Dead wrong. Intel drivers are excellent and I and many others have had great success with them. They also usually work quite closely with the kernel community as a whole to make sure things work as expected; that's why what this article is saying seems to out of character for Intel. For instance, try searching for "intel.com" in the git commit log. Lots of kernel developers are on Intel's payroll, including core people like Alan Cox.

You also gotta wonder, what exactly is that technology supposed to be anyway? Tor is readily available, as are file uploading sites and message boards accessible with tor. If you have the technical know-how to use the "Openleaks technology" to publish your leak, you probably already have the know-how to use Tor anyway. Wikileaks also offered "mail-your-leak" dropboxes, a very secure option that has nothing to do with source-code.

The real thing Openleaks could add is vetted technology to remove things like embedded tracking of documents, for instance the metadata in jpegs and word documents, as well as technology to defeat stenographicly hidden per-file tracking codes. I haven't heard of anything from Openleaks even mentioning that stuff, yet defeating can be vital if a leaker wants to remain anonymous. It's a much harder problem than the actual publishing as well. It's also a problem more easily solved by human efforts, such as trusted individuals that re-word and summarize documents and publish the summaries rather than the originals directly.

What Openleaks can't do with technology is vet the leaks to ensure authenticity. For that an organization like Wikileaks makes much more sense, as does traditional journalism.

What makes you think they never had that "critical mass of readership"? I'd argue their most important readership was other journalists, and pretty much every leak they've ever published has been picked up by the press in some form or another. That the general public can read the leaks easily is a side-effect, necessitated by the fact that they want to keep the journalists honest, and by the fact that the term "journalist" should be interpreted fairly inclusively.

It's better than that: sure helium can leak out of your hard drive enclosure, but it's also the only think that can leak into the enclosure as well. Helium is present in the atmosphere in small quantities, so the pressure in the hard drive will track atmospheric, albeit very slowly, yet still maintain a nearly pure helium atmosphere.

Linus Torvalds has an interesting comment regarding your style of programming: plain-old C can make a lot of sense precisely because abstractions wind up being implemented with relatively verbose code. Compare C++ code using operator overloading, vs. object-oriented C code using functions, or at the very extreme, lisp where the language itself is different due to heavy macro use. It's the C code that's most likely to be readable by other maintainers, even if it takes more lines and typing to do what it does. The Linux kernel has a lot of developers with relatively little management structure or co-ordination, as well as being relatively simple programming in the sense that they can avoid a lot of the more complex logic more often seen in, say, GUI code, so you can see why that kind of trade-off can make a lot of sense.

I'm an electronics designer and I'm in the exact same boat, although where I work we're pretty lax on all the other stuff, so often we just have schematics to work from. Granted I'm at a schedule-driven startup, so being lax can be a valid tradeoff between time spent on documenting stuff, and time wasted due to forgetfulness and the bus factor. Just like with programming (a former career path of mine) even just some simple notes on what equations you used to design something can be hugely useful I find.

Nah, that won't be an issue. The secret sauce of my plan is actually the marketing people; required to convince the business world that a shitty low-resolution black and white company logo and unreadable signature really is going to make a better "personal connection" with their customers...

...and if you are using it for an actual application, it's frustrating how expensive the damn stuff is. As an electronics designer I'd much rather just gold plate everything for durability, rather than having to fight the cost-engineering department every time.

Gold bugs are a very real part of the reason why consumer electronics are unreliable. Remember that every time a crappy tin-plated connector fails. Heck, silver bugs too: for high-current connectors and RFI shielding silver is usually the best option, but again it's unaffordable for a lot of applications.

If anything, BitInstant's paycard is an attempt to avoid this problem by giving you a way to easily spend Bitcoins with people who haven't adopted it. If they can eventually find a way to somehow allow you to deposit funds from a bank machine onto the card in the future, even better.

What they're doing is kinda like an early fax machine company setting up a nice fax-to-postal-mail service. Imagine being able to skip days of letter delivery overseas by simply faxing an office in the same country and city as your recipient...

Actually, excuse me, I need to find some investors and a time machine...

As I said, the currency conversion could work either way. It'd be easiest doing it the way you describe, however a big part of BitInstant's other business doing fiat-Bitcoin conversions is that they have good mechanisms to complete the conversions as fast as possible with accurate pricing. This may enable them to do the conversion fast enough that they'd somehow load the dollar balance on the card as you attempt to spend it. Effectively they'd intercept that "over-spent" notification somehow in the computer system, and instantly transfer over dollars as required. They then take the risk that they can't buy the dollars back at a profitable exchange rate.

Re: risk, any balance actually on the cards would be the responsibility of the bank actually issuing the cards. This isn't any different from any other pre-loadable card. BitInstant itself is a registered company in New York, NY with a physical office. The identities of the people running it are well known. On IRC they hinted that they're looking into allowing you to hold the private key associated with the Bitcoin balance on your card, so if BitInstant failed you could recover the Bitcoin balance yourself. (the BTC funds wouldn't be pooled in that case) The issue there is the person holding the card spending the coins before BitInstant can take them to sell on the market. However, the card isn't anonymous, so they know who you are if you try to pull that scam and will just ask you to give them Bitcoins to replace the ones you took. Also in practice even a transaction that's been on the Bitcoin network for only a few seconds is difficult to reverse.

Currencies aren't really backed by anything these days, other than the fact people in certain economic regions generally use a given currency.

All this stuff about governments backing currencies by accepting them as taxes ignores one critical thing: How does the government decide how much to take? Suppose you were in Canada, running a business happened to accept US dollars exclusively from your customers, and at the same time happened to have all your expenses in US dollars. (quite easy to do with an cloud-using internet startup actually) The Canadian government will of course demand you pay your taxes in Canadian dollars, but as for how many Canadian dollars you owe they'll use the exchange rate between CAD and USD.... which is in turn determined by the currency markets, which figure out the price by, guess what, how much people want Canadian dollars vs US dollars.

Now imagine for some reason that every company in Canada becomes such an internet startup overnight. Even though the Canadian government demands that you pay taxes in Canadian dollars, they're going to have to ask how many based on the exchange rate at commercial exchanges. Chances are the value of 1 CAD compared to 1 USD will plummet as soon the only reason you'd want to have Canadian currency is to pay your taxes. Eventually some weird equilibrium will happen where the value is totally determined by how fast the government sells of their Canadian dollars to buy things that the government wants, which means the government might as well just peg the value of the Canadian dollar to the US dollar at some arbitrary multiple.

Of course, this doesn't happen because Canadian companies often want to be paid in Canadian dollars so they can pay their employees and other Canadian expenses. Said Canadian companies often export things to other countries, hence the exchange rate being determined by the supply and demand between Canadian stuff being exported and non-Canadian stuff being imported.

Bitcoin, modulo people holding it due to speculation, acts quite like this sort of thing, except that you can picture Bitcoin as the intermediary between two different currencies. Therefor the exchange rate of Bitcoin is determined by how long it takes to convert from a currency to Bitcoin and back again, times how many people are doing these types of transactions, divided by the total amount of Bitcoin on the market.

Really the value of Bitcoin is based on the velocity of money, and speculation. The latter is just a matter of supply and demand, it's the former that's more interesting. Suppose Alice, in the USA, wants to transfer value to Bob, who lives in Germany, using Bitcoins. The actual exchange of value is going to look like this:

• Alice gives Charlie, who runs an exchange, some US dollars. Charlie gives Alice some Bitcoins.
• Alice sends her Bitcoins to Bob.
• Bob gives Dan, who also runs an exchange, his new Bitcoins, and Dan gives Bob Euros.

Now if this goes on over and over again, Charlie is going to be short of Bitcoins, and Dan short of Euros. So on Charlie's exchange the price of Bitcoin relative to the US dollar will rise, and on Dan's exchange, it'll fall. Enter Anna, are arbitrage trader:

• Anna sells US dollars on the Forex market, buying Euros.
• Now Anna wires those Euros to Dan's bank account in Germany, and Dan gives Anna Bitcoins
• Anna then gives Charlie Bitcoins, getting US dollars back.

Why didn't Alice just wire Euro's? Well, bank wires have high fixed fees, so Anna can amortize the fees over one huge wire transfer in a way that Alice can't. Also, Anna doesn't care if either side know who she is, but Alice might.

The key thing is that all these steps take time, which means Bitcoins are tied up in the system, reducing the supply. So by basic supply and demand, the price goes up, based on how useful it is to use Bitcoins to transfer value between people, and how fast the whole process can happen. Some people find them very useful to do this, especially pseudo-anonymously, notably the drug trading site Silkroad. Note how curiously a more efficient Bitcoin market, in terms of fiat conversion, will lead to a lower Bitcoin price.

The key thing is that Alice and Bob don't actually care that much what the price of Bitcoin relative to other currencies is. They just care that the price doesn't change fast enough that one side or the other gets ripped off during the time it takes to complete transaction. The faster those transactions happen, the less of an issue this is. If either side can quickly buy and sell their coins, something made easier by things like BitInstant's new paycard, they don't have to worry about exchange rate volatility as much. And speaking of, while people in the US don't have this problem, ask someone in, say, Iceland, with its tiny economy and its own currency, how stable the value of the Krona is relative to other currencies...

As for what happens if cash systems collapse... frankly if that's true, chances are something is seriously wrong with society. If the internet is still functioning Bitcoin might still be working well enough as a technological project that your coins are going to be worth something. If the internet isn't work, well, you're coins are going to be worthless because people can't transfer them. Of course, if you're in such bad straits that what you really want is goats and chickens, you might also quickly find out that you can't eat gold...

Like any other tradeable thing: on an exchange. If you don't think the exchange rate BitInstant gave you is fair, it's easy to go to a site like BitcoinCharts and look at the spot prices at the time on any of the exchanges in existence for the currency you were converting too. After all, the back-end of BitInstant is just a computer program that automatically submits sell orders as required to meet their float requirements. Buyers then buy the Bitcoins, giving BitInstant fiat currency in return, which they then send to the bank handling the debit card backend via wire transfer.

The key thing from BitInstant's point of view is to have solid reliable backend software that submits orders fast enough, and statistically understands the volatility well enough, to figure out what instantaneous exchange rate they can make a profit on while still giving a good enough rate that their users don't feel ripped off. This is easy if the price of Bitcoins is rising, but if it's dropping they could easily be in a situation where they give you too much fiat currency for too little Bitcoins. Part of the issue too is that spreads between buy's and asks on Bitcoin exchanges tend to be much larger than you'd find in government-issues currency exchanges, simply because the market is smaller. They're looking at spreads in the region of one percent, rather than hundredths of a percent.

Ultimately though the process is no different from what your bank does if you use a USD denominated debit card in Europe.

Specifically one re-loadable with Bitcoins.

From what BitInstance has said it either works by simply converting your BTC into USD or some other currency when you deposit the BTC to the address on the card, or they've done some clever thing where they convert the appropriate amoun of the BTC balance associated with the card it as you make a transaction. (depositing the currency to the actual account associated with the card just in time for the payment to go through) Credit is never offered.

Of course, this is why they don't really need much help from Mastercard: the technology powering the card is already common for branded payment cards like those ones you can get to give as gifts, just without the crazy fees and with a much more reasonable %1 currency conversion fee and the flat $10 or so fee to buy the card. That conversion fee being pretty similar to what you'd pay at a regular bitcoin exchange anyway.

Ultimately this is like Paypal's debit card: just an easy way to spend your Bitcoin balance. If you happen to have a lot of coins for whatever reason, be it you run a business, mine, or invest in them, this card might be for you. Otherwise don't complain that a niche product doesn't make sense for you.

From a security point of view it's a very easy product to offer. When Bitcoins are sent you can almost guarantee the transaction has gone through, no charge-backs, after one confirmation, or an average of 10 minutes. That rises to essentially guaranteed after six confirmations. (I'm pretty sure no-one has ever managed an actual double-spend with even one confirmation) The rest of the process is subject to the exact same risks as any other pre-paid debit card, hence the transaction limits they've said they'll apply of something like $1000 a day. Basically they just want to limit losses if someone steals/clones your card, or they screw up somewhere.