So, you're saying that, because our daughter was reading and doing arithmetic at a fourth grade level by the age of 6, my girlfriend and I are exceptionally good parents?

Interesting. I've always had lots of trouble with Intel graphics, but never any trouble with Nvidia. My current (2 year old) Dell laptop has an FX 1800. It really kicks tail compared to the newest ones some of my coworkers are using. Their (also Dell) laptops have Intel graphics (don't know which).

We skipped Vista and only started using Win7 in March of this year. Similarly, we started using XP in 2004. If we follow the pattern, we might be using Win 9 about 3 years after it becomes available - assuming we stay with MS Windows.

One of my coworkers used to work in one of those sweat shops. Then, when she was 30, she met an engineer (electronics) from a customer of my company's. She's been working here for over 2 years, now, and is very happy - has learned a lot more, is paid much, much more and only works 8 to 10 hours per day (like the rest of us SW people in my company).

I also liked the G4 Cube. And I still get a laugh out of one reviewer's comment on how fast it was compared to other PCs at the time: "Where's the drama?"

I liked the Newton, but it was too expensive for me at the time. Fortunately, I had a friend who had one, so I got to play with it. I really liked the "graffiti" writing-to-text feature. Palm also had a similar writing-to-text feature. I still have an old Palm T3, that I used for many years. I'd still be using it except the only way to exchange data between it and anything else is with SD cards (well, it does have IR communications, but nothing else I use does.) I am mostly happy with the Android tablet I now have, but I really miss the writing-to-text feature that worked so very well for me on the T3 and the Newton.

Difficult is relative. An admin trained on Linux is going to have (some) difficulty retaining to admin Windows. It won't be as easy for her/him as for someone starting fresh (and has nothing to unlearn).

Whatever the "tool", there is always the perception that the "other tool" is harder. And there is always the perception that learning even a radically different new version of the "same tool" is easier than learning the other tool, even if the other tool is lot more like the current tool than the new version compared to the current tool.

Emulate an x86 or supported ARM on a 6502. Of course, performance would likely be seconds (or minutes) per FLOP. Aside from the slow instruction cycle rate of the fastest made 6502, all your "emulated" RAM would have to be on disk.

It's Amazon's cloud, so Amazon might own them. Depending on how the government interprets the contact between Amazon and the authors.

(In the brief on EFF's website,allegedly a copy of the government's brief, the government is presuming that the combination of the contracts between Carpathis and MegaUploud, and between MegaUpload and Mr Goodwin sufficiently restrict Mr Goodwin's ownership rights to the data that he effectively has surrendered his rights to that data. This is not quite as the same as saying Mr Goodwin has automatically lost his ownership rights in said data. Rather it is saying that Mr Goodwin effectively and voluntarily surrendered his rights.)

I saw one of those 1U rack modules like you described a few years ago. It had a tiny LCD display and a 2x13 keypad on the front so you could do your key management with out relying on any external equipment. (it also supported SSH and direct serial plug-in) According to the network admin at the office where I saw it, it had 2 Ethernet ports - one for local clients on an isolated LAN, the other connected to the insecure, internet connected LAN. Clients on the "secure" LAN could upload sensitive files to it and specify where the encrypted copies would be sent (mostly to backup storage, both on- and offsite.) or where to fetch an encrypted file and download the decrypted copy.

Also, years ago, I used an USB-key based encryption engine. I forget who made it, but it was amazingly easy to use. It used a USB-to-serial converter profile. For text files (including UTF-8 encoded), a program could use it by opening the device file, write setup commands, then the text followed by either control-D (aka EOF) or control-Z. At same time, you read the encrypted/decrypted text (or read signature or verification after all the input text is wriitten). Non-text data required padding the input data to a multiple of 8 bytes and telling the device the padded length. You would then write/read that amount.

The app itself should not need to know about Phone State.* The OS's SMS_API support should be able to queue messages and hold them until the state allows sending. Ideally, the SMS_API should allow the app to specify Cell Only, WiFi Only or Either, then only send queued messages when the appropriate connection is available (and via the appropriate connection).

As for Identity, an app could maintain it's own identity. Not as convenient, but some people would be willing to make this trade off. Of course, over a cell connection, the network is going to know the ESN of the phone that sent the message. I suppose SMS-over-WiFi would also have the ESN, but for other IM protocols, that could be avoided. (This assumes the OS itself doesn't unnecessarily leak personal information)

Likewise, an app could maintain its own contacts data. Again, less convenient, but some people would be willing to make the trade off.

* For apps that want to pause while a phone call is in progress, the OS could handle that for apps that request being paused.

Pay the victim? Probably not. Paid their lawyers. Most certainly.

No. They'll either start charging for their services or figure out more ways to monetize the users. (or both)

Economy of scale. Get a larger audience by morphing it into something they think will appeal to the masses. Product uniformity reduces unit cost. The sales pitch for cable TV was hundreds of channels so there will be something special for everyone. That's not what we got. We got hundreds of channels of slightly different flavors of the same old same old. Why? Because it's cheaper to produce 10 shows that differ only in the cast and the names of people, places and things in the scripts than it is to have even 5 truly different shows. If a new little guy comes along and does something truly different,fine - as long it stays little. But once it gets large enough to start taking away audience share, it either gets assimilated and homogenized, or it gets quashed. Of course, any that dont get that big will likely run out of money and die.And even if you don't get assimilated or quashed, you will feel the pressure to homogenize.

There is also where part of the problem lies. If you burn something that is owned by some one else, you could be guilty of a crime. Said crime being comesserate with the value of the thing burned. There are Christians who consider burning a Christian Bible an act of blashemy.