Most of the petrol station was destroyed. It was amazing to see so many people run so fast...

OK, so I know from you calling it "petrol" that this wasn't the US, but...they don't have required fire suppression systems in your country?

Here in the US, every gas station has to have an automatic fire suppression system. When they let go, it's very, very impressive...

They've been hammering at this for years, and we have yet to see anything more than a jacket with buttons for your MP3 player enter the market.

What's the obsession?

They're a bit late to this game, Amazon is pretty entrenched with Kindles already in most people's hands, and at prices which are far more approachable than iPads.

What's entrenched: Apple, with millions of iPhones, iPod Touches, and iPads in the hands of teenagers and college students.

That's bullshit. You're only adding 10% of a *FUEL*. If you added 10% water, and it still ran, you'd expect an approx. 10% loss in efficiency. You could mix in kitchen oil (which will burn) and if you could get it past the injectors, you wouldn't expect a loss anywhere near that.

NO, you're the one full of bullshit. You're operating on the incorrect assumption that the only (or worst) effect a contaminant will have is to not burn. Stoichiometric ratio changes, burn speed (flame front speed) changes, etc.

Ethanol has a completely different stochiometric ratio from gasoline; it's more like 9.7:1 for E85, versus 14:1 for gasoline. That 10% ethanol requires twice as much oxygen to burn than the gasoline it replaced.

Ignition timing is based off a lot of factors to provide ideal burn, because it's a BURN, not an explosion (that's called detonation, and it cracks/blows bits of your engine when it happens.) A flame front travels from the spark plug outwards in a designed way, and it takes time to do that - it's not an insignificant amount of time relative to motion of the engine, especially at higher RPMs. Depending on the mixture, temperature of the gas/fuel mix, engine speed, and more - the engine computer decides when to fire the spark so that the burn is appropriately timed. When the burn is timed can dramatically affect torque generated and the kinds of emissions produced, because the pressure in the combustion chamber is always changing. A fuel mixture burned at one pressure burns differently from another - different temperatures, and thus different kinds of emissions output.

There's more. Rich mixtures burn slower and cooler; lean mixtures burn faster and hotter. Slower burns are less efficient, faster burns moreso. However, lean mixtures tend to blow/melt things, so everyone tries to avoid lean running if at all possible. Flame front speed will be dramatically affected by contaminants and additives.

If you put 10% cooking oil in your car's tank and managed to get them into a homogenous mix, you'd be lucky if the car started at all. If it did, the fouling of the spark plugs, valves, and catalytic converter would take minutes, if that.

Water? Well, aside from the fact that water and gasoline literally don't mix: the water would cause almost instantaneous rusting of the fuel pump, fuel pressure regulator, and fuel injector pintles.

Cold air is denser, and causes the engine to run richer, i.e., inject more fuel into the engine. This gives you a bit more power, but at the expense of fuel efficiency.

This is complete bullshit written by someone who has no idea how engine fuel systems work. Any fuel-injected vehicle sold in the last 20+ years uses a mass airflow sensor which provides the correct amount of fuel, no matter the ambient temperature or pressure. There are various styles, but the most common is a hot-wire based sensor. Porsche and others used a vane/flap-based sensor in the 80s before switching to hot-wire sensors. Mechanical fuel injection systems used a sensor plate linked to a metering valve.

Further, EFI is closed-loop because of the O2 sensors - O2 sensors have been in cars since the 70's. In vehicles made since around the mid 90s there are two; one before the catalytic converter, and one after. The sensor detects the amount of unburned oxygen in the exhaust, and thus the fuel mixture ratio. The engine computer modifies the mixture based on the sensor's output; computers made starting the very early 90's kept track of those measurements to adapt to air leaks and whatnot over time.

We'd be getting jokes about bestiality, tales of practical jokes involving other people's bicycles, lewd comments about hot people at the bar, and oddly-endearingly-angry rants about how the world would be a more peaceful place if everyone smoked pot and got laid once in a while.

Be sure you have enough RAM or you're going to be in for a heck of a surprise. 2GB of dedicated RAM per TB of disk usage is recommended as a rule of thumb. I found this out the hard way when it was new.

