Unfortunately, Dr. McCleese, instead of describing the likely future program, spent his entire half-hour explaining in detail why NASA at the current time does not have the slightest idea what its future Mars program should be (except for the 2001 Mars Surveyor Orbiter) -- and won't know for some months.

The original hope was that NASA would announce the revised plan this summer, but McCleese told me that this fall is more likely.

The reason is very simple: Mars exploration has turned out to be much more difficult than NASA had optimistically hoped in the wake of Pathfinder's stunning success in 1997. Since then the failure of the two 1998 Mars Surveyors has changed everything. But even with these equally stunning failures, the program was headed for a radical revision, as over the past year it's become clear that NASA's hopes to return the first samples from Mars in 2008 for a total cost of about $400 million were wildly optimistic.

In an earlier interview, McCleese told SpaceDaily that the cost estimates for such a mission had soared to over $1 billion; at this conference he told an MSNBC reporter that he now regards a sample return mission as a "Cassini-level program" -- and the Cassini Saturn probe cost over $3 billion and took a decade of planning.

In retrospect, given the extraordinary complexity of such a mission, it's hard to understand how NASA had ever hoped otherwise; their current plan involves using a rover to collect samples over a range of a kilometer or more and then return them to a mother lander that would blast them into orbit around Mars with a two-stage solid rocket -- after which a separate French-built orbiter would rendezvous automatically with the baseball-sized sample container in Mars orbit, retrieve it, and launch itself out of Mars orbit and back to Earth.

McCleese says this plan is now open to complete review, and that NASA may ultimately use some other scheme, such as having the sample-return lander process Mars' air into rocket propellant to fuel a more powerful launch directly back to Earth - leaving France to find other way of spending its taxpayer's money exploring Mars.

But in any case, a sample return mission will require a lot of time and money -- and that raises another question: given NASA's limited Mars funding, should it immediately focus on developing a sample-return mission, or should it instead devote the next several Mars launch opportunities to using smaller orbiters and landers to carefully survey Mars to locate the best possible landing sites for sample-return missions?

After all, sample return flights will always be expensive and relatively few in number, and each one will return at most a few kg of samples -- which means that we may have to put a great deal of thought into where to land them to maximize the chance that any samples successfully returned contain evidence of ancient micro fossils.

According to McCleese, this question is currently bitterly dividing NASA's Mars science advisors -- and he rather plaintively asked the astrobiologists assembled at the Conference for their input on the question.

He said that in the past, NASA had traditionally decided early on about the overall form of a space exploration program, and then only asked scientists for their recommendations as to the details of experiments to fly on the selected missions -- but that in the case of Mars, this philosophy must change and the scientific community must play a central role from the beginning in designing the structure of the overall program.