On 14 January 2011 22:55, Mark Goodge <[email protected]> wrote:
>
> I seem to have found myself in the position of defending - or, at least, of
> being perceived to defend - restrictive approaches to government data.

I think that's because you have lost track of what it is we were
discussing and ended up confusing yourself. Its easily done and
something I do all the time.

Consider the proposition that the land registry or companies house
registration should be made public (in the sense I have already
explained) and free. There might be an objection that this would allow
the discovery of facts about individuals that ought not to be made
public.

The counter to this is that _some_ facts about individuals are
appropriate to make public. The solution is to adopt an approach that
avoids those facts being part of the publicly available data.

Your confusion is in thinking that:

(1) the decision that it is OK for particular data to be made public,
means that our policy will always allow that data to be discovered
(2) (implied in what you said) the proposition "removing private data
from a dataset will make it less useful" somehow implies that the data
sould not be removed and the dataset kept private.

So, the fact that names aren't unique identifiers for directors isn't
relevant to whether the Companies House data ought to be released. The
most you can do with the Director(name, company) relation, is to infer
that a particular individual is the director of that company. Once we
have decided that that data is legitimately public, it doesn't matter
that you won't always be able to make that inference. Sure, it would
be nice to have better data that did the job, but that's another more
difficult question.

I am sorry if I am not being clear. If I was allowed to use Heyting
Algebras this would all be clear, but I fear that they are disallowed.

-- 
Francis Davey

_______________________________________________
developers-public mailing list
[email protected]
https://secure.mysociety.org/admin/lists/mailman/listinfo/developers-public

Unsubscribe: 
https://secure.mysociety.org/admin/lists/mailman/options/developers-public/archive%40mail-archive.com

Reply via email to