Politicians love to use the hot-button topic of oursourcing to pretend like this is a zero-sum game. i.e. a job outsourced is a job lost domestically. Americans have to choose: Do you want your companies to have access to cheap callcenters so you can grow your core business and create more jobs, or do you want to pay high local callcenter rates just so you can brag that "we buy american", while possibly stiffling growth.

I've found a solution of sorts for my pump problem. See Air Lift Pumps. There are already commercial windmill-powered systems for drawing well water so the problem is well understood. I've even seen people experiment with these in aquaponics using aquarium pumps (massively underpowered).

I now think that a cheap $50 diesel air compressor might be made to run using biodiesel produced from duckweed.

Thanks for all the comments and suggestions !

I'm currently running an AC waterpump capable of delivering 3000liters per hour at pump exit, and less than half of that at 1.5 m head height. This pump uses a whopping 40W of electricity.
I think that a windmill is an excellent idea, however since the wind can be rather fickle, I don't know how one would keep the nutrient-rich water flowing, and the fishtanks aerated.
An alternative is to have a biodiesel pump. There is a particular waterplant called "duckweed" which makes an excellent fish food, and also just so happens to have enormous potential as a biodiesel. Estimates are of delivering 200L of biodiesel from a modest planting of the stuff.

It certainly is a sticky problem and one which I've wrestled with for some time now.

This is also well worth a read :

How to Grow More Vegetables: Than You Ever Thought Possible on Less Land Than You Can Imagine - John Jeavons

I have recently started an aquaponics system at home. I'm African, but an expat living overseas. I am massively impressed with the potential for this particular technology to allow for microfarming on small tracts or even in your backyard.
Benefits I persieve so far:
      a) High yields over comparable soil-based techniques
      b) Allows for both protein and carbs to be sourced from one system
      c) Staples like corn have been successfully grown on *very* short cycles
      d) Small family-sized setups can be built to supplement a small family's needs or large "community systems" can be built to leverage economies of scale.
      e) Highly efficient water use compared to soil-based methods with only losses due to evaporation.
      f) Once it gets started the system is self-stabilising

Challenges I see:
    g) Technically not the easiest thing to get started
    h) Cycling the system to establish the nutrient and bacterial load can take up to a month
    i) First fish harvest can take up to 9 months (Tilapia)
    j) A typical flood-and-drain system needs a waterpump running 24/7 as well as potentially an airpump for the fishes. Electricity !?

I would be very much in favour of aid which goes toward establish self-sustaining community farms. I'm not a fan of aid which breeds dependency.

I was at the keynote for the devcon in Singapore today. I was surprised to note that the OS version seemed to jump from the up-and-coming 7.1 to 10 next year. Then while queuing up for the free playbook, we had to sign a "license agreement" for the 2.0 beta OS loaded on some of the devices.

On a sidenote the keynote and all other opening prezzies were delivered using a Macbook pro which had the back covered to obscure the apple logo. I would have liked it better, seeing how much they were loving the HTML 5 on stage, if they'd actually eaten their own dogfood and delivered an awesome HTML5 prezzie using the Playbook+HDMI

... with apologies to Hustle and Flow.

Actually, there is nothing wrong with the word Mashup. I've seen several people now complaining that yet another neologism has creeped into the already crowded web vernacular. It's just a "relatively" new word that means something "relatively" new. It's not even indicative of the web being overrun by the mindless masses and buzzword junkies, that happened long ago.

S.