This sort of surveillance does sound more like what what you would expect out of the CIA -- which is hampered by federal laws limiting them to spying on international communications and foreign nationals -- or the NSA -- which has invested in a huge new facility after admitting that there's just not enough power to come close to breaking a significant amount of encrypted traffic. The big question is why the FBI would jump into something it's never been a major player in before.

Best guess: they're trying to update wiretapping. They've been getting increasingly alarmed and vocal about just how little wiretapping actually buys you now. If you really want to keep something secret, you can just use an https encrypted connection to any one of numerous services that keep no records and have no mechanisms for spying on their users.

They recently floated the idea of requiring backdoors be installed into such service, the way telecom hardware is legally required to support conventional wiretapping. that idea had no real support in technical or public circles. Even if you trust your government, it's much hard to game a system that requires someone to go to a location within the your country and physically connect to equipment owned and operated by a someone else than it is to find an exploit in a protocol that can be prodded by anyone online and which would have to be implement by everyone from Facebook to Club Penguin.

With no widespread support for spying-as-a-service, they're stuck traffic-tapping the hard way: inspect every packet for the start of an HTTPS handshake so you can break the connection, or somehow crack an encrypted stream with incomplete knowledge. They still have no idea how they would reliably accomplish either of these. However they do it, it will probably require new laws to make it feasible. It sounds like the program casts a wide net in an attempt to find something that works, and is trying to keep it quiet because they don't know what solution will rise to the top, or how knowledge gained about the process now could be used to defeat it technically or legally later.

You could just use a kernel password and tag on a date: MyPass112011, MyPass122011, MyPass0112012,.... That's what most of the bank officers do when they're forced to do that same thing.

I agree with your general sentiment that many of these icons are not particularly outdated so long as you accept the underlying metaphor to begin with. For example, a magnifying glass is probably no more or less used today than it was 60 years ago. It was always a very loose metaphor, referring more to a caricature of Sherlock Holmes than anything else. Bookmarks are also quite easy to grok if you accept the notion of the web as a "book" of independent documents (which even in the 90s seemed weird to me, as places in cyberspace metaphor worked much better for the web than the documents in a sequential book metaphor did, even then).

For some of your specifics, though, I have to disagree. First, there is a definite bias towards items a paper-heavy office. That's fine, but the largest consumers of technology don't work in those anymore. Some are not in offices, and others are in offices where all of their work is through a company system on the computer.

Anything that's based on technology from 60-100 years ago is definitely dated, because they have to pic a single incarnation of the technology that stands out as much as possible from other items. Modern design aesthetic is to smooth corners, hide the pokey bit, and as much as possible reduce every device to a rectangle with a screen (which maybe you can touch).

Polaroids look like Polaroid prints. Most pictures look like Kodak prints (rectangles with a picture covering it completely) and pretty much no one prints their photos anymore. They are stored on their computer instead of in an album, or carried on a phone instead of in a wallet.

Many people receive bills in the form of an email saying either, "it's time to log in to the web site and pay your bill," or, "we have deducted the required amount from the bank account you provided. Thank you for using auto-pay."

Microphones used in bars an stages look something like a metal ice cream cone -- a conical grip and an a wire mesh screen -- not in the studio style, like a mesh hot dog suspended by a forked base.

The voicemail icon is wrong on a couple of different levels, because the answering machines that were replaced by voicemail hadn't used a removable reel-to-reel cassette in a decade. They really had to reach back.

I'm a little surprised how little I've seen so far on how difficult this makes security for browsers. Because most of the TLDs now are country codes such as .uk, and those countries in turn have their own sub-TLDs suck as example.co.uk, browsers keep a list of which TLDS and sub-TLDs are real suffixes. This lets them know that mail.google.com can read/set cookies for google.com, but evil.co.uk can't read/set cookies for all of "co.uk", much less safe.co.uk.

As you may have guessed, this doesn't always work out properly. It's kind of a crap shoot with sites that use the country TLD directly, such as nhs.uk. With unlimited and variable TLDs, this implementation becomes even more questionable.

Does anybody know if browsers have gotten smarter about this in the past few years, or are we racing towards one of those security nightmares that forces content companies and standards bodies to actually get their acts together?

