From the New York Times, hints of a worst case scenario.

AGHDAD, Iraq, Aug. 10 — Riots over severe fuel shortages continued today in Basra, Iraq's second-largest city, as some officials who have studied the matter warned that fuel shortages could recur in other parts of the country.

United Nations officials said there was a "near certainty" that Iraq would face winter shortages of kerosene, a vital fuel for heating homes in northern Iraq, because of the same refinery problems that have led to the gasoline shortages.

To paraphrase General MacArthur, suppose the word comes to Washington, "Send me kerosene or send me bullets". Most of the people in danger of dying of cold aren't the same ones trying to kill U.S. troops - but we can't quite tell which ones are. Occasionally our kerosene does get shared with our soldiers in the form of molotav cocktails. How much will the president who prefers cutting taxes for the rich to helping America's hungry and homeless be willing to spend there? I know, I know, the tax cuts for the rich are supposed to create jobs. That's part of the point - money spent in Iraq isn't going to create many jobs in America.

It isn't helpful to repeatedly contend invading Iraq was a mistake - we have to talk about what we should do now. The first thing we should do is deemphasize hoping for the best and reemphasize preparing for the worst. We don't want to repeat that mistake again.

This is the worst dilemma I can think of right now. Even if we spend massive amounts of money helping ameliorate the suffering of the Iraqi people, it may only keep attacks on United States troops at the level they are at now. The violent wing would claim we were only partially moderating problems we ourselves had caused - even while they sabotage all attempts to rebuild Iraqi infrastructure. On the other hand, doing less could lead to widespread rioting such as MacArthur feared.

I've thought hard trying to come up with something more helpful, but the best I can do right now is to look more at worst case and less at best case scenarios.