Edit: It has been decided I was going off on an unnecessary tangent here.

If we want the MCU to be able to turn the power supplies for the FPGAs on and off, it should be bus powered. If USB is being used, we are talking  5 volts, 500mA (100mA without permission).

If the FPGAs have a small enough 2.5V current draws, no extra part is needed. According to page 7 of the datasheet "...Spartan-6 devices do not have a required power-on sequence." (Table 6 note 2).

I couldn't really find anything saying you can power up V_ccaux and V_cco with V_ccint unpowered more explicitly.

I am a Free (as in freedom) and open source software fanatic. I will be using FPGAs if all the ASIC modules are literally epoxied black boxes. What are you trying to hide? Will I be able to write my own miner to make use of the chips?

Note: I don't like FPGAs that much either due to the lack of documentation.

I looked at the scripting opcodes and my eyes kind of glazed over. I am currently assuming an automated loan is not possible.

The reason I am potentially concerned about a double-spend attack is what happens if somebody uses these bonds to buy goods or services? On the appointed date, they would "cash out," leaving the bond-holder with nothing.

Thinking about it a little more, I decided a customized client is probably needed to use "weird" scripts anyway. The default Bitcoin client probably mainly sticks to "standard" transactions, which are also scripts. Again, I have not carefully checked if my impressions are true or not. The default client will relay "weird" transactions with no problem as far as I can tell.


A "Fork" in the traditional sense won't work: it is all or nothing. The clients not recognizing the fork will studiously ignore it.

That said, there is reason to believe that if there is a compelling reason, all (or at least the majority)  of the participants can agree to a protocol change. If the change is not urgent, it can take place with decades of lead time. However, in order to get the (vast) majority to agree, the change should be non-controversial. I think your bond proposal would be too controversial to gain widespread acceptance for a protocol change. A non-controversial change would be something like: changing the hash function if collisions are demonstrated with SHA-256. Or, possibly, increasing the sub-divisibility beyond 8 decimal places if too many bitcoins are lost over the years.

Edit: I already explained why I think these non-controversial changes won't happen: I think bitcoin fail (and be replaced with something better) before the specific design choices become too inconvenient. (That is to say, I don't think Bitcoin will fail based on its own technical merits.)

Both links and achors are described in Section 12 of the HTML 4.01 specification. Essentially, instead of using an 'HTTP' url, you use a 'bitcoin' url. Some browsers/Forum software get confused when you present it with a url type is does not recognize. For example: Floodgap Systems' official gopher server.
edit: Forum software assumed I forgot to include "HTTP". The Http gateway is here.

Umm, what if there is no original author?

Monkey Business: Can A Monkey License Its Copyrights To A News Agency?

Photographer David Slater Claims That Because He Thought Monkeys Might Take Pictures, Copyright Is His

HTTPS includes authentication, because without it man-in-the-middle attacks are trivial.

Encryption isn't magic security pixie dust. Without authentication, the attacker can simply set up two encrypted sessions: one to the user, and one to the server being impersonated.

I think it is still prudent to investigate is it is possible using scripting. If so, it would be a double-spend attack with a potentially years-long lag time.

I disagree completely. I think starting a new block chain will almost always be better than forking the existing one. Forking the chain undermines the currency; thus eroding confidence and value.

A rival blockchain is not something to be feared. There already exists the testnet and namecoin blockchains. If somebody starts up a "Bitcoin with bonds" blockchain, some people may move over to it. If you are correct that bonds are needed for long-term stability, it may even become the new "standard." It does not matter if people have money invested in the first bitcoin: If the transition happens slowly enough, the people who want to sell bitcoin can sell it in exchange for "bitcoin with bonds".

The reason I am so cavalier about the failure of bitcoin is that I expect any crypto-currency, no matter how good, to fail in the medium-long term. Computers are just not secure enough to handle a crypto-currency in my opinion. Modern computers systems are complex and unproven. Subtle errors missed by this oversight can likely be leveraged into a devastating attack on any crypto-currency; given enough time an resources. For the record, I expect the computer industry to finally mature into reliable, proven correct systems after about 6 generations (150 years) or so. As precedent, I point to the Agricultural revolution, the invention of the wheel and the printing press. All took centuries to fundamentally change society.

This won't happen as described. If bicoin is not replaced with something more latency-tolerant, the Mars Colony would start their own blockchain with their own version number. The reason a split of the existing chain would not be attempted is that the longest blockchain wins. That implies the Mars fork would never be accepted.

Of course, exchanges supporting latency-tolerant transactions would be set up. A bigger problem, is what prevents the larger network from manipulating the smaller one?

The problem is that currently transaction blocks create coins at a predictable rate. The difficulty is tuned such that transaction blocks happen at a predictable rate. If you throw new types of transaction blocks into the mix, either you will affect the rate of coin creation or the rate at which transactions are processed. You would break the automatic tuning mechanism.

The only way your proposal can work is if you come up with a new type of transaction that does not directly affect coin creation. In any case, it is likely you will still need to start your own block chain. This may be a case where Bitcoin should focus on doing a few things well.

Guaranteeing bonds does not appear to be something bitcoin is designed to do. Bitcoin works by keeping a public transaction log. Participants trust that the majority of participants don't want to tamper with the public transaction log. The individual participants still need to trust each other when exchanging goods and services. The payments can't be reversed, but the seller can fail to send the goods, for example.

It complicates the protocol. The system guards against fraud by using a lottery to determine who processes a block. The difficulty of producing a block is automatically adjusted such that one block is processed approximately every 10 minutes. Adding new types of blocks without the same constraints will hurt the security guarantees of the bitcoin network. You can likely mitigate this by only creating new types of transactions instead of new types of blocks. You would still have to convince everybody to agree to support them.

