Since I love building decks, I decided to build another deck.
When I bought my house in November of '01 there was this strange, square, 16' x 16' gravel pit, surrounded by a flat, 14" wide concrete enclosure in my backyard, more-or-less abutting the house. Perhaps some previous owners had planned to put in a patio? It was situated such that it divided the side yard from the back yard and I could never figure out what to do with it.
Earlier this year I started to dig it out. I was going to put something there! But I wasn't sure what. Originally I tossed around the idea of putting in another patio or pouring a concrete slab. I even entertained the idea of pouring my own concrete blocks made to look like Scrabble tiles. Just thinking of all that expensive and heavy concrete put me off the project for a couple months, though. It wasn't until it occurred to me to simply build a second deck on top of the space that things finally got under way.
I decided to build it just like I built my initial deck: pre-formed concrete footings supporting 2x6 pressure treated pine joists with 16' 2x6 boards for decking. Because I was using pre-formed concrete footings I wasn't going to have to dig any holes or pour any wet concrete. I also wouldn't need to grade, level, fill or tamp the ground especially well. There was a bit of prep work I would have to do before I could start placing the footings, though.
At some point in the past a chain-link fence completely encircled the yard. Most of the fence was still there when I bought the house, except that some thirty feet had been removed for who-knows-what reason. Consequently, a twenty-foot, isolated segment of this old fence remained in the corner where I was going to build my deck. I decided to completely remove it, concrete footings and all.
Once the ground was clear it was time to position the footings. This was the hardest part because it was hot and those things are heavy and I'm very lazy. I started by tossing them roughly where I thought they should go, forming an evenly spaced 4x4 grid, then I started measuring. I decided to cantilever the edges a foot so beyond the footings (due to the fact that I couldn't get the footings under the corners, anyway, because of the existing concrete ring).
I picked the footing nearest to the house and made a command decision as to the height of the footing: it had to be high enough so that all the joists would be above ground level but as low as possible (since the goal was to make a low deck that didn't require any stairs). Since I had dug up the top few inches of ground and hadn't bothered tamping it down I also decided to recess the footings into the dirt (so that they were sitting on the hard packed soil). The first footing was simple, because its height wasn't dependent on any other footing. The others became progressively more difficult.
For each subsequent footing I had to position it so that it was level with all the previous footings. This wasn't too difficult at first. I worked, footing by footing, moving away from the house. Within a couple of rows, though, the grading of the ground presented a problem. With the first several footings I had to remove soil to recess them into the ground to get them level; but at some point I found I was having to elevate the footings to maintain levelness. I believe I went back to the local nursery three times to buy more leveling sand (you will always underestimate the amount of leveling sand you'll need!). For the final concrete footing I found myself sifting gravel out of the ground in order to build a sufficiently high and incompressible mound.
There were some points where I determined that I was working with a four-dimensional manifold. Footings A and B were level relative to one another; footings B and C were also relatively level; but then I would find footing A was an inch higher than footing C. There was one day where I had been at it so long: raising this footing - lowering that one - that I finally gave up and decided to try my luck the next day, hoping that normal, three-dimensional space would again present itself.
Eventually all of the footings were level with one another (within an acceptable margin of error) and I was able to attach the rim joists. The main reason I spent so much time leveling the footings was that I had decided to build this deck without vertical posts (in order to minimize height). Normally it doesn't matter if the footings are level, because you can much more easily adjust the height of the posts in order to make your joists level. Since my joists were to sit directly on the footings, they had to be as level as possible or the surface of the deck would be irritatingly, noticeably askew. Which, if I were building a Tilt-A-Whirl, would be fine... now I kinda want to build a Tilt-A-Whirl.