fabricate: way better than make. Well, that isn’t saying much.

I had a lot of plans for all the work I was going to do today getting ready for my upcoming travels. Instead I am just trying to survive the heat: I don’t know what the temperature is, since it’s beyond the upper limit of my thermometer, but it is Definitely Not Fun.

I celebrated my birthday yesterday with a backyard barbecue at Sunrise. It was fun; people started showing up around six-thirty and the last few left at two. We drank beer, margaritas, and even some champagne; lazed around on blankets, sat around a fire, talked, caught up, enjoyed life. It was comfortable and fun, and I felt happy and grateful for the company of so many good friends.

Inspired by a backyard party at Divide & Jacqueline’s house last year, I decided to install a firepit. People will happily sit around a fire for hours doing nothing but talk and relax; a fire turns any random patch of ground into the center of a social space. So I dug out part of our sloping back yard, packed the dirt level, and built a three-foot circular brick platform, on which I placed one of those metal fire dish things. It was hard, dirty, sweaty labor – I can’t remember when I last got myself covered in so much dirt! It always feels good to build something tangible, though. And, sure enough, once we set up a bunch of folding chairs, people sat around the fire til the end of the evening, talking and drinking and laughing and enjoying the night.

Today’s project, after cleaning up from last night, was a work party for the groovik’s cube. I took my sewing machine over to G-haus and sewed strips of velcro onto white ripstop squares (the faces of the cube) and black lycra strips (covering the gaps between the faces). It was a lot of work – we have about half of the lycra strips done now, but only seven of the fifty-four ripstop panels. There will be another couple of days at it for sure. The project is going well overall, though – the power boards are all done as of yesterday, which I’d been worried about. It’s going to work.

I’m leaving Wednesday morning for a big trip to California. I’m going to ride my FJ600 down the coast to San Francisco, then hang a left on Friday and take I-80 to Sacramento. Saturday morning, my brother Andrew will fly me and his friend DJ up to Bishop, where we’ll pick up a car and drive to a trailhead, from which we’ll hike up to meet the John Muir Trail. Later that day, or the next, we’ll meet my mother, who’s been hiking down the trail for a few weeks now along with my four youngest siblings. The three of us will unload our packs to resupply theirs; then Andrew and DJ will hike back home, while I’ll continue along for the last week of the trail. At the end, we’ll climb Mt. Whitney – 14,500 feet, the tallest peak in California – and then trek back down to civilization. A day or two to rest, and I’ll get back on the bike and head home, probably the faster I-5 route this time.

I can’t wait.

Hello, world.

This tutorial describes the implementation of a simple compiler using LLVM, an increasingly robust code generation library which is rapidly becoming the obvious solution for any compiler targeting x86.

Fascinating long-form email post called my evolution as a programmer: mostly about the author’s progress through programming concepts and the process by which he developed abstract thinking about software development.

It has been a challenging year, and I feel drained. I’m glad to have friends who can support me through rough times, but it’s time I got my life back to a more peaceful state; time for a change in focus.

I moved to a new house last weekend. It’s a pretty, comfortable, welcoming little place, a 1910-era Craftsman on the eastern side of Capitol Hill – not far from where I’ve been living, but closer to shops and friends and activity. Cat moved in a couple of weeks ago, and Sam will be joining us at the end of the month; the house is maybe half the size of the one I’ve been living in, but there’s plenty of room for three.

The upper floor of this house was designed as a master suite; it’s a sprawling big room with two closets and funky ceiling angles. I’m going to cozy off a little nook to use as a bedroom and turn the rest into an art studio. I feel good when I make things, even more so when I can share them with people who will enjoy them, so I want to arrange my living space around that. I imagine a “clean” work table for sewing, a “dirty” workbench for electronics and general fabrication, a desk with computer and electronic music hardware, cabinets full of tools, a whole closet full of materials and stored projects…

I’ve also given notice at Microsoft. This job has been a frustrating, difficult experience from beginning to end; I kept on struggling through, at first because I thought it was just an unusually difficult adaptation period, then because I couldn’t imagine where else my career could go, and ultimately because I simply didn’t want to abandon an unfinished project. The project is almost over now, and I still don’t know exactly what my next career step will be, but I’ve seen enough to be certain that this job will never offer the kind of work I need – nor, it is clear, are my skills or temperament particularly well suited to the group’s needs. I’ll spend the next couple of weeks wrapping up a few last bugs, writing some design documents, and passing on what little there is to know about the feature I’ve been grinding my way through – then I’m out of here, glad to be done with the Eastside and ready to enjoy the summer.

