note 97474 deleted from function.uniqid by danbrown

Note Submitter: jaimthorn at yahoo dot com ---- On 04-Mar-2010, Robin wrote: "A longer and more diverse unique id in the same hexit scheme can be achieved by instead digesting uniqid() (or a series of mt_rand(), f.e.) through md5() or sha1()." Although what Robin says is correct, this statement should not be mistaken to imply that creating a md5/sha1 digest of uniqid() is more likely to create a unique sequence: given two uniqids, the md5s of those uniqids are the same IF AND ONLY IF those uniqids are the same. The md5s of two *different* uniqids, however, typically don't resemble each other very much, even when the uniqids DO resemble each other a lot --e.g., when they were generated in quick succession.