It's worse than that. Even with plenty of RAM, I had a ZFS pool of a couple TB and dedupe turned on. Turns out virtually none of the data in the pool was duplicate, and disk access slowed to a near complete crawl because the machine was going through an enormous dedupe table.

Took us months to recover after turning dedupe off - fortunately, the backup software we use periodically re-processes its archives, and in doing so, re-writes them, which removed the data from the dedupe tables.

...which is why there's a soft-metal tab that attaches them to the frame, called a derailleur hanger. It doesn't take much at all to bend them, and the derailleur is also pretty fragile, in general. The more gears you have on the rear cassette, the more precise everything has to be. Having the chain exposed like that, and using a derailleur/cassette, is pretty stupid. Ask anyone who commutes in the winter; it's all going to clog up and stop working. They should have gone with an internally-geared hub (with suitable oil for the temperatures) and a full chaincase covering the chain. She's going to have issues with snow and ice getting everywhere, in the cables and more. Won't be much she can do about it, either, at those temperatures.

I don't know how small your library is but if it's large enough to warrant a card catalog then I'd suggest first putting all the books in the correct order and making sure the card catalog is accurate.

Speaking as someone who volunteered for several library projects - that's not how you convert a library over. For your database, you process books as they go through circulation. Ie, book returned? It goes into the "enter and re-shelve" pile. Or, alternatively, you at least initially enter the book when it's checked out so you're tracking it with the new system.

This prioritizes the most popular books, assures the highest & quickest conversion in terms of transaction volume, and thus the quickest efficiency jump (which frees up time for dealing with the other books.) Otherwise, you'll spend years dealing with books mostly not entered into the system.

You can deal with low-circulation-volume books by pulling them shelf-at-a-time. If the library is small enough, simply re-shelve them in a new location. If the library is more than a few shelves, then tag the books in some fashion (the various color dots and whatnot you used to see on books in various libraries were, in fact, how the library was tracking this sort of thing.

You'll thank me in the long run when you're not stuck with a million lenses for a camera you've outgrown.

Uh - all the major mounting systems (Canon, Nikon, four thirds, Leica) are not going away soon. They'll work just fine with newer cameras. There is a little risk with the APS-C lenses (Canon calls them "EF-S", Nikon "DX"), but both companies sell buckets more APS-C cameras than they do pro cameras which have larger sensors. They're not going away any time soon.

What you should NOT do is buy an emerging, unestablished lens mount system, like the Nikon 1. Also, if you have a large collection of old Nikkor AF lenses (or anticipate wanting to collect old, probably overrated and outdated lenses) that need a camera motor drive, some cheap Nikon dSLRs don't have that.

In terms of "outgrowing":

• EVF cameras are poor for shooting anything even slightly moving because of lag, and don't have good dynamic range (ie, bright areas blow out, dark areas are hard to see anything.) Even Sony's newest A77, which was lauded by reviewers for having a "great" EVF, is getting complaints about these problems. Driving the EVF also means the sensor is powered up and generating heat all the time, which increases noise.
• LCD-panel (no viewfinder) cameras suck even more for anything even slightly moving.
• The cheaper dSLR models usually cheap out on controls, and sometimes functionality. Most famously, the _0D series (serious amateur to most working photographers) from Canon had a back thumbwheel and index-finger wheel, whereas the Rebel (the "prosumer" line) had buttons. The thumbwheel usually is used for exposure compensation, which once you get shooting more, you use constantly. The thumbwheel works great; the buttons are a royal pain. Nikon's EV adjustment procedure on some of their cameras REALLY blows; you have to hold a button AND use a wheel.
• Don't pay much attention to video abilities. They pretty much all suck in lots of different ways, from rolling shutter issues to lets-protect-our-video-camera-market recording length limits, to crappy codecs, to poor focusing, and so on.
• When you start shooting for commercial purposes and your clients are reproducing your work on the sides of buildings, by all means, consider full-frame. Until then, go APS-C/DX like everyone else.