[XPOST] A unique number for every "person" - can it be done?
From: TGOS (tgos_at_invalid.invalid)
Date: 02/27/05
- Next message: Willem: "Re: c code, works on linux, crashes on windows"
- Previous message: Jerry Gaspard: "c code, works on linux, crashes on windows"
- Next in thread: Virgil: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply: Virgil: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply: Ioannis: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) Jacob Sheehy: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) spinoza1111_at_yahoo.com: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) Tim Smith: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) john_ramsden_at_sagitta-ps.com: "Re: A unique number for every "person" - can it be done?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sun, 27 Feb 2005 20:10:57 +0100
Hello,
First please excuse the xpost, but the topic of this post does not fit
into any single NG and I didn't want to miss the right people by posting
to the wrong NG. Maybe we can find out where it really belongs to in the
discussion process and move the thread to the right group.
For halve a year now I'm thinking about creating an algorithm, this
sounds like math, but without some form of hashing used in cryptography
it will most likely get nowhere, and the topic is program related. You
see, it's very hard to categorize the topic.
The problem can be summarized in one sentence:
Calculate a number for every human being, company and organization on
earth, that is guaranteed to be unique till the end of time.
The rules in detail:
1) Every "person", real person or corporate body (like company or
organization), needs a number. Every means every one world wide.
2) Then number must be unique. Neither may there be two "persons" having
the same number at the same time, nor may there be two "persons" having
the same number at a different time. Even two billion years after my
death, nobody may have my number. For cryptographers, if it were a hash,
it must be guaranteed to be collision free, whereby the possibility of
collisions is allowed if the likeliness is close to zero (one in a
billion of cases) and if you can handle it (e.g. in case a collision
should one day be noticed, it is solved by ...) to still give both
involved parties a unique number.
3) The numbers may be earth bound. The system does not have to scale
beyond earth, in case mankind can one day life on a planet other than
earth. But they may not be country bound. Countries come and go, they
join and separate and dissolve again.
4) The system must work without a central registry. Establishing a
registry and saying, every "person" gets a number when registering, that
is one higher than the last number, would work, but a central database
like system is required.
5) Once a person has a number, he can always recover it, in case it got
lost or forgotten, so the number creation is reproducible. An algorithm
working with the current time in ms since 1970 for example will
certainly add some randomness to the number and such making collisions
unlikely (in combination with other data), but you could not recover the
number again. This also means that all data for calculation must be
constant and can't change over the time.
6) The length of the number is not limited, but the shorter it is, the
better. However it may grow over the time; in thousand years having a
slightly bigger number would be acceptable.
7) You may use whatever data a "person" has available, but you may not
private data. If you tell someone your number, he should not be able to
gain any non-public knowledge about you through the number. If private
data is used, it must be one-way-hashed, so the original data is not
recoverable. On the other hand, the number should clearly show if it is
from an organization, from a company or from a real person.
8) Everyone must be able to calculate his/her number, without doing a
lot of research and it must be data everyone knows for sure (the exact
time of birth in ms does not qualify). On the other hand, people may be
forced to use a computer for doing that and have access to the Internet.
E.g. if you want to know the coordinates of their town of birth (using
whatever coordinate system you choice), this could be acceptable, as
there could be a search engine telling people the coordinates.
Be creative, try to find data useful for the purpose. Things you may
want to use:
- Date of birth / Year of foundation
- Place of birth (consider not always known, names can change over time,
better go for coordinates)
- Name (First, Last / Name of company/organization)
- Name of parents (consider orphans / companies)
- Blood Type (consider companies have no blood type)
- Gender (consider companies)
- Eye color (should be constant, consider companies)
and so on.
Just because some people may not know something or something may not
apply to certain people doesn't mean you can not use it. But explain
what to do in the case it does not exist or is unknown.
I came up with plenty of ideas, but they were either too complicated or
creating collisions was too likely, that are not easily resolved.
Simply writing down some data and hashing it creates a decent number,
but how long will this be collision free? How big must the hash be to be
secure for thousands of years and 6 billions of people and millions of
companies/organizations?
-- TGOS
- Next message: Willem: "Re: c code, works on linux, crashes on windows"
- Previous message: Jerry Gaspard: "c code, works on linux, crashes on windows"
- Next in thread: Virgil: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply: Virgil: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply: Ioannis: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) Jacob Sheehy: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) spinoza1111_at_yahoo.com: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) Tim Smith: "Re: [XPOST] A unique number for every "person" - can it be done?"
- Reply:(deleted message) john_ramsden_at_sagitta-ps.com: "Re: A unique number for every "person" - can it be done?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|