I don’t know what I’ll do next, but there are a handful of options in the air already, and I have a good long while to figure it out. I’d like to find something a bit more low-key – perhaps another telecommuting position – something I can do well without having to pour all of my free time into it. In the meantime I will probably spend a fair bit of time working on the Groovik’s Cube and riding around on my motorcycle, enjoying the sunshine.

I went in to work late today so I could pick up my FJ600 as soon as Felony Flyers opened. I dropped it off just before going to California, hoping it’d be finished by the time I got back, but finding a valve cover gasket for a 25-year-old bike apparently takes a while. The bike is grubby-looking as ever, underside all coated in oil and road grime, but for the first time ever it left no drips on the floor when I rode away. I’ve been topping up the oil every month or two since I got it, which is kind of embarrassing, but I’ve finally worked my way far enough down the to-do list that the leaks became the top priority.

While he was at it, Steg also tightened up the chain, synced the carbs, adjusted the valves, replaced the brake pads & fluid, and even wrapped some insulation around the exhaust joint underneath the peg, apparently just because he felt like it. Unfortunately the carburetor still isn’t set quite right; the bike is running smoother and quieter than ever, but it’s missing some power, and has a weird new tendency to hover around 2500 rpm for 20-30 seconds before dropping down to idle. Ah, well, it’s perfectly rideable, and I’ll bring it by again next week to have him readjust it.

I’m really happy to have this work done, since I’ve now reached the end of my mechanical to-do list. There will always be more maintenance work, and the bike can certainly use plenty of cleaning and cosmetic work, but aside from the carb adjustment issue it is now in perfect mechanical condition.

This bike was only supposed to be a temporary commuter while I got the Maxim fixed up, but a year of 100-200 mile weeks has made me rather fond of it. It’s not the sexy beast I was originally looking for, but it’s fast, nimble, reliable, and even graceful, in its chunky mid-80s way. It’s a big old slug by modern sportbike standards, but what they called a sportbike in ’84 actually comes off pretty well in ’09 if you think of it as a light touring bike. I might still upgrade to something like a Honda 919 or a Speed Triple next season, but for now my plan is to clean it up, make it pretty, and keep on riding.

I have a pretty clear cosmetic plan now, actually. The unique taillight lens shape is a key part of the FJ600’s look, so I want to preserve it, but I want to get rid of the side blinkers, and I am tired of dealing with incandescent bulbs. Sometime this fall I want to replace the whole taillight assembly with a custom LED board, incorporating wide-angle red LEDs for running lights, high-intensity 20-degree red LEDs for brake lights, and two strips of yellow LEDs for turn signals. I will probably even grind off the existing turn signal mounts, just to make the body line that much sleeker; the new turn signals will be hidden behind a red lens, so you’ll only see them when I’m actually using them, but the yellow color should still punch through clearly, and they should be substantially more noticeable than the current system.

I’ll need to repaint the fairing after I repair its various cracks and scrapes, but instead of duplicating the existing color scheme, I want to simplify it. I am definitely going to give the engine and exhaust a nice clean coat of flat black; but I just can’t see this bike as a rat or streetfighter. Instead of flat black I plan to strip off the 80s-fabulous decals and redo the silver & red bodywork in semi-flat grey, then give it a clearcoat. It’ll look a little cleaner, a bit more functional, but it’s really more of a happy bike than a bad-ass, so I want to keep it light. Buff up the aluminum accents, replace the windshield with some new lexan, and I think it’ll be great.

The Groovik’s Cube project has a web site now. It is still a big ambitious engineering project with significant unknowns, but we continue to make progress and the design looks increasingly solid. All we need are more money for parts and more people to help assemble them.

I am working on a project for the silent auction at next week’s fundraiser: a new microcontroller-equipped light-water-pack, running the same dimmer board software used in the Groovik’s Cube. The dimmers have a demo mode which runs until the central computer takes control; the backpack will simply run the demo forever, fading in and out through slowly evolving color cycles, on four independent channels. I like the idea that someone can help support the cube project by bidding on an object built around a piece of its technology.

The groovik’s cube project is coming along well. Chris, Brian, Barry, and I spent a couple hours discussing the software yesterday. We reviewed all of the existing code, talked about each module’s constraints and expectations, and kicked around ideas about how we might solve problems we think we might encounter. It’s in pretty good shape: we have working prototypes for every component, and the system works when we plug all the components together.

