RE: unique row identifier data type exhausted . . .

On Wed, 26 Apr 2000, Tom Cook wrote:

> Is this necessarily a good solution? If you use 64-bit OIDs, some joker
> will just hook up a several-terra-byte disk array to his machine, try to
> store the location of every molecule in the universe and break it.

That's not going to work anyway. To store information about a molecule you
need at least one such molecule to hold that state, barring major
revolutions in storage technology. :-)

> Admittedly, ~2x10^20 is a very large number, but that's what they thought
> about 2000, also...

A while ago I said that in order to exhaust the oid space you need to add
1 million new records a day for more than 10 years. Then someone said, ok,
what if I have an email service with 1 million users that each get 10
emails a day. Then you're talking about 1 year. But in order to exhaust 64
bits, you can have 10^9 users (i.e., everyone), getting two million emails
a day for 1000 years. That seems pretty safe for as long as I care.

Of course to store all molecules you really need more like 384 bits.

> What I'm saying is, is there a better way of doing this?

Transfinite numbers ;)


--
Peter Eisentraut                  Sernanders väg 10:115
peter_e@gmx.net                   75262 Uppsala
http://yi.org/peter-e/            Sweden