The cited section basically talks about widespread French spying on American companies, and then claiming it was all a big conspiracy to make the French look bad once it came to light.

The fact remains that even if the U.S. government were willing to steal information and share it with American companies -- and this is pretty unlikely given that the U.S. doesn't have the sort of cozy, formal overlap of public and private sectors that France, China, or even Great Britain have -- most other countries haven't had anything we want. You have to go back to 1793 Pawtucket to find a good example of the U.S. gaining an edge through industrial espionage.

Don't get me wrong, the U.S. government has shown it's willing to co-op private technology for its own ends. (For example, when it co-opted the patent for Phillip French's Crater Coupler and then used that state secrets privilege to get the dispute tossed out of court.) They just haven't been shown to help private U.S. firms with any of it, or to do it specifically to improve the competitive advantage of a U.S. company.

Or they could, you know, not bring weed into Texas. It's not like it's hard to find anywhere in America. Sure, there'll be a premium for competing with other tourists, just like with food, lodging, and gas. But does that really make taking it across state lines (on a week you know they're going to step up their presence on highways to deal with all the out-of-towners) seem like a good idea?

Going through old comments. This will probably never be read, but what the hell:

re: equivocating
Forethought + disagreement = shilling. Got it. Not a very sound opinion, but certainly a convenient one.

Regarding the more substantial part of your post on the free market:

A market is free, not free, or somewhat free. Economists aren't thrilled by it, but the fact is that in practice most markets are not completely free. A bigger mistake, however, is to refer to the Market as a singular. There is no market. There are markets. And many of them are not even remotely free.

Most countries do not permit prostitution: sex is effectively free to distribute between people, so long as it's a gift. It's a service you just aren't permitted to sell.

Emissions trading was a hot point in debate in recent years because it introduced a market where none was before. The issue, of course, was that many people thought it was evidence of corrupt politicians letting big industry buy their way out of failure. In fact, a better way to think of it was the government attempting to put a price on something that (up to a limit) had been given away for free: clean air for industry to sully. This was to be an application of market pressures to regulate something that was previously very binary. And, for the harshest toxins, policy experts all made the same recommendation: don't allow such emissions allowances to be traded at all.

The printing of backed currency is reserved to various governments, which is pretty much necessary for those commodious little pieces of paper to have any value. Back when the value was backed by specie, such metals ALSO had to be traded through the government, or else through licensed buyers (a fact most people who fantasize about the gold standard seem to have forgotten).

Food and devices are regulated for safety and fitness for a particular purpose. Labeling, fitness, and safety laws exist to cover gaps in what a consumer can know. A consumer would never choose to buy poisonous food, or faulty cribs, but since they can't know 100% what they're buying, the market distortion is necessary. This returns to what I said about a perfect solution: as consumers approaching perfect information about the products, their processes and compositions, and the alternatives, free markets become optimal solutions for trade. But in the real world, we can never get perfect information, at least not without cost.

Pretty much every product on the market today is subject to some form of regulation, even raw labor. In the real world, a free market is a nice goal, insofar as improvements in consumer awareness allows us to consider a free market to be a better and better approximation of our ideal means of maximizing resource utilization. But the real world has limits, and every industry fudges on the "free market" notion a little, and some industries are complete non-starters because of it.

It would be naive to hope that a few examples and a relatively short post on the internet will change your mind. I try, but I don't hold much hope. I will simply say that this last comment of yours illustrates _exactly_ what I said earlier about the Free Market as an object of religious devotion. The invisible hand does not guarantee you cheap goods, or even long-term stability. It guarantees you that and small discrepancies between wants and means eventually stabilize (though they are of course free to be replaced by _other_ disturbances to the equilibrium). It's an observation about the long run, and nothing more. Saying that that is all ye know of the markets, and all ye need to know is just ridiculous. It's like saying "1 != 0" is all you really need to build a Cray supercomputer.

