Unfortunately, right out of the box there was a bit of a problem (aside from the fact that I had to locate 6 D batteries – who uses those anymore?) The gun would only fire in certain orientations – only if it was held sideways, or pointing down, or even upside down. It wasn’t consistent, either. Pulling the trigger wouldn’t do anything at all – no noise, no motion. I figured there was probably a switch that wasn’t getting pressed except when it was held in a certain way. I took the whole thing apart trying to figure out what was going on. There’s a lot of really fascinating little plastic parