stuffduff's Journal: What alternatives to RDBMS are Slashdoters using & why?

From my first experience with Fortran IV data blocks in the 60's to CouchDB I've seen a lot of methods to handle data. But for some reason, the appearance, in this day and age, is that the RDBMS is the 'only' suitable method for handling large volumes of data. If you've been through the mill and decided not to use an RDBMS I'd like to know what alternative method you've chosen, what benefits and drawbacks made you look for alternatives, why you chose an alternative method, and last but not least, how the choice is working out for you. Please include a short description of your primary development environment, how your decision has developed into a solution, and if there is one, what is the one thing about it that you could change if you could?