Path: utzoo!genat!perle!dave
From: dave@perle.UUCP (Dave LeReverend)
Newsgroups: can.general
Subject: Re: SIN Number
Keywords: Checksum Procedure
Message-ID: <420@perle.UUCP>
Date: 20 Dec 88 17:22:21 GMT
References: <302@idacom.UUCP> <723@apss.apss.ab.ca> <1902@pembina.UUCP>
Reply-To: dave@perle.UUCP (David LeReverend)
Distribution: can
Organization: Perle Systems Limited  Scarborough, Ontario, Canada
Lines: 95


In article <1902@pembina.UUCP> cdshaw@pembina.UUCP (Chris Shaw) writes:

>[Quotes from Herbert Presley and Chris' responses to them deleted.]

>SIN's are self-checking[.]

They certainly are.  The last (ninth) digit is a "check sum", and it can be
determined using the 9 simple steps shown below.  I tried this out on my own
SIN, but would never be so naive as to post my own SIN on the net :-).  As an
example, I used a 9-digit SIN with the following form:

        123 456 78C

where "C" is the check sum digit.

Steps:

1) Use the even digits to form a four-digit number.

   In this example, it's: 2468

2) Double this four-digit number.

   2468 x 2 = 4936

3) Sum the digits of the number determined in step 2.

   4+9+3+6 = 22

4) Use the odd digits (but not "C") to form another four-digit number.

   1357

5) Sum the digits of the number determined in step 4.

   1+3+5+7 = 16

6) Add the results from steps 3 and 5.

   22 + 16 = 38

7) From the result of step 6, determine the next highest multiple of 10.

   40 

8) Find the difference between the results from steps 6 and 7.

   40 - 38 = 2

9) The result of step 8 becomes the last (ninth) digit in the SIN.

   The complete SIN becomes: 123 456 782


This algorithm is useful if one wishes to verify the first 8 digits of a 
given SIN.  I have not looked into the "strength" of this type of a check
sum.  If anyone out there has the knowledge and inclination to do this, I'd
be interested in the result.

Of course, with this algorithm it is possible to chose any 8-digits, calculate
the proper check sum, and produce a "valid" 9-digit SIN.  

Perhaps portions of this discussion should be cross-posted to "comp.risks".

> [So,] to give a fake one, switch any pair of digits
>3 and 5, 4 and 6, 5 and 7 of your SIN. You can switch more than one pair.
>
>123 456 789 (3&5) goes to
>125 436 789 (4&6) goes to
>125 634 789 (5&7) goes to
>125 674 389 ...

Chris is saying how to re-arrange the digits of your SIN so that it will still 
have the same check sum.  I haven't given any thought to this.  If anyone 
cares to, I'd like to see if Chris' assertion can be proven based on the
above nine steps. 


David LeReverend

-------------------------------------------------------------------------------

My grade 12 math teacher saw this algorithm in a government-published book
of math problems, so I assume that there are no rules against my telling 
other people about it.

-------------------------------------------------------------------------------

"I believe the aliens have to take a physical form on this planet.
 So why not one with 13 channels?"

			From Joe Jackson's "TV Age"

-------------------------------------------------------------------------------