A Blind Sample allows you to see a small example of the data inside a dataset,
without seeing any of the actual data.  The sample is random data created to
reflect the kind of data within the actual dataset.  Strings are similar
lengths, integers are in the same range, and floating point numbers have the
same precision.

The sample is not sufficient to reconstruct the dataset in any meaningful
way, but it does provide a realistic glimpse at the types of data in a
dataset.

There are actually two kinds of samples which can be requested from a data
asset:

  * Mock Data
    Exactly 10 sample records for the dataset.  The mock data can be
    hand-crafted by the data owner or random set created by the Blind Sample.
    Regardless, the mock data is always the exact same 10 records.

  * Blind Sample
    Random synthetic data that is representative of the contents of the
    dataset.  Unlike mock data, the number of records can be specified. and a
    new sample is generated every time the sample is requested.

    Additionally, the data owner can set a MockType to produce more realistic
    samples of some sensitive data types.  For example, fake social security
    numbers, addresses and names.  Using these continues to protect the privacy
    of the data, but produces a more representative dataset.
