> a) Not sure how you'd explain hashing and salting to someone.
I'm not 100% I've used the right terms, but a hash would be 'a unique
identifier for each address which doesn't reveal the address but shows
whether two addresses are the same' - avoiding discussion of what
actually happens behind the scenes and neatly sidestepping the
question of salting. :)

> b) With only a few tens of millions of addresses, even with a salt, it could
> be trivial to brute-force someone's address hash. You'd have to estimate the
> current and future cost of the resources involved.

With a sufficiently large secret salt (hundreds of bits, I should
think), it should be infeasible. Or we could potentially encrypt the
normalised address (which I think is broadly equivalent and should
neatly avoid any issue of collisions!). Granted, if the secret got out
then there would be major issues (think of any of the many 'X
department lost Y data on Z mode of transport' stories), but this
secret isn't something that should ever need to leave a datacentre.

These are ways of doing it, but I'm fairly sure they're not the ways
of doing it *right*.

Dave.

_______________________________________________
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