Is mining worth it
Returns to mining should be in the aggregate near zero over the long term, your mileage may vary depending on your incremental electrical cost of mining. The energy efficiency of mining from least to most efficient generally goes: CPU mining, GPU mining, FPGA mining.
The most tenuous aspect is whether or not you will be able to recover an initial hardware investment, due to both the expected near-zero returns over time and that any payback period could quite possibly exceed the lifetime of bitcoins actually being of value. Of course, lots of folks will tell you the latter will *never* happen.
If your only cost is the additional electricity to mine on a video card in a system that is left on anyway ( say a 7x24 firewall box ) your only initial investment would be your time setting up the software. Beyond that point any outlay for hardware should probably incorporate FPGA hardware, which is unlikely to ever be of any use for anything else. I suppose if you were to buy a video card for gaming you could also choose one that could be used for mining.
Setting up a CPU miner program is easiest, cheapest in initial outlay, and would give you a feel for what takes place.
Some people might tell you mining is a sucker's game since the return to mining is rigged to begin with by the difficulty level, which is adjusted every 2016 blocks to result in one block being solved every ten minutes or so, thus the difficulty changes every two weeks or so. The more capacity ( hardware and electricity ) that is thrown at mining, the less the return per unit of mining capacity. From this one might conclude the best way to mine is not to mine, although there are evidently big galvanized steel tubs of Kool Aid from which you are encouraged to drink in order to help you believe otherwise