Apart from everyone listed here.

Trouble there is that you then have to deal with the shortcomings of the X Window system (just try talking to Jonathan Blow about that one) and then anticipating anything the user might have activated on the other end that might conflict with it. In theory, it's easy: in practice, it's highly complex to get it working on all Linux/BSD systems due to fragmented and variable configuration.

[citation needed]

So it doesn't work if:

• the software is stolen (yes, I know this definition of piracy is moot, but still)
• the OS you're using is ten years old and will cease to be supported in July

So that's still the vast majority of users covered for free. Sounds like a pretty good deal to me.

True, the App Store model is monopolistic and overall has rather nasty side effects, but the rejected description stated that the app had been a finalist in Android's developer challenge.

This information was simply irrelevant, because the Android app is a completely different beast to the iPhone OS one. It's like Microsoft advertising Office 2007 by saying that Office 5.1a for the Mac won an award, somewhere. True, it was probably inadvertent, but if I was browsing the synaptic repositories and came across the Skype package, I wouldn't want to know about how well it runs on AmigaOS.

This is a very flawed way of thinking, because it fails to take into account the fact that expectations change. User requirements will change, as will operating environments, user expectations, communications protocols, hardware standards, standard formats, user interface standards, etc. etc.

In essence, you're wrong because as far as software goes, anything that plateaus and stands still will stagnate—and Microsoft are smart enough not to let that happen.

In fact, OpenOffice.org, in its present form, is pretty poor tech (there are many reasons for this, which I shan't quote here for the sake of brevity.)

BorgOffice is superior in practically every technical aspect. However, as Microsoft knows all too well, if the price is right, the sheep will flock to it, even if it is complete and utter shite.

You definitely don't want grade crossings (or level crossings, or whatever you want to call them): then you get crashes like this http://en.wikipedia.org/wiki/Ufton_Nervet_rail_crash and hair-raisingly near misses like this http://www.youtube.com/watch?v=4nggx7yklaA.

As I understand it, it's not permitted to build new road/rail level crossings in the UK (and probably the rest of the EU, too) because they are simply too dangerous.

That doesn't really count: (a) Transrapid is a German company, and (b) these are maglev trains. These require specialised (and very expensive) rails, and "normal" wheel-on-rail rolling stock can't use maglev rails.

Of course they're going to push for replacement. But they're still good enough to maintain the old product, which gives the old product a longer lifespan, therefore another selling point. QED.

So not at all like the ~300mB hunk of Qt libraries I have to download if I want to get, say, LyX running on my Ubuntu system?

Citation, please?

Not taking sides on the actual case of not permitting laptops in the cinema, my local establishment (the Camberley Vue) has lockers, I believe, so that customers can dump their bags before entering. Maybe Cineworld would do well to implement a similar scheme, which would make the idea of not being permitted to take your notebook into the cinema far more palatable.