Ron Cowen has a nice piece at Nature on the "long-standing but little-publicized software problems" of NASA's Fermi Gamma-ray Space Telescope. Since its launch in 2008, Cowen reports, the world's leading gamma-ray telescope has had only a clouded view of the highest-energy gamma rays.
A better view of these gamma rays, with energies greater than 10 billion electron volts, "could yield clues to dark matter and the powerful stellar explosions known as gamma-ray bursts," Cowen writes.
This might have been reported elsewhere before, but Cowen is the only one I could find on Google news who's reporting it this week. And the occasion for the story is that a workaround for the problem was uploaded to the space telescope two weeks ago, and it is now being tested. When this software comes fully online at the end of 2013, it will boost the data on the highest-energy gamma rays by 60%. One researcher tells Cowen this will mark "a complete renaissance in the science this instrument will do."
So what was the problem? Two years before launch, tests showed the instrument collected much more information than could be stored in its memory. One scientist calls that "a woefully stupid error," but one the researchers couldn't risk fixing before launch, because that could have delayed and possibly threatened the mission.
Check out Cowen's nice story to be reminded that the even smart guys sometimes screw up. It's reassuring.