>>I had never tried to redefine a sequence object as a four-byte integer. To be honest, I think it's nearly dangerous if they do allow it to be redefined. If you have enough tables in a database, you increase the possibility of having 2 billion+ rows.
Fair enough, so you might prefer more than one sequence.
>>As for being more efficient than a GUID,, well, that's debatable. Yes, the footprint is much smaller.
We did some volume testing. Measured in nanoseconds to retrieve bigint sequence result. GUID is not quick. The other issue is that the results are saved into the table ascending sequentially which (as you note) is an index benefit, unlike GUID.
>>A sequence object will only be unique to a database
Right, but for the purposes of a FK before inserting the parent row, that's useful.
"... They ne'er cared for us
yet: suffer us to famish, and their store-houses
crammed with grain; make edicts for usury, to
support usurers; repeal daily any wholesome act
established against the rich, and provide more
piercing statutes daily, to chain up and restrain
the poor. If the wars eat us not up, they will; and
there's all the love they bear us."
-- Shakespeare: Coriolanus, Act 1, scene 1
