ctcgag_at_hotmail.com wrote:
> Karsten Farrell <kfarrell_at_belgariad.com> wrote:
>
>>Well, it might be different now ... but when I got my SSN (admitedly a
>>very long time ago), the first 3 digits indicated what region of the
>>country you were in when you applied for one. The fact that it has that
>>wee bit of "smarts" designed into it makes it a natural key. It's not a
>>truly generated key ... except maybe within the region.
>
>
> That's the same as for OPS when you use CACHE and NOORDER. Each local
> 'office' gets a chunk of numbers from which it allocates without refering
> back to the central office. The only difference being how large
> those chunks are and how many offices there are. Do you consider
> Oracle sequence numbers to be natural for this reason?
>
I don't see how the next set of cached sequence numbers is the same as
the pre-assigned region number used in SSNs. Sequence numbers still
increase in value. An SSN of 555-xx-xxxx can be assigned, followed by an
SSN of 123-xx-xxxx. SSNs bounce all over the place. Sequence numbers don't.
>
>>I know some people who know some people who will, for a small fee,
>>produce a "legal" Social Security Card with a unique SSN.
>
>
> Do all the people who do this coordinate with each other and the
> government? If not, how can they know that the number is unique?
> I bet you get a card which they hope is unique.
>
I'd be very surprised to find out they coordinate with the government.
They're not dummies ... they've been doing it for years ... and they
don't hand out duplicates. Don't know how they do it ... just know it
happens.
>
>
>>Don't know
>>where they hijack the numbers from, but it's enuf to fool the IRS. All
>>this tells me the SSN is not an arbitrary key.
>
>
> I know people who know people who can get a legal PK, which is good enough
> to fool a database into accepting it. Does that mean Oracle PKs assigned
> from a sequence number are not arbitrary keys?
>
And your point is? As stated above, Oracle sequence numbers do not
contain any smarts; but SSNs do.
> Xho
>
Received on Mon Nov 25 2002 - 19:12:55 CST
