Depends on if your goal is theoretical economic efficiency or robustness. Consider for example 9/11, the banking crisis, and hurricane Sandy. The country would be better off if there were fewer centralized institutions. That is, there is no reason that brokers and traders based outside of the US North East to not have been able to trade during Sandy. Shutting down the exchange merely ensured that the NY based bankers didn't miss out on days of trading. If there were more medium-large banks rather than fewer enormous banks, then individual institutions would cause less of a shock to the system when they get caught in their own poor decision making and fail. Not to mention the rest of the country wouldn't be playing a heads they win, tails we lose situation of having to bail out the risk-prone banks. And if more of our financial institutions were spread around the country, there would be less possibility for substantial damage caused by direct terror attacks.

An additional effect of having more medium sized institutions scattered around the country is that it allows for economic development to occur in multiple states and also provides more laboratories for innovation. There is a point where big is big enough and it better serves the interests of the populace to start branching out. Even if it is less "efficient."

TLDR: it's easier to kill all the elephants on the Serengeti than all the ants.

...having a history of breaking backwards compatibility at a whim...

It's almost like they're writing the Linux kernel.

Maybe you should go read the ruling, it's conveniently linked from Apple's website. But to save you the trouble, here is the link to the ruling of the Court of Appeal:

http://www.bailii.org/ew/cases/EWCA/Civ/2012/1339.html

Go read paragraph 85. I'll wait.

What do you know, Apple is in complete compliance.

What's more if you read paragraph 82 you'll note that the Court of Appeal didn't even have a problem with the quotes ("not as cool", etc.) being included. It was their inclusion of the ruling from the German courts that the Court of Appeal felt was confusing. Oh, and in paragraph 86 the judge reduced the amount of time that they have to maintain the statement.

Looks to me like Apple is complying and the Court of Appeal isn't too fussed in general.

They have to take an ad out in the papers. I don't think they give a fuck about the "Streisand effect."

No, what they did was include code that automatically resizes the hero shot to take up all the available space.

That is not what's happening. In every screen shot shown in this thread the iPad mini image is scaled down , not up. As your browser window gets larger it increases in size until it is it's real size. After that it stops scaling. Also the page switches to a portrait layout if your browser window get's beyond a certain height. If one trusts timestamps provided by the web server, the scripts and images that make this happen have been on the site since the day before the iPad mini announcement.

The image in question (http://images.apple.com/uk/home/images/hero.png) is 828x809 pixels. It's been there since 22 Oct 2012. The day before the announcement of the iPad mini:

$ telnet images.apple.com 80
Trying 184.84.183.155...
Connected to images.apple.com (184.84.183.155).
Escape character is '^]'.
HEAD /uk/home/images/hero.png HTTP/1.0
Host: images.apple.com

HTTP/1.0 200 OK
Last-Modified: Mon, 22 Oct 2012 21:24:46 GMT
Server: Apache
nnCoection: close
Accept-Ranges: bytes
Content-Length: 448138
Content-Type: image/png
Cache-Control: max-age=2070
Expires: Sun, 04 Nov 2012 05:46:54 GMT
Date: Sun, 04 Nov 2012 05:12:24 GMT
Connection: close

They're not making the image bigger when you make your browser window larger. They are making the image smaller to fit crappy small browser windows.

Not sure what you are seeing, but I just went to the site. Turned my monitor so that it's 2560 pixels high and then maximized the browser. There the statement was, right at the bottom of the content, with about a thousand white pixels below it. Only thing different is that the layout of the page changed to be more appropriate for the shape of the display (1440x2560 rather than 2560x1440.) The script that manages this was last changed two weeks ago, long before the notice was added.

One would think that leaking less is out performing. Apparently presidents are like gaskets. Or sphincters. Take from that what you will.

You're putting words in his mouth. What he said was:

...porting their games to Linux showed a massive performance increase over the Windows version.

They didn't port to linux and see a "massive" -- 20% is not massive regardless-- improvement. They ported to OpenGL because they had to in order to run on Linux and found an improvement. Then they took that code base to Windows and found the same improvement. Under OpenGL the performance is effectively the same. Anyone who wants to imply that there is some large performance advantage for gaming on Linux and wants to use this as an example is being intentionally ignorant at least or just plain lying at worst.

You're missing the point. It's not about which implementation is better, we all agree that the opengl implementation is faster than the directx implementation. The G*P made the argument that two different implementations running on two different OSes tells you which OS is better. That's false. You cannot tell because you have two variables on each side of the equation and both variables are changing in the test. If A = opengl, B = directx, C = linux and D = windows, then A + C > B + D tells you nothing about the validity of C > D. What's more it seems that when removing directx from the equation and replacing it with opengl, it turns out that A + C ~ A + D.

There's a chip between the RAM and the bus. Also known as buffered.

This makes no sense. If you have two different implementations you cannot draw conclusions about the OSes by comparing those two implementations. If I have a bubble sort and use it to sort seven billion strings and run that on linux, and if-statement used sort two numbers on windows, does that mean that windows is "massively" better than linux? You need to be using, as near as possible, the same code if what you are measuring is the underlying platform performance. If you want to measure the difference in two completely different implementations that's fine, but it doesn't tell you anything about the underlying platform.

...Valve has the clout to pull this off...

Valve is a big fish in a small pond, now they're seeing a blue whale heading their way and they are desperately looking for a new small pond. Who needs Steam when there is the Windows Store?

If a 3.8% advantage is "massive", what words do you reserve for things that have advantages/improvements on the order of 50%+?

I think they gave them one of those one things. You know. Those things. What're they called? It's right here on the tip of my tongue. I hate it when this happens. One of those one things. Ah-ha! I got it: a paycheck.

The card the the iPod are just thank you's. Going above and beyond. Ya putz.