Today is Good Friday, and everyone here took vacation, except me. I spent most of the day trying (and failing) to beat my CardBus problem into submission. I fixed a couple other minor problems with my kernel configuration (figured out why USB mass storage had unresolved symbols, for example) but merely changed the nature of the CardBus problem from an inability to reserve a resource to an inability to reset a socket. I believe that the only change between those two states was making the PCMCIA support modular instead of built in. I'll play with that more later. Most of the day I spent waiting for kernel recompiles, rebooting, fscking, and so forth. I never made it out the door of the house, though I at least left the window wide open for most of the day to try to make up for it.