1997-06-02 - Re: Creating a unique ID number for a dollar

Header Data

From: “William H. Geiger III” <whgiii@amaranth.com>
To: jonathon <grafolog@netcom.com>
Message Hash: 34f552e0a9f6df60e3c546a2b728adef45d12e885698af0bdcb2fa6715ca44f0
Message ID: <199706020349.WAA09406@mailhub.amaranth.com>
Reply To: <Pine.SUN.3.95.970602014607.25651C-100000@netcom2>
UTC Datetime: 1997-06-02 04:12:27 UTC
Raw Date: Mon, 2 Jun 1997 12:12:27 +0800

Raw message

From: "William H. Geiger III" <whgiii@amaranth.com>
Date: Mon, 2 Jun 1997 12:12:27 +0800
To: jonathon <grafolog@netcom.com>
Subject: Re: Creating a unique ID number for a dollar
In-Reply-To: <Pine.SUN.3.95.970602014607.25651C-100000@netcom2>
Message-ID: <199706020349.WAA09406@mailhub.amaranth.com>
MIME-Version: 1.0
Content-Type: text/plain


In <Pine.SUN.3.95.970602014607.25651C-100000@netcom2>, on 06/02/97 
   at 02:24 AM, jonathon <grafolog@netcom.com> said:

>On Sun, 1 Jun 1997, Tim May wrote:

>> At 5:25 PM -0700 6/1/97, William H. Geiger III wrote:

>> >stored on individules a SS number is quite convienat as everyone has a
>> >unique one. Most employee, payroll, medical, insurance, credit, databases

>	Somebody here is forgetting that the Social Security
>	Adminstration, back in the mid-eighties claimed that at least
>	10 percent of the numbers in use, were improperly issued.

<sigh> This is really no suprise considering who is issueing the numbers.

>	The same number was issued to _two or *more*_ people.  The worst
>	case was a number that several thousand people used, thinking it 
>	was issued to them, exclusively, when it was in fact never issued.

This is rarely a problem from a data managemant posistion as SS # are key
fields. Any attempts at adding duplicates produces errors which then have
to be resolved usally manually by human intervention. Rarely are SS #'s
used exclusivly but in combination with other data (name & DOB is usaully

>	A further complication is that the same individual could have been
>	issued two or _more_ different numbers, either by design, or
>	accident.

I doubt that 2 or more SS# would be issued delibratly. The only reason for
multiple SS#'s would either be screw-ups by SSA or by design of the person
applying for the SS#. I imagine that new SS# may be issued in some special
cases such as witeness relocation, perhaps after adoption, but then the
old SS# is not being used so it is not really an issue.

>> >unique qualities that make it perfect for this use: 1 every person has a
>> >unique #, and 2 it never changes. This can not be said for any other
>	Both premises are false, and the SSA has said so on several
>	different occasions.

For all practical purposes it is. with the exception of screw-ups by SSA
ones SS # is unique and with the exception of a few rare cases mentioned
above it never changes. compare this to other identifiers and it is
obviously the most convienient id # available. Atleast 50% of the
population has 1 name change durring the cource of their lifes, Addresses
change numerious times during an average americans lifetime, and DOB's are
not unique enough to be used.

While problems do exsist with using SS#'s as id they are quite small when
compaired to using other less stable data to generate id #'s.

>> without entry of any allegedly random numbers, and without any hashing
>> of personal data. It's not necessarily a real short number, certainly
>> not as short as an SS number. 
>	One proposal I'm familiar with was:
>	date of birth << year month day >>
>	time of birth << hours, minutes, seconds >>
>	longitude of birth  << degrees, minutes, seconds >> 
>	lattitude of birth  << degrees, minutes, seconds >>
>	sex		     << one letter >>
>	mother's initials    << first, middle, last >>
>	father's initials   <<  first, middle, last >>

>	so you'd end up with something like

>	19970601185500-0300000.00-300000.00mxyzwvz

>	<< A number which would be issued to a male born today 
>	somewhere slightly north of Port Shepstone, and slightly west
>	of Pietermaritzburg, RSA. >> 

>	However, there are several problems with it, the two most notable
>	being the lack of accurate birth times, and that most people have
>	a very hard time remembering 42 digit numbers.

>	I don't know how solvable those, and other not so apparant
>	problems are, but I suspect that it has been intensively
>	studied by more than a few governments and organizations, since
>	it was first proposed, fifty something years ago. 

Really much to complex to be of use not to mention the lack of reliable
data to form the id #.

The use of DOB + Geographic Identifier + Unique Code would work quite

19970601 - DOB.
0123     - Sample Geographic Identifier (say NY City).
0142     - Unique Code added to handle collisions of the above two.

I beleive that this is very simmilar to what the SSA uses though I beleive
that they only encode the year of birth when calculating SS #'s. Using Hex
rather than decimal for encoding would help greatly in redicing the number
of digits required. I would imagine that the SSA will have to go to a Hex
or complete Alphanumeric codings system as the population increases.

- -- 
- ---------------------------------------------------------------
William H. Geiger III  http://www.amaranth.com/~whgiii
Geiger Consulting    Cooking With Warp 4.0

Author of E-Secure - PGP Front End for MR/2 Ice
PGP & MR/2 the only way for secure e-mail.
OS/2 PGP 2.6.3a at: http://www.amaranth.com/~whgiii/pgpmr2.html                        
- ---------------------------------------------------------------

Version: 2.6.3a
Charset: cp850
Comment: Registered_User_E-Secure_v1.1b1_ES000000