You may want to investigate if contracts and scripts can do what you want. I think you would be able to force a coin to be unspendable for a length of time, but I don't think you would be able to loan a coin.

The problem is that your browser trusts a lot of signing authorities. Many will let you a buy a certificate that in turn lets you sign your own certificates. That means if your attacker has money, they can easily impersonate any website they want by signing a certificate in their name. Source: Certificate Patrol Website

Edit: You also need to be a company with over 5 million in assets, 5 Million in insurance and has clear audited policies on how you use the certificates. For the GeoTrust root anyway.

I only started caring about ubiquitous encryption over the past 3 months or so. I am dismayed how prudent sending everything is clear-text really appears to be: HTTPS can easily give a false sense of security. I have not had time to tease out a list of best practices. My own website does not support HTTPS at the moment.

Edit: I have an interest in the gopher protocol. I may investigate using IPSec for encryption or authentication. I currently care more about authentication. Authentication can never be automated and easy: That is the situation we have with Certificate authorities. Users need to exchange information out-of-band: over the phone, in person, or by letter mail.

I think both methods suggested open the door for fraud. Also, miners processing the transaction are chosen by lottery. If miner not recognizing bonds processes the maturing block, the payouts will be incorrect. Because of the possibility of block chain splits, such an error can't be corrected for about 120 block, were another non-bond recognizing  miner may process the transaction.

On second though, this will split the block chain. The "standard" miners have no way of recognising a "non-recoverable" address: in fact, there is no way for prove that such addresses will never be generated by chance. Any miners and not knowing about the "bonds" will reject any block containing the bond reimbursement transaction that appears to create money out of thin air. The way transactions work; all transactions are tracked back to the coin creation. Coin creation is governed by a set of rules all nodes must agree to in order to recognise valid coins.

I am confident that bonds can not be introduced directly into the blockchain without creating an new block chain. If you are confident there is demand for bonds, you can try to convince people to join bitcoin2. I will be staying with the original bitcoin for at least 2 years (to see what happens when the block-reward drops).

SSL/TLS Vulnerability Widely Unpatched.

Add to that that most browsers treat Clear-text HTTP as more secure than self-signed certificates. That means most websites either use no authentication, or rely on a Certificate authority that may be malicious.

In an ideal world, organizations would post the "fingerprint" of their self-signed certificate on their premises. If you are using e-billing, the billing company would send you a copy of the certificate or fingerprint by mail.

In any case, your ISP knows which server you are connecting to when you use HTTPS, but not necessarily which pages you visit.

I am assuming the -1C temperature is really false.

My guess is the power supply. If the ground on your "free hanging" power supply (at the card) goes above 0.7V (relative to the rest of the computer) due to high current draw, you can likely cause all kinds of problems.

You should have the ground on the two power supplies connected together. In another thread, the method suggested for starting a pair of power supplies is to tie the green wires together as well. If it was me, I would want the Power good signals tied together with an AND gate, but that requires some research (is 'Power good' a digital signal?) and circuitry.

I was sort of hoping that if the motherboard is compelling enough, people would build ASIC boards that are pin-compatible

Edit: the FPGA prototypes would be a "proof of concept".

To be honest, I am not sure what you are proposing. If you want to "lock-in" a difficulty 6 months ahead of time, that will likely screw up block generation. On average, one block is produced every 10 minutes or so. If the next block was already produced 6 months ago, we would have to artificially raise the difficulty level to avoid generating coins. Worse, block generation is what processes transactions. Artificially raising the difficulty would slow transaction processing: you can't just "add" transactions to the 6 month old block now maturing.

As becoin said, you can make your own bonds: Generate a savings wallet using Bitcoin Off-The-Grid (BOTG), and store it in a safe place; preferably two safe places. Send money to it. After the "bond period" is up, retrieve the savings wallet and import the private keys. You can then spend the coins again.

I think one of the major benefits of a backplane is density. You can pack the modules in closer together without them getting scattered all over the place every time you bump the table.

I have little industry experience as well

Edit: By electrical, you mean circuits over 50V/20A?

They both set cookies. They can personally identify you all they want; though Google seems to promise you need to opt-in first. You (and possibly your users) need to review the relevant privacy policies.

- http://www.facebook.com/privacy/explanation.php

- http://www.facebook.com/policy.php

My reading of that is that you are not anonymous until 6 months have passed: by which time you may have visited the page in question again, resetting the timer.

- http://www.google.com/intl/en/privacy/ads/

- http://www.google.com/intl/en/privacy/ads/privacy-policy.html

They go on to say that they don't use your personally identifying information without your consent, but I fail to see how information about your interests is not personally identifying.

If I am more knowledgeable about these things than the average person, it it probably because I have a bad habit of reading fine-print. That is one of the major reasons I am interested in bitcoin: on the surface, less onerous "Terms of use" to follow. Though, some of the onerous terms in Credit Card agreements seem to stem from anti-money laundering and terrorist financing laws.

Yes, for the foreseeable future, it will always be possible for the value of Bitcoin to drop back down to 0.

The only thing that gives bitcoin value is the ability to buy things with it. That means: we need merchants to invest in infrastructure for accepting bitcoin. Few merchants will invest in bitcoin if they feel it has little support.

If bitcoin gains widespread support as a payment alternative; mining and speculation will be very profitable. However, if everybody hordes without spending any Bitcoin; the value will stagnate or even drop.