I use their data to demo my medical software. I obviously can't show real patient names, so I import a whole bunch of realistic looking people from here instead.

I've used it in my software as well. Except that the names actually end up looking fake to me. Not sure if its cause I know they are fake or what. Doesn't matter to me either way.

Except it doesn't do addresses, especially postal codes, which is what I'd use it for. Making up names is not the problem, finding a matching postal code with street name and a house number (this database is published and many sites check it) is the issue.

There are lots of syntactically valid addresses at the bottom of Lake Washington in Seattle: useful if you want to avoid spamming random people with postal mail in case that contact information later gets sold on.

The idea is to bypass syntax checking while trying to be anonymous, not perpetrate identity fraud.

It does addresses, broken down by StreetAddress, City, State, PostalCode, and Country. But it sounds like what you want isn't fake identities, you want real ones.

You don't see addresses on this website? Look just below the randomly generated name.

Not in Dutch

I don't like the idea of doing this on a website. Why not just do it offline? Why fetch the data from a server where it could be captured and logged?

What data can they capture and log, beyond data captured and logged at any website?

Would be cool to have an API. Could be useful for mass address book creation.

Faker (for ruby here https://github.com/stympy/faker) originally for perl, available elsewhere also

Nice, anything like this available in Python or C# land? Or even a simple desktop app rather than a webapp?

There's something similar in python https://pypi.python.org/pypi/fake-factory

C# - https://github.com/slashdotdash/faker-cs

Very useful for generating fake data when writing test cases.

I was reminded of this as well.

Who would need mass address books?

People writing address book software.