Michael Glaesemann <grzm@xxxxxxxxxxxxxxx> writes: > On May 31, 2007, at 11:46 , lawpoop@xxxxxxxxx wrote: >> However, I wondered, if I were in an environment where there were many >> concurrent inserts, would it be possible that I didn't get the serial >> number of the insert that *I* just did? > No. Uh, yes, because he was using "SELECT last_value FROM seq" which is indeed subject to race conditions. currval() would be far safer. regards, tom lane
