Thanks to your question I've finally gotten around to doing something I've meant to do for a long time. I wanted to post sample output from ICE so that you could evaluate it. Since I'm now running linux, I used "dosbox" (an emulator under linux) to run ICE and output a file for the moon in May '07 for an observatory near my location. I pulled the resulting ascii file into openoffice so that I could generate a .pdf file that would hold formatting on a post to APUG. Here is the output you'll get for 31 days of lunar rise and set times starting May 1, 2007. Note that there are some times that are preceded by "f" or "p". This indicates "following" or "previous" day, since each rise/set pairing of the moon can straddle a solar day. The azimuth (compass bearing) for rise and set and the maximum altitude in degrees is in parentheses beside the times for rise/set and transit (crossing a line from N to S through your local zenith).
You can also run the numbers for the moon in tiny increments during a single day to give you alt/az for any time as it crosses the sky (in case you want to calculate exactly where it will be relative to an object in your scene). Be sure to adjust for local magnetic deviation when using a compass to predict the location of the sun or moon. In my location it's currently off by 6.5 degrees to the west of due north.
And for the linux users here, you now can put ICE to use for yourselves.
DOSbox is also available for OS-X users, and the ICE program is so simple I'd expect it to work very well there as well.
I know this isn't as simple as some might like, but it works, it's free, is very accurate, and you can print as far in advance as you like and use it without toting along a computer.
Lee