Its a bit sad that any progress comes with a cost and the cost of the various space programs is cluttered satellite orbits and large pile of junk in the space. Hope some one will start thinking about cleaning some of the mess.
@Krisi, @Larry and others: Worden is a huge promoter of nano sats. He noted some sat contractors are still resisting the move preferring to sell a few big $100M giants than trying to sell dozens of $1M nanos, but methinks the times are changing for many (though not all) the sats.
Even so, the same argument applies. Previously this type of system failure was catastrophic, because you could only afford to send up a system once. A generation of satellites can be tried, the data on their deterioration can be collected, and an improved generation can replace it. This cycle can be run many times within a fraction of the cost envelope of the budgets of prior systems.
Interesting perspetive Larry...but without rad hard design the electronics will slowly but surely drift our of spec...this is not the question of glitches, it will just stop working, period, due to the damage to silicon lattice, this is not reversable
As the price goes down for putting up satellites the risk premium for the onboard electronics goes down as well. NASA has been experimenting with using cellular electronics instead of the ultra rad-hardened but ancient and expensive chips that have been used in the past. Likewise, a number of student and hobbyist groups are designing micro and nano satellites around cheap commercial electronics. If something glitches then you try something else the next time. A single failure doesn't cost megamillions of dollars, so people are taking more chances. This is definitely getting exciting.