How do I group different rows in a field?

What's more efficient when working with mySQL? Save a date in unix timestamp within an integer field or using unix_timestamp() function on a string field with a date formated like 'Sat May 12 22:52:01 +0000 2012'?  I will be using this field to count and group rows.

• Answer:

  The integer, which is 4-bytes, blows the string away, which I counted to be 30 characters which usually ends up being approximately that many bytes. Also note that there is a TIMESTAMP column type in MySQL that is used for this very purpose: it stores unixtimes in 4 bytes and lets you use all sorts of nice date manipulation functions instead of manipulating the integers yourself. TIMESTAMP is the most efficient way to store a timestamp, that is until 4 byte unixtimes wrap around in 2038, at which time it will be Y2K all over again.