When StarLight's budget outlook grew dim three years ago, there was only one option: Figure out a way to put one of the daughter telescopes on the mother ship.

But doing that would mean light gathered at the mother ship would already be at the final collection point, whereas light gathered at the one remaining daughter craft would have to travel to the mother ship.

For interferometry to work, the two beams of captured starlight must travel exactly the same distance before being combined. So the engineers created a little detour. By bouncing the light gathered by the mother ship back and forth several times, its effective travel distance could be extended.

"It's all done with mirrors," Blackwood says. House joke, of course.

Then two of the team members, Peter Gorham and Bill Faulkner, remembered their 10th-grade geometry. A parabola.

"Bingo. It was like a light went on inside someone's head," Blackwood recalls.

A parabola is shaped like a U, except that the upward-pointing lines spread forever apart. Light hitting any point on the U will be reflected to a common point above the bottom of the U, called the focus.

Now imagine a daughter spacecraft out on the left side of the U. As a wavefront of light reaches the daughter craft, a beam of light will be reflected and travel a known distance to the focus. Call it beam A.

Directly above the focus, a beam of light in that same wavefront will travel to the bottom of the U, called the vertex, and also be reflected back to the focus. Call this beam B.

The law governing parabolas dictates that beams A and B are identical in length when measured from the same wavefront crest.

So the StarLight team realized that they could put the mother craft at the focus of a parabola.

It would capture Path B, and then use mirrors to make up the known distance to the vertex and back. The result was that if the mother craft sits 7 meters above the focus, then the daughter craft could sit 600 meters away, along the parabola, and the baseline would be 125 meters.

"Now that's interesting," Blackwood says. "That was our breakthrough. And it saved the mission."

Blackwood is a Star Trek fan. Voyager. Like a lot of people at JPL, he grew up on a diet of science fiction, dreaming of other intelligent life forms, envisioning the discovery of that next pale blue dot.

Because it is not the money that lures people to JPL.

"They're here because of the vision thing," Blackwood said. "That's why I'm here."

But there's a catch. The vision thing is not achieved overnight. Nor is it realized by any one individual. And vision takes money, of which there is never enough by the time it trickles down to you and your project or department.

"I think a lot of people come to JPL thinking they can change the world quickly," Blackwood says. "And what we all learn when we're here is that there are only so many dollars and so much time. We can't just say 'hey this is a cool mission, let's fly it.' The mission's got to be exciting to the public, it's got to fit within the money that NASA has, and it's got to be interesting to NASA and to Congress."

And so Gary Blackwood and his colleagues are in a race against time and money to prove that their vision, their wish upon a fake star, is one that needs to soar and find real targets.

The team figures they'll be ready to launch in 2006. But before that happens, NASA headquarters has to come through with real funding, to the tune of $220 million for the two spacecraft, instruments and operations, plus $70 million for a launch vehicle, in 2003. The team would grow from 56 to around 200.

This year, the mission went through "a pretty serious set of reviews," Livesay says, "and it's looking pretty good for StarLight."

But NASA headquarters has a lot of competing missions to choose from. There's still one little planet on the outskirts of our own solar system, Pluto, that has yet to be visited by a spacecraft. Other scientists are eager to spend limited funds on exploring asteroids, comets, and other objects closer by. Every scientist at JPL knows they must fight for funding, or else their last mission may be their last.

And so it must gnaw on Blackwood's brain that the craft he has devoted four years to, and will work on at least into 2003, may never fly.

"I think we don't think about that. I don't."

But he does think about it, or at least now he has.

"I'll be disappointed if it doesn't fly," he says in his next breath. "But if this doesn't fly, I think we've already learned a lot about what a formation-flying telescope needs to be."