PostScript, on which PDF is based, is Turing-complete. I think PDF was specifically designed not to have the capability.

However, once the PDF format became popular (because it acted like paper), Adobe decided to screw it up. At about version 6 of the Acrobat reader (not sure which PDF version), JavaScript support was added, adding back turing completeness. Adobe also added DRM features, making PDF files act decidedly unlike paper by betraying the user (refusing to open, print, or edit).

Even with a turing-complete language, I would not trust the quality of any random number generator running inside the printer.

I find the idea impressive, but I don't think it will actually work. As far as I can tell, the proposal is technically sound, but abstracts away real-world concerns like bandwidth and CPU time. It is like you took Computing Science to its logical conclusion.

I think Scalabilty has the potential to be a real problem for bitcoin. In the example given, a "VISA" level of transaction processing (2000 transactions per second) is expected to have blocks over 1GB in size. I expect this will force many miners on home or even business Internet connections off the network (Colocation with fiber access would be needed instead).

Even with the incredible compression you expect to achieve, I doubt a global turing machine will need less than 2000 transactions per second. This implies a huge amount of bandwidth usage and local storage demands (for saving every fork).

I also don't think confidential processing (such as generating vanity bitcoin addresses) will be possible without Fully Homomorphic encryption. Found out about that concept reading this very forum

Most certainly not. Patents grant a time-limited monopoly on the use of a novel, publicly dislcosed invention. Even if the concept of using logos for branding was once patentable, it no longer is, due to ample examples of "prior art."

Trademarks on the other hand, grant an indefinite monopoly on the use of distinctive marks. They are designed to reduce confusion in the marketplace, such that you know whom takes responisibility for manufacturing a product.

Copyright may also apply to the logo, but it is my understanding that it is under a relatively liberal public copyright license.

Point of order: "Windows" is a trademark, not a patent. Microsoft may use software patents to force you to "agree" to the EULA just by using the machine, or force you to pay licensing fees even if you are using a competing OS. However, none of those Patents are called "Windows." They are numbered and have titles like "No. 6,339,780: placing a loading status in the content viewing area of a browser" Source. I would have let it go, but you quoted yourself.

From an "any publicity is good publicity" perspective, percieved defemation may not matter. That bitcoin is so innovative that is must be illegal may be a good thing.

As a socialist I can give you a little bit of insight. I assume by "memo", you mean the fall of the Soviet Union. It is entirely possible that Capitalism is just as unstable, but the USSR just collapsed first.

Actual communists go even further. They see the central planning employed by the Soviet Union as "state capitalism," mirroring the central planning often seen in large corporations. In that light, the fall of the Soviet Union can be seen as a failure of the corporate structure endemic to capitalism. An actual communist society would rely on local democratic decison making. When central decisions are needed, representatives would be selected. Much like Holliday was advocating actually

On the topic, for small transactions, bitcoin is still expensive (due to the banking system you must use).  It is kind of a "chicken&egg" problem.

My bitcoin cheque idea resolves some of those problems, but is no magic bullet.

Essentially, the biggest risk mitigation measure is their single-use nature. While a few may end up being used as a "savings wallet", the majority should be redeemed within weeks of being funded. That implies that a would-be cheat never quite knows the optimum time to ruin their reputation for a quick buck. It also makes counterfeiting (economies of scale will hurt small runs) less profitable since it is not known ahead of time how much funding will go to a specific address.

I also plan to leverage existing trust neworks. When printing my bitcoin cheques, the printer will be putting their reputation on the line. They will have a real business outside of bitcoin that they would not associated with bad press. The reputable printers will do everything they can to prove to themselves that they, their employees, or equipment have no way of knowing the private key after printing. This may involve banning cameras during printing, keeping employees with a photographic memory away from the floor during printing, using a hardware random number generator, wiping the local storage on any devices connected to the network during printing (including the press). The point is the printer has only their reputation to prove they are trustworthy. Proving to yourself that you can't know the key simply removes any tempation to tarnish your reputation.

Of course, as I mentioned earlier, the printer has to be able to prove whether or not they printed a given document. This may involve using things like microprinting with access to a high-resolutuion press. My printer does not have that resolution, but it may be able to print that Eurion Constallation even though modern printers may stop printing. I am also considering making each printing its own series with its own silk-screen. The silk-screen would be deliberately damaged (eg: with holes) and stored for analysis if a dispute ever arises. Large print-runs may want to make use of traceable paper as well.

I don't get it.

Do you plan of giving out bitcoins, then pressure the recipient to transfer those same bitcoins to the charity illustrated on the bill?

In any case, you appear to be breaking the paper currency metaphor.

I have my own laptop I can leave running for days. Since my timeframe is months, I can generate a much more complex address. The idea of asking for help is I might be able to go for a much longer one, but the exponential rise in difficulty will kill you. DeathAndTaxes' suggestion has merit to. I am considering writing a Postscript file that will do most generation (including ECC math) for me (would likely be run using GhostScript, rather than a fancy printer). The only thing that would change between documents would be the randomly generated 32 byte number the key is derived from.

Currently, my computer situation is a bit of a mess. It is one of the reasons I don't actually have any bitcoins yet.

An yes, the tamper-evident seal was going to be the scratch type. The problem with generic holograms is that a dishonest person can go and buy the same generic holograms to stick on their forgeries. Printing the name of the printer on the hologram is better, but I suspect few people will check such details.