I’m not sure how many hours I spent on the dimmer code this weekend, but it felt like a lot. I’m on my fourth rewrite; I’ve figured out how to set up an interrupt-driven timer, and how to do fast writes by updating the IO registers. This is an enormously compelling project, constantly tickling those learning-driven pleasure centers in my brain. This problem is really all about timing, and the precision thereof; in order to render the full range of 24-bit color without flickering, the dimmer code has to be able to emit a light pulse of exactly one-eighth of a microsecond. This is at least theoretically possible with the 16 MHz microcontrollers we are using, but it is going to require some extremely tight code. I have not had to think about performance at this level of detail in years, over a decade at least; it’s kind of a guilty pleasure, and I am thoroughly enjoying the opportunity.

Understanding Engineers: Feasibility. Funny-because-it’s-true definitions of terms like “trivial”, “hard”, and “unfeasible”.

Since I bought my FJ600 last July, I have replaced the battery, regulator, turn signals (all four, twice), clutch cable, throttle cable, speedometer cable, front fender, both tires, chain and both sprockets, rubber peg covers, rubber shifter cover, left-side body panel, ignition key, gauge cluster, rear brake pedal, spark plugs, and muffler. The only non-cosmetic problems left are a minor oil leak in the valve cover gasket and another in the oil filter assembly. This is excellent condition for a motorcycle made in 1984!

The bike’s sluggish performance on my recent weekend trip to Eastern Washington left me thinking it was time for something newer. Now that the new muffler has smoothed out the engine’s power issues, however, I no longer feel so much like I’ve outgrown it. Maybe it’s time to clean it all up and give it a paint job.

My bike came with a beat-up little slip-on muffler which had never been adequate for the size of its engine and had in any case long moved past its prime. I looked into replacing it, and the custom 4-1 header, with stock parts from eBay, but costs were daunting. I finally bought a muffler up at Bent Bike, stock part from a wrecked ’94 FZR 600; same size engine and reasonably similar geometry, so I figured it ought to work.

This weekend I finally made time to drop the bike off at Felony Flyers; three hours later, Steg had fabricated a pipe adapter, welded it on, bent and cut some flat stock into a bracket, and bolted everything on. Steg is great like that.

Now that it’s done I wish I’d made this project more of a priority. It had never occurred to me that the distinctly frustrating dead spot in the bike’s power curve right around 6000 rpm was not some flaw in the carburetor tuning or some consequence of the engine’s age, but simply caused by the bad muffler! Why did I put up with this for so long? Same bike, same engine, now with smooth, quiet, powerful acceleration all the way up the range. No more need for all this jackrabbitting up to high RPMs just so I can downshift without bogging down…

What’s more, I hadn’t realized just how awesomely noisy my bike was. I am embarrassed, now, to have subjected friends and neighbors to that racket for so long – I simply hadn’t realized there was anything out of the ordinary.

I went by Danger House last night to look at the first full end-to-end test of the Groovik’s Cube hardware. The electronic components made a complicated sight, plugged together in a nest of wires and ziptied together onto a big pegboard; at the far end of the wiring, a scaffold cube supported one five-foot square of ripstop nylon, glowing brightly as an LED trio illuminated the interior reflecting surface. It’s fun to watch just by itself: with nine of these per face, fifteen feet on a side, suspended some eight or ten feet off the ground, it’s easy to imagine the finished cube being one of this year’s notable Black Rock City landmarks.

My part of the project is a set of twelve-channel Arduino-based dimmers. Red, green, and blue each count as one channel, so each dimmer manages four of the cube’s fifty-four facets. The central computer controls the dimmers via USB, and each channel’s output is a logic signal feeding into a power module. The power modules, in turn, govern the brightness of the LEDs. We could definitely solve this more cheaply if we delegated the PWM control to dedicated LED driver chips, but this solution is simple and does not require any custom circuit boards. We just buy a bunch of Arduinos, upload my driver software, and plug everything together. I’ll do some experimenting now that we have a working solution, but as much for my own interest as anything.

I haven’t gotten a good picture of the new LED taillight on my FJ600 yet, but it’s pretty bright. I can’t see it while I’m riding, of course, but it’s using 20-degree LEDs aimed just slightly above the horizontal, so it should do a good job of getting the attention of the driver behind me.

I will leave this unit installed, since it is a good auxiliary brake light, but the point was to prototype the full replacement taillights I want to build for Adam’s bike and my Maxim. It is clear that I will need a mixture of wide- and narrow-angle LEDs, perhaps a curved mounting surface, and possibly a diffusing screen.