serious

Good Good. Canada bodyswerves son-of-StarWars

The BBC reports that in spite of previously supporting the idea, Canadian PM Paul Martin "has reportedly decided that his country would not join a missile defence system being developed by the United States."

Tests of the system - which, remember, was supposed to have been officially deployed by now - have not been going well. After a two-year hiatus (following tests widely criticised as not even vaguely resembling real-world attacks) the most recent attempts at interception both resulted in the antimissiles failing to even leave the ground.
There's an exchange which concerns the Star Wars project mentioned in a Stephen Pinker book on language. It goes something like:

"What if the enemy develops anti-anti-missile-missile-missiles to shoot down our anti-missile missiles?"

"We will simply have to work on anti-anti-anti-missile-missile-missile-missiles."

You have to admire the forethought even if it is probably an urban legend.
I always admired him for including an analogy involving Smarties, with a footnote explaining that "For the benefit of North American readers, Smarties are like M&Ms, only better."
Huh? We *always* had Smarties in Canada. If anything, M&Ms were harder to come by.
Heard the story back in the early eighties, but with two contradictory locations: an American test out in Nevada or similar, and an RAF one in Germany. Can't find it on Google, but I believe in the American version the rocket was a "Sergeant York".
M247 DIVAD, apparently. And as an indicator of how accurate that version's likely to be, the Sgt York had dual machine guns rather than missiles. The other version's probably as reliable, mind.
Unlike our previous PM, however, Martin is capable of speaking out of both sides of his mouth, so though I'm pleased to hear this I also wouldn't be surprised to an about-face come along later.

Though as SoSW will likely be about as successful as its predecessor, I suppose it's really a moot point.
Tests of the system have been reported as completely successful by their developers. Admittedly, this appears to be because every time some awkward milestone such as "actually works" or "can fly in a straight line for more than a second" fails to be met, they scale back the tests until they reach the current level of testing which ensures little more than "when you hit the power switch, the red light comes on at least 50 percent of the time." Not bad for 80 billion dollars, eh?