I will go back to my original point:

Going through old comments. This will probably never be read, but what the hell:

Business method patents are not identical process patents. All sorts of processes are patentable, such as teh processes for distilling chemical components, manufacturing steel, etc. Business method patents are a recent invention arising from a Federal Circuit case, where the judge ruled that pretty much anything can be patented. Amazon attempted to patent the idea of an online "shopping cart": not a specific implementation in code, or even an abstract management system, just the entire notion of "a list metaphorically referred to as a 'shopping cart' of items a person has selected as those he might like to purchase is the near future." These have been written to cover whole business _models_. Including, ironically, a business method patent for patent trolling.

Looking back, I think that my post above would have better served as two separate posts. Had I not been in such a hurry, I probably would have written one post commenting on the similarity between the unwritten rules and basic patent protection, then another later on the difference between software and hardware open source.

Again, my goal is not to belittle the movement or the practices the community is using now. I'm just concerned that the fact that there are fewer protection mechanisms for hardware projects will make licenses more difficult to enforce, and ultimately will make open source hardware projects more difficult to manage and keep in line than their software counterparts.

Fortunately, there are still software protections available for applications requiring controller firmware. That's usually the difficult part of a project to even experienced designers, and the firmware covers a lot of the real knowledge that goes into a working device (proper timing, error checking, signal processing). If an application is trying to avoid costly embedded operating systems, then the tasks become more difficult and there is even more reason to embrace an open source solution rather than try to start from scratch.

If the open source hardware movement wants to become sustainable as more than a hobbyist endeavor -- and bear in mind, that's not strictly necessary, as there are a lot of electronics hobbyist out there -- they should focus on what's useful to developers as much as end users. Open source software is successful mainly because developers leverage the free operating systems, development tools, and packages to minimize R&D time on other products like servers, monitors, and mobile devices. The end user doesn't care about the fact that 90% of their toy is based on FOSS, but the companies that built the toy benefit from it, and so too does the FOSS community.

Coordinate with FOSS developers to get better tools, and for the love of God standardize on and optimize a good suite the way so many standardized on GNU in the early days. (Sadly, most people I know use Eagle, which is a good schematic capture and layout suite, but is only the best free-as-in-beer option. The free-as-in-speech options I've looked at are ages behind.) Embrace and develop a good, stable RTOS that doesn't require a $10,000 per end product fee, and work to ultimately get it certified for safety-critical systems. Grow the tools, the tools will grow business, and businesses will grow the community.

I think what's being proposed is actually a weak form of patent protection.

"So I see you're selling something called 'noTV'. Is that a clone of TV-B-Gone?"
"Yes."
"Did you improve upon it somehow?" (see "Cloning ain't cool")
"Yes."
"Great, then you're doing something useful! How did you improve it?"
"Okay, so that was a lie. It's a direct clone."
"That's not good. You shouldn't do that. At the very least you should pay royalties you work out with the TV-B-Gone team." (see "We pay each other royalties...", "we credit each other, a lot")
"No, thanks."
"Well! Expect a stern look the next time we see you!" (As I said, weak protection.)

If you like the idea of patents, but ultimately want them to be toothless and enforced only by social mechanisms, then these ideas are for you. Which is about the right level of enforcement, given that most of these things can't be protected under patent (not novel) or copyright.

Open source software actually has stronger protection mechanism under copyright (and in some instances such as a Linux kernel, software patents) to make up for the lower barrier of entry for imitators (copiers). At the very least there are licenses that let you stipulate what applications you don't want your software being used for, how you can brand it, whether improvements MUST be fed back into the original project, and what kinds of other software it can interface with, if the author is so inclined to place those restrictions on a work. And ultimately, those agreements have legal teeth.

For hardware of this sort, the barrier to entry is only cost to build and market such hardware, and the protection is very weak. There are some trade secret laws that electronics manufacturers can usually invoke for direct rip-offs before a product hits market, but after it reaches the market tear-downs are legal, and products are easy enough to copy. Most designs boil down to "reading the IC manufacturer's intended application circuit from the datasheet," and that's about it. Very difficult to protect. That's why most cases today (such as Apple vs... well, everyone) involve using software patents to disrupt a competitor.

I expect that the open-source hardware movement will have an increasingly difficult time enforcing these unspoken rules as it gains traction. And none of this touches on problems arising from applying the open source model to hardware, such as whether or not updating an old designs based on EOL'd parts to use newer parts is a new design, a major improvement, or a trivial change.