Exactly. Don't want to got into more detail until mock-ups are ready. Don't want mock-ups until the vanity address is ready. I am also thinking of using the Eurion constellation to discourage printing with proprietary drivers or hardware (which can cheat on you).

I was planning on the public private key only for "proofs". I want a way to alert people that the test documents are *not* secure.

The actual cheques would use a tamper-evident seal like bit-bills, covered with a silk-screen bearing the name of the printer and possibly series/batch-specific marks.

Edit: I guess I could go for a less ambitious vanity address like: 1*PubLic*.
 

I have decided I am interested in producing bitcoin cheques over the next few months. They would be single-use, so durability is not too much of a concern. Security of the bicoin cheques would rely on two things:

• The printer can prove that they printed a given document.
• The printer can prove (to themselves at least) that they or their hardware has no way of knowing the private key once the document is complete.

I decided that holograms would just make the cheques more expensive to produce without really enhancing security.

One thing I want to do for printing proofs is generate a bicoin address of the form: 1*pubLic*private*key*. Does anybody want to help with that? The Private key would be public, as implied.

While I like the Shatoshi challenge/response, there has to be a way to discretely announce you are not interested in buying or selling. Maybe immediately changing the subject such as: "What do you think of the weather?" or "How about that local sports team?" (Inserting actual sports team name optional.)

EBay sellers are not allowed to accept bitcoin.

Scott Charney, Microsoft Vice President of Trustworthy Computing has the opposite vision. The secure Internet of the future will require Trusted (Intel) hardware and (Microsoft) software for doing such things as voting or e-mailing aunt Tille.

Since about 1996, the powers that be have been trying to suppress the general-purpose computer. That was when the Internet+MP3 files+Napster were revolutionizing music distribution. World governments got together and wrote updated Copyright and Perfomance and Phonograms treaties. More recently, world leaders aggreed to the Anti-Counterfieting Trade Agreement that was negotiated in secret.

The point is the general purpose computer is being phased out. Windows Vista was a major step in that direction with the Protected Media Path. When Vista was first released, no video drivers worked properly because manufacturerers were required by contract to make the drivers hard to debug (to prevent reverse-engineering by the end-user). Windows 8 will continue the phase-out by requiring hardware support for Secure Boot for logo certification. Whether the big computer companies lock out GNU/Linux remains to be seen.

The point is that after 15 years of pressure, your computer now works against you if it suspects you are guilty of copyright infingement. Do you really think world governments will hesitate to try to have computers computing consoles automatically block bitcoin by 2026?

Sometimes a summary can help. I stopped reading at the Python interpreter written in JavaScript that took gedit about 60 seconds 4 minutes of CPU time to prettify. (Edit: the module 'msg' starts at line 22588) Edit2: It appears to use the MD5 hash function which is no longer considered cryptographically secure.

If you mean encoding short, human readable messages in the block-chain; I don't think that is what the OP was referring to. You can be a lot more subtle while proving you have access to a certain text with the block-chain. Simply take the SHA-2 hash, convert it into a public/private address pair, send a token amount to that address, then (optionally) spend the coins back to yourself.

Even back-dated about 6 hours. Nice.

Edit: Nevermind. The forum is using UTC for the time-stamp on posts.
Edit2: See, I was really first:
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

When one sign a text with private key, he can prove, that he had a text
If one will add a timestamp from some block
one can prove that he have that text before any other.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQFN6rhFPPYMXhz9onsRAk6lAJ9KROC1FpJMHAqABzhe4uHKIWttsgCdFEWj
PN8OQlvYN8uCM7Qq3COEfNo=
=k+Qs
-----END PGP SIGNATURE-----

To go back any further, I would have had to back-date my signing key as well. Then there are archaeological considerations like: should I use the Debian "Weak keys" just for extra detail?

This is a valid concern, even for the example passphrases given.

If the passphrase has ever been published on the Internet, it may cost a suprisingly small amount to determine if any substrings on the the Internet (of limited complexity) correspond to a public address in the block-chain when hashed. If you use a dictionary word, you have no chance: it does not take any special hardware to check.

You only need 3 things:

• The ability to buy CPU time.
  
• The ability to break the problem into small parts.
  
• Access to a copy of the Internet. (The block chain is easy to get a copy of.)
  

Now the SHA-2 function is likely expenive enough, as is the address conversion, that simple queries will likely cost more than $100. However, I would be surprised it the cost rose more than an order of magnitude larger (Read: $10,000 budget, not including coding time).

I feel I may have broken the analogy by trying to pin it down on specific bits. The Block hash is freely copied. It only proves that a specific block has not been changed by a malicious party.

I contend the ability to manipulate coins represented by the block-chain is precious. Therefore the private address corresponding to a funded public address would be precious. Such 'bits' have no value without the block-chain, which is what makes them "precious". Also, not all private addresses are precious: those corresponding to empty addresses are worthless.

I suppose (most) the alternate block-chains would be considered semi-precious.

I like the fact "precious bits" seem to fit nicely into gaps in the Fintrac Guidelines. Until now, I was assuming that a "Money Service Business" would be the closest analog. (Though I fail to see how "alternative money remittance systems, such as Hawala, Hundi, Chitti, etc." can be reasonably included.) Precious Bits neatly fall outside the scope of "Dealers in precious metals and stones", at least until the law/regulations are changed to include "bits" (with the clarification that copyrighted works are not considered precious because they can easily be copied).

Edit: Which bits, specifically, are precious? Only the block-chain (and some novelty addresses) are computationally intensive. The private key (for a funded address) is what lets you move/create/destroy/merge/split funds precious bits.