I would also point out Bitcoin with a lite client (or ewallet service) doesn't need to download the blockchain either.

Not having to download the blockchain is OPTIONAL in Bitcoin depending on how much trust you need.

I doubt any major merchant wants to run their wallet without access to the raw blockchain though.

How many coins are we talking about?

There is 1 private per address.  A public address can be trivially recreated from the private key so the only thing which needs to be saved is a pricate key.  The wallet file of the default client maintains all used public/private keypairs as well as a keypool of the "next" 100 private keys to be used. 

As long as you haven't used more than 100 new private keys since your backup no funds will be lost.

Generally speaking you use 1 new private key every time you:
a) SEND coins (receiving coins is "free").
b) click the NEW ADDRESS button.

You should periodically update your backup to ensure the backup contains all private keys in use by the wallet.

Probably not.  GPU mining has killed CPU mining and slowly FPGA (dedicated hardware miners) is killing GPU mining although it likely will take a year or two.

As stated above 1% is well inside the natural variance in a 2 hour test.   It is even within normal variance in a 12 hours (~2GH/s) test.  Run your test and control for 72 hours+ and you might have something.  I mean you could have not fart, "tested", farted, "tested" and concluded that farting before starting a run improves GPU performance by 1%.

That isn't to say that there isn't a real change but 2 hours doesn't prove anything.

Mining is a Poisson process.  At 1.732 GH/s your expected # of shares per 120 minutes is 2404.  A 1% deviation would be 2433 actual.

Expected: 2404
Actual: 2433
The odds of getting 2433 (or greater) events when expected in 2404 is ~ 30%.  
So there is only a 70% confidence that your observation is the result of an actual chance.  That sounds good but in statistics 70% confidence is worthless.

Now over 72 hours the expected # of shares is
Expected: 104526
1% over expected: 105571
The odds of getting 105,571 or more shares in 72 hours would be <1% so there is a 99% confidence that the observed difference is "real".

Thanks johnyj.  I didn't know about the 1.035V step.

Simple answer is ... no.

Complex answer is that when you "mine" you are solving a specific block.  All the inputs of the block including the merkle tree (which includes the coinbase which include both the reward address and merged chain hash) plus the nonce form the block header.  A hash which solves that block only solves that exact block header.   There is no method to both solo and pool mine at the same time.

Well they can't ignore us anymore (on edit: no idea why I put "yet" here and worse it was quoted by multiple people).

Mahatma Gandhi (origin disputed)

No those are made up nonsense terms & definitions used by nobody except yourself. In the future if you wish to make up nonsense definitions it would be less confusing if applies them to new made up terms.

Like this:

When you went to school it was on the short bus wasn't it? 
Did you mom tell you it was because you were so important you got to take the VIP bus?

Well she lied.

Calling an FPGA an ASIC defeats the entire purpose of the definition.  Definitions exist for a reason.  
In related new I have a miner working with quantum computing*.  It's true.

* Quantum computing in this case doesn't refer to any commonly accepted definition of the word it refers HD 5970s.  Definitions? We don't need no stinking definitions.

Calling a FPGA and ASIC is asinine.  It is like deciding you are going to call floating point numbers, iinteger or the language the rest of the world knows as English, Spanish, or the period of time when the sun is down "daytime".

FPGA = Field Programmable Gate Array
ASIC = Application SPECIFIC Integrated Circuit.
The definitions are mutually exclusive.

Even if they weren't lets boil this down.  Your (accepted by nobody else on the planet) definition of an ASIC is so board as to include virtually all silicon chips?  Ok genius you honestly think it is a breakthrough that we "now" know that BFL uses Silicon chips.    Prior to that revelation I honestly thought they built their boards out of potatoes.   Thanks for that clarification.  What address should I tip you some coins.

Don't confuse DOA w/ orphans but yea getting DOA rate down is more important.  A higher MH/s rate and high DOA rate simply means more electricity, heat, noise for "nothing".

Donation sent and a feature request.  Have a column "shipping time" which shows # of days between ordered and shipped (for the orders which have shipped).

For the first entry it would be 122 days and the last (shipped) it would be 76 so shipping times are getting shorter but we are still far from the 28 to 42 days promised by BFL.

Hopefully that number should be getting lower over time.

I get no such error using Chrome.  Cert looks good.

Issuer:

Subject:

Alternate Name:


Either it is a bug in Firefox 10 or you got some url redirection going on.

The rake.  Roughly 50% of all funds deposited get converted into rake (the other half gets transferred from fish -> regs ).  CC fraud cuts into that but enough that they still remain profitable.  Gambling sites in generally lose a ton to CC fraud.

For player vs house gambling Bitcoin as an option alongside reversible deposit methods can reduce that fraud but for player vs player the only way to combat CC fraud is to ONLY accept Bitcoins which so far no poker operation of significant size is willing to do.

Yeah the only option to fight block witholding is an unavoidable "minting bonus", and I honestly wonder how effective a minting bonus is.  Say someone joined BM (or p2pool) with the intent to drive down luck by witholding blocks.  To have any detectible effect they would need to be large.  Say at least 10% of the pool.  The "cost" of the attack is a % of their hashing power which is the same % as their share of the pool's hashing power.

Example:
BM hashing power: ~130 GH/s
Attacker hashing power: ~13 GH/s
BM % of pool's hashing power: =~10%.
IF attacker withhold all blocks all miners lose 10% revenue.
Attacker's cost is 10% of 13GH/s = 1.3GH/s lost revenue (for as long as the attack continues).

A minting bonus doesn't really change that cost much.   A 1% minting bonus mans the attacker loses 1.3 GH/s (normal) plus 1% of the blocks he finds (0.01 * 13 = 0.13 GH/s).  A 1% bonus has increased the cost to the attack by 0.13 GH/s (~10%). 

Since there is no economic value to withholding blocks the attackers motivation is non-economical.  The minting bonus doesn't really materially change the cost. 


Sadly the "real fix" would be blind hashing (miner doesn't know if the hash solves a block or not) and that simply isn't possible with Bitcoin without a breaking protocol change.  Maybe the "next" blockchain will incorporate ideas like that.

To add to the other factors:
a) they are an established business with their own merchant account that gives them at least a fighting chance
b) Bitcoin is only a portion of their business.

If they were to switch to Bitcoin only sales with a $5000 maximum I would imagine they would be bankrupt within a month.

Well calling a custom FGPA an ASIC kinda defeat the entire purpose of using standardized terms.

ASIC = "application-specific integrated circuit"

A customizable programmable application specific integrated circuit is kinda an oxymoron wouldn't you say?

occam's razor says it is a Stratix III FPGA (my total guess would be a EP3SL150F780) purchased at huge discount to retail price because the Stratix III is EOL.

The voltage, power draw, dimensions, pin layout, package type, voltage, and board characteristics (1MB flash loader, JTAG header, etc) all match that hypothesis.

Try changing threads per GPU to 1.

Yeah but a pretty brutally bad run of it.