On Thursday 06 January 2011 7:14:00 am Bill Moran wrote: > In response to Scott Ribe <scott_ribe@xxxxxxxxxxxxxxxx>: > > On Jan 6, 2011, at 1:52 AM, Stuart Bishop wrote: > > > If you are looking at these extreme > > > improbabilities, your SERIAL isn't guaranteed unique either when you > > > take into account cosmic rays flipping the right bits in your ECC > > > memory or on your disk platter. > > > > Yes, that's rather the point, the probability is so extremely low that it > > in most cases it should be treated as 0. Some people seem to have a > > problem wrapping their heads around relative magnitudes that extreme. > > > There. I Godwined the damn thing. > > -- > Bill Moran Maybe a wrap up is in order:) As I said earlier this is one of those arguments that could go forever because everyone is right, so to summarize: 1) UUIDs can have a very to extremely large namespace but less than infinite. 2) There are other alternatives i.e SERIAL 3) Managing the above is based on the interaction of three components - software,hardware,wetware(people). Any one of which can have a weakness and in combination their are many permutations. 4) DBAs need to plan for the worse. Worse being somewhat contextual. Real time control of a nuclear plant versus Web social media. Choosing a unique number generator and dealing with possible collisions is contingent on this context. -- Adrian Klaver adrian.klaver@xxxxxxxxx -- Sent via pgsql-general mailing list (pgsql-general@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-general
