I've been using Linux as my primary desktop environment since '97, and I agree with you.

I think too many the people who bother to involve themselves in coding for Linux (and particularly towards the end-users) are lacking in .. well .. empathy: They are simply not able to reset their mindset such that they can see how your application will be perceived from a first-time user. I know the tricks, how to google effectively, the knowledge base in my head is sufficient etc etc so I'm happy using Linux. But if I was a less curious and stubborn person, I'd given up a good while ago.

Linux is overall also far too anarchistic in its nature to produce a good Desktop OS. It will remain as a superior server platform, a good development platform, a breeding ground for new stuff, but it won't ever gain end-user traction. KDE and Gnome will continue to live side-by-side, Ubuntu dares not be controversial enough to completely drop either. And the desktop will therefore be both slower and more confusing than it should be. We make stuff work, but we rarely make them shine.

The Linux desktop even missed the netbook wave, which was almost custom-made for it to shine. XP on my HP Mini 311c is good, solid and misses fewer frames (especially in Flash) than Ubuntu running the official nvidia ION driver. What a missed opportunity! (I realize M$ used dirty tricks to lock Linux out, but those frames don't drop themselves - Ubuntu is slower than XP)

Linux is adequate for almost anything, but good at only a few select things, forever trapped in the 80/20 rule. That's the curse.

Coal is relatively plentiful. An estimated 56 years is a long time to our current system.

I doubt we'll be forced out of our hydrocarbon dependency in a long time. By then, it'll be too late. Depending upon natural limits is a poor strategy anyhow - once we hit those limits, it may be too late because hitting the limit may limit the overall production (by humans or by nature, that devil is hiding in the details, sniggering).

I don't watch TV (aaarhhh!), I use adblock, and most of the stuff I subsribe to comes with only a few ads. My phone is not listed for telemarketers. And on my mailbox, there's a "no advertising, please" sticker.

Life is good. At least in that respect.

(I still get hassled a bit by pushy salespeople in the streets, though. I'm just waiting for the Norwegian law enforcement to become so inefficent that I can punch them without risking getting hassled by the Man. And the public space is filled with advertising.)

Good point about the startup cost. However, if you download the files in parallel rather than in order as we do today, the start-up issue will be mitigated. Also, most people have spare upload bandwidth. Most dedicated file-sharers don't. And anyhow, most update services will download in the background anyhow and then notify the user when the update is ready. The end-user will not notice the increased lag.

And maintaining the cache is really simple: Seed until ratio hits .. say .. five or there has not been any access to that file for five days. Then remove the .torrent from the bittorrent client and delete the file.

Ok, so to summarize a good, working architecture:

1. The current mirrors will also run bittorrent clients and will act as primary seeders for the torrent server.
2. The torrent server will only provide peers that have pinged back to verify they are indeed patched.
3. The torrent server will list non-mirrors first when listing peers. But only those who are patched.
4. The servers themselves will run a bittorrent client that is transparent and will seed until the ratio hits x or there has not been access to that package for x days. It will then clean its cache.

Who's up for the task?

(I got max karma already, so I'm not whoring)

Another global problem in a nation-based world.

That's interesting. Seriously. I thought the machines were able to distinguish between living matter and printouts (like looking for signs of movement, adding an IR channel). If the iris scanners can't do that, the product is DOA.

No need to shoot the account owner. It will actually become easier to rob someone. Now you gotta threaten them to give up their PIN. With iris scanning, the robber will only need to drag your ass to the ATM and hold your head in front of the iris scanner.

Not that I like the technology, but your example seems wrong. Also the other assertions about the risks of shared authentication may be misguided. Those who hold your iris info will not necessarily be able to use this to authenticate themselves as you. Manufacturing an iris is considered difficult. And the iris scanning may be digitally signed when authenticating yourself online to get around the problem of using the same authentication to several accounts.

"I think there is a world market for maybe five computers"

But who said it? Not Thomas Watson

Well duh. Or the web site you link to re-engineers its layout without providing back-compatible redirection...

Anyhow - I store bookmarks in my bookmark collection. Twitter is mostly a way of quickly sharing and discovering ephemeral information, anyways: If I want to keep stuff, I export it from the cloud.

Multifunction office printers/fax machines/scanners have horrible, horrible UIs with lots and lots of buttons. And they must - because they are multifunction, and the buttons are function-specific. Hopefully, we'll start to see such machines that can be used be mere mortals and not specially trained secretaries.

Come to think of it - WHAT A HORRIBLE IDEA!

methinks you suffer from an acute lack of imagination and inspiration if you can't do something useful with your iPad. That said, I have a Galaxy S (which shares a lot with the Galaxy Tab) and it's a nice little piece of hardware with a few flaws:

• Not as responsive as the specs indicate - I eventually rooted my Galaxy S in order to reduce the lagging to a tolerable level.
• GPS falls out after some uptime. This is supposedly fixed in the Froyo update.

That said, the Android platform is coming along nicely and I much prefer the Android openness to Apple's strategically founded App Store policies. I just wish you had more imagination. :)

He did not do much damage to the economy - he just scammed a lot of folks out of tens of billions over 30 years.

But to scam the public of trillions in 5 years, that takes CEOs in Wall Street and The City.

It would be useful if the phone informed the caller that you were angry and offered the caller to go straight to the answerphone. I'm not envisioning this to be mandatory functionality, but I'd prefer getting such a message upon calling someone rather than being barked at. In real life, you can look at a person and observe their mood before engaging them in conversation. This sort of technology could/should be a matter of lowering the threshold for giving someone a call. After all, most of us prefer not to inconvenience someone, so there is a reluctance to call someone in case they're busy. Or in a bad mood.

Anyways: Application of technology matters. After all, Windows Mobile and Palm were around for a decade and got nowhere serious until Apple entered the arena and applied the existing technology in a better way. Good technology engineers is not sufficient - you also need good product engineers.

I'm thinking the lists should be called "reasons why people of different cultures remain single".

Unfortunately, I'm wrong.

You are right regarding video playback. However, if the CPU is set to do "real work" for 1080p displays, then adding a core is a good thing.

Also, this trend of ever faster ARM platforms is interesting for the netbook market. Intel Atom is probably consciously crippled in order to avoid it eating market share from their more profitable "upmarket" processors. However, these ARM producers are indeed competing to make their processors ever faster. On a two-year perspective, Intel may find it's lowend offerings being threatened by new platforms. The iPad is just the beginning.