That's a bad analogy. Patrick Harnett had full control over who he lent to in order to reduce correlated risk. He had information on what exactly applicants claimed to be using the money for and the ability to demand as much evidence of this as necessary. Based on this evidence, he falsely assumed that his borrowers weren't exposed to Pirate and got screwed - that's the incorrect belief that's the problem here, and the people who loaned Patrick money didn't have this information! They had to rely on Patrick's promise that he was competent to vet applicants and that he'd made sure not to lend money to people who'd just invest it in Pirate.

A closer analogy would be if one party entered into a contract in which they gave another party money which the second party was to buy a lorry-load of cherries with, and they'd split the profit from reselling them. If the second party then goes and buys off the back of a truck in some parking lot and gets crates full of rubble instead, which party should be liable?

A user on the BFL forums posted the entire article and it doesn't appear to contain any information about die size or process node. The only interesting thing it says is that BFL is considering a future, next generation ASIC offering that moves to a smaller 45nm process node, or possibly even smaller than that. So we know that the process node they're using is bigger than 45nm - which we could've guessed anyway - but that's it.

Not true. There's a good chance pirateat40 still has funds which could be recovered in order to partially refund investors - this has happened before in other ponzis - but if the PPT investors' funds were never invested in Pirate then there's nothing to recover.

If Pirate did actually pay out in full, how exactly would this hypothetical "synthetic PPT" be able to pay their investors though? They wouldn't, which means they had no intention to pay their investors regardless of whether Pirate fails, which means they themselves were operating a ponzi and scamming their investors.

Well, Something Awful does actually have a huge creative side, but... they seem to do this thing where they get so obsessed with some group or website that a lot of the members just collectively lose touch with reality. They don't just do that with furries or Second Life or Bitcoins either - I'm catching up with the Awful Kickstarter thread, and a whole bunch of people spent pages trying to argue that Penny Arcade's Kickstarter somehow violated the site's rules, even though it was exactly the kind of thing Kickstarter was created to fund, just because they hated Penny Arcade so much. They're also chronically and totally unable to grasp the fact that TV Tropes has always been really popular with feminists, because they've collectively decided it must be a bunch of misogynistic male geeks with Aspergers, and that's lead to much weirdness and cognitive dissonance whenever their perspective collides with external reality.

Actually, the Bitcoin thread is mostly relatively reasonable, the odd weird brainfart aside. (Apparently if you install the Bitcoin client it starts using your CPU to mine Bitcoins automatically, and building and launching satellites is impossible without a billion dollar budget. This no doubt comes as news to the people who actually develop the Bitcoin client and to the amateur radio groups and schools who've built microsats and put them in orbit, respectively.)

Yeah, that branch doesn't work with CONFIG_LOOP_LOG2!=1. You probably want http://www.makomk.com/gitweb/?p=Open-Source-FPGA-Bitcoin-Miner.git;a=summary de0-nano-hax branch, projects/DE2_115_Unoptimized_Pipelined project. The voltage regulators are also indeed horribly inefficient on the DE0-nano.

Nope, single core each. Spartan-6 chips are actually quite a bit worse at Bitcoin mining than comparably-sized Virtex ones, it's just that the price difference more than makes up for it.

It does - one heatsink per chip, all cooled by a single fan blowing air downwards onto them.

His wording makes it sound like that, but someone asked whether they had any prototypes that were actually hashing at all and Josh said no:


Admittedly that was a week ago, but he was using similarly confusing language back then too. For instance, he said: "We are waiting on the bulk chips from the foundry, that is really the major delay right now" as though there were some previous non-bulk chips, which apparently don't actually exist.

Those caps are tantalum, so they don't use a liquid electrolyte. Apparently the usual failure mode of tantalum caps involves them turning into a dead short and exploding violently as a hundred amps or so of current surges through them.

Very strange. Either someone's using an unusual web browser that ignores RFC 2616's recommendations or they're faking their referrer somehow

Your site doesn't use HTTPS whereas this forum does. If I recall correctly, that means web browsers won't send any referrer information when someone clicks on a link to your website from here to avoid leaking unencrypted information about what HTTPS pages people have been visiting.

Tom's probably too polite to say this, but frankly BFL have done that because they're run by skilled scammers, sorry, marketers. The pictures look convincing but are basically worthless - they're non-functional mockups without the actual ASICs and it took a huge amount of badgering before anyone could get them to confirm that. BFL won't know if their current designs work or if they need redesigning for several weeks even if they manage to keep to the announced schedule.

Quite a bit less than it's going to affect anyone unfortunate enough to do business with BFL, I should imagine - they won't even have anything complete enough to test for RF emissions until shortly before they're scheduled to ship based on their own statements. Also, I don't think BFL's existing products have been tested for emissions and registered with the FCC either.

Engineering samples are by definition an actual, mostly-working version of the chip, possibly with various issues and really poor yields, but functional enough to test with. BFL have said they don't have any chips like that yet, so these pretty much have to be fakes (though I believe the more polite term is "mock-up").

Try the following exercise: take a map of the US interstate system. Imagine that, instead of being government-run, it was owned and run by a private operator. Explain how you would build a competing long-distance road network, remembering that since all those roads are privately owned you need the permission of the owner to build anything that crosses them them.

What's more, since this monopoly is going to be more profitable for the road operators than running competing roads, without government intervention the free market would inexorably turn privatized roads into a monopoly.

They dropped kerosense prices to that level in the complete absense of functioning free-market competition. For example, they had secret deals with, or in many cases even owned, many of the pipelines and railroads which meant that even if someone had a cheaper and more efficient of refining oil they couldn't compete with Standard Oil because they'd have to pay far more to transport it and far more to buy crude. Likewise, competing railroads and pipelines couldn't make money because anyone using them had to rely on Standard Oil-run transportation as well and was driven out of business that way. I'm pretty sure the contortions Standard Oil went through to put their competitors' transportation costs up screwed over everyone else who needed to transport goods and distorted pricing in totally unrelated markets too, such as food.

It doesn't actually matter how intelligent they are, crowdfunding investors are inevitably going to make poorer, less well-informed decisions than proper venture capitalists. It makes no sense to spend more than the maximum amount you stand to lose on investigating an investment opportunity, so the smaller the individual investments the less investigation the investors do before putting money. Even if some of the crowdfunders are actual venture capitalists they're still going to make worse investments, though they'd probably do better at it than random outsiders with no investment experience.

(I'm pretty sure this is why so many investments are restricted to accredited investors too. Before the Internet, raising money from large numbers of small investors was very expensive, so the only reason to do it was if you wouldn't be able to raise the money from larger investors - for example, because the investment was a scam and you were taking advantage of the fact that smaller investors are less able to sniff this out.)

The best solution is probably to not force users to hit the back button in the first place - instead display the usual comment/PM submission form with their comment already filled in and a note at the top telling them they need to wait.

I'm pretty sure that applies to everyone - by making proportional payouts on the long blocks where it pays less than PPS and then retroactively switching the payout method to PPS as soon as they had a cluster of short blocks where proportional would've paid out more, they effectively short-changed every single miner even if they didn't pool hop.

Which - if I'm reading the spec right - is quite hard to do with the current protocol. Even if you change the difficulty at the exact same moment as sending out a new work unit, the difficulty change applies to all the previous work units and miners are allowed to continue working on those at the new difficulty unless you invalidate all of them at the same time. Except that if you do that, you're making the miner throw away work they've done without submitting it. There's a mismatch between how the protocol wants you to handle difficulty changes and the way you actually need to handle them to prevent cheating by miners.

So basically, the protocol's flawed in exactly the way kano said it was - in order for it to work reliably and without opening pools up to exploits, difficulty changes need to be tied to work units rather than happening in the middle of a work unit.

That would be 200M $ worth of BFL hardware at current prices. There's a substantial financial incentive for them to drop their prices once business slows down - the marginal cost of ASICs is fairly low and the up-front cost fairly high, and they probably want to make as much money on that upfront investment as possible. Of course, that doesn't leave much room for purchasers of BFL's devices to cover their investments either.