stand whith that sort of
> thing? that could clean up the postcode data awesomly.

Hehe, I think using it without asterisks is still using it and quite wrong :)
In OSM, we don't need or want to use proprietary data to make our free
map. Freethepostcode and npe-maps are both efforts to free up postcode
information regardless of the powers to be. Similar to how OSM didn't
hang around waiting for the OS to release Mastermap, these efforts
aren't waiting for the Royal Mail to release the data. And the most
powerful point of doing it ourselves is that we really no longer need
or want the OS to release their data anymore. Ours is better.

So, whilst it would be illegal, and wrong for us to take that leaked
dataset and use it, all academics and students in the UK do have
access to the very same, or better, dataset and could do such a study
quite easily and publicly - similar to Muki's comparisons of Ordnance
Survey's datasets and OSM. Might not clean up our data, but it could
be a good starting point to help improve our own data collection
efforts with postcodes.

> Would this be a good time to petitioning the government to give us access to
> this data, after all, I'm sure we have people around here that are capable
> of wording such a request convincingly.

Whilst I think internet petitions are rubbish, and requests for it to
be released have been written several times over the years, it is a
good time to raise the issue of closed public geodata in general. The
guardian's free our data campaign for example have been banging away
at this for some time now.

Incidentally, what I would be interested in finding out is how many
trap postcodes are in the dataset. Together with the number of
postcodes that are changed added and deleted with each iteration, and
their list of who (and therefore who hasn't) has a licence for the
database, would it be enough to fingerprint that wikileaked dataset?

Cheers,

Tim
http://thinkwhere.wordpress.com