The following is height of crypt-dream outline of becoming the richest man in
world, profoundest intellectual so far on earth, and how to bootstrap to it. It
is believable but provably useless for anything but belief test! It is long,
logical, absurd but fun imagination. I sincerely believe in it.
My generic blockchain world has consumers (or consumer
agents) and Miners, A pair of consumers called the giver Alice and getter Bob wish transmission of entity product through some miner
representative agent who interacts
with other miners and maintains a distributed ledger. It is assumed that there is allies that wish to ensure taxes
are paid and can do warrantful
decryption by force. There is also many evil easvesdropper Eve and evil modifiers Malloy
who might subvert arbitrary miners. For understanding, consider coins from
giver to getter.
Aalan
Unlike other languages, syntax is not a matter of choice but
existing inadequacies. Eventually it is designed to win compulsorily – it
provides the best of all others and is currently the only way to get these
1.
Library primitives to privacy preserving
blockchain
2.
System-level efficient multiple inheritance
3.
Non-deterministic programming
As well as ideas from other contemporary great efforts of
1.
Concurrency from Go
2.
No linguistic garbage collection from Rust
3.
First class scope from angular js
4.
Functional operations from shell
5.
First class function values from javascript
6.
Multiple simultaneous error returns from Go
And small but crucial syntactic sugars of
1.
() versus {}
2.
Arguments by category
3.
Meaningful compaction
Rest is overview of the implementation goals and why!
Doubt 1 – As stated,
there is NO solution!
Reply 1 – For NO one on earth but me. Think of practical
cases. Suppose a Canadian firm allowed undecodable encryptions. Indian police
would be rightfully upset as it would be also used by every criminal! So they
would demand forced decryption with proper warrants. Possible if the secrets of
the encryption field were given. Such a doing by Canadian company gives India
police enormous powers of forced decryption of all else, no Canadian need
later, never mind corrupt folks in Indian police later! So Canadian company
says no. So Indian police convinces GOI to banish products of this company!
Similar scenarios have happened. No police can allow
unbreakable encryption. Feds v. Apple was so. Enormous number of users can be
mis-decrypted by state or an enemy using one turncoat. China USA is big battle.
How can I do it, uniquely bypass the lacuna?! By having nearly as many
encryption fields as the number of some multiple of consumers! By changing
every field every year or month or week!
Doubt 2 – There are
plenty of smart encryption fellows!
Reply 2 – One must consider security and privacy, a tension throughout
human history. No privacy leads to big brother and tyrants of one kind; while
total privacy leads to crime, underage sexuality, cannibalism, destability and
tyrants of other kind. The USA solution is greatest approximation with strict
warrant-ful control on breach of privacy! One can never allow unbreakable
encryption. But every exposure of the field parameters leads to turncoats and
destruction of other people privacy.
The only solution to this conundrum is as many IBE encryption
fields as members or more. It is easy for me, impossible for others! And no one
on earth can do what I can!
Among other applications, I can build unlie-able trademarks,
untappable phone conversations, guaranteed secure mail, unusable if pirated
video etc.
Doubt 3 – There is no
way to bootstrap!
Reply 3 - Beyond some point, my empire will grow from money
raised internally from products sold. How does one get to self-growth point? Especially
if I have deep doubts about business acumen of self! The answer is venture
capitalists, who must have a convincing demo. So that is job1. But how?
If I gave away my root extraction, that is crown jewels! But
wait a minute! A toy implementation on 16 bit bignums can be made that does
only operations on 16 bits now, but can work for 1000 bit primitives directly!
This simulation is not done yet! What about square root
which will leak even from toy implementation! However for 16 bits, one can do
something impossible for 1000 bits – make a table of square roots! There are
64K numbers. So one can make them for 1-9 people in the world for half a
million element tables. Anyone can ask for square root of any number in family
semi-prime and it is just a table lookup!
In fact, I solve the problem of impossible bugs in my 1000
bit arithmetic! The toy requires simple 32 bit operations. In fact, difficulty
for evil in my world can be appreciated by 64 bit arithmetic!
Enormous development
One utterly remarkable upshot of this documentation is that
not only does it present very simple and convincing outline to blockchain that
is central to new-21, along with many other encryptions, it is all based on a
problem only I can solve on earth! What is more, demonstration programs can be
built that simulate this primitive and all essential user code can be open
source without empowering competitors in any way! Even this property can be
demonstrated! So let me expose nearly everything, also nothing!
Only secret, earth-wide, that I have is computation of
square roots of 1000 to 10,000 bit integers, modulo semi-prime. Computation of
squares is easy and requires very little hardware. Hence checking a square root
is cheap. But computing it is exponentially hard.. Simulation for a small
integer, 16 bit, is easy as all the squares can be enumerated. One can make all
things discussed here, except root calculations, open source. In turn, the
software can be built using copy-left. The square root extraction can be done
secretly inside protected hardware. The results can be checked.
Nearly half the numbers to a (semi-prime n : a/log n) are modulo
squares. For 1024 bit integers, 1013 bit numbers are. This means 1013 bit keys
are available!
TO USE ENCRYPTION THAT IS IMMUNE TO ADAPTIVE AND QUANTUM ATTACKS,
one uses a double encryption – outer-quantum-computer-proof-inner-my-IBE.
Knowledge-able will correctly guess the details. For all else
Encryption (Message-1gb, key-1024, Alice-private-data, Bob-public-data)
=> eMessage-1gb+10%, key-square-1024, more-3000
Decryption (eMessage-1gb+, sqrt-key-1024, Bob-private-data,
Alice-public-data) => message
To send encrypted to bob, encrypt message using
consumer-choice- key and send the encrypted with key material. Provided I
secretly send back the sqrt (after authentication) decryption happens. Bob does
not care about how I did, sends me asked square encrypted and I answer sqrt
encrypted.
The sqrt is required in this scheme! It can be cheaply
checked. But NOT extracted. The quantum computer enclosure is needed. Much
smaller key-material can replace it with OAEP+ RSA. However New-Hope-Simple is
fast enough.
Two semi-prime
protocol aa2
I have built my IBE around two semi-prime, one whose factors
I know and one I don’t! This allows messages undecode-able by me to be passed.
Imagine a nation where Alice resides. It may require key escrow deposit before
the semi-prime can be used. The message is still secret to me. Associated with
any allowed semi-prime is a signature which must be attached by any sender. False
attachment is Warrant-evade crime! Police permit properly escrowed semi-prime,
fail only with warrants, detected reliably and receiver punished without
further cause. False entrapment fixable. My semi-prime only eliminates
man-in-the-middle by providing identity proof!
Let us call escrow deposit and signature process
registration. Sender needs to register in the destination, while the
destination must be registered in own nation. The sender might be required to
send copy to self of all external mail. This allows warrant in send and receive
jurisdictions. Eve and Malloy are useless.
The aa2 allows the destruction of any address without
exposing any previous mails! It is done causelessly to automatically destroy
any leaks every year or faster. Any address may be destroyed at once for cause.
In one application, locks are changed whenever an employee e leaves in all
locations where e might have a key!
Signatures
Trademarks are essential but enter extra costs. Clearly,
provider is interested in enforcing trademarks while consumer wants the quality
but not have the extra cost. Legal method is attaching symbols and distinctive
containers to products. Breach is done by false copies of marks to malicious
smuggling.
It may be to goods, even money. Suppose every nation had a
semi-prime. All the notes would have a field from which a uni1que serial number
was extracted. It would be followed by square root. A local check can be made
that the serial number is genuine and that the square root squared reproduces
the serial. This is a fool-proof local check to devastate Pakistan! Same works
for Gucci making bags. Trademarks in general. What can be better than
eliminating the criminal PTO of USA!
Another is automatic fee payment to continue encryption
address beyond current year.
In one application, banknotes lose value fully every year –
they must be exchanged in banks every year after a record is made. The record
itself may be encrypted to allow signatured recorded permitted access only. One
can reap demonetization benefits without one!
Apart from becoming the richest, a number of warrant-ful
applications can only be possible in my encryption! USA judges complain
incessantly about use/abuse of privacy by suspect’s lawyers! Civilization
forward is the continuous battle between privacy and security. Only I have technological
not legal ideas – one you can breach, others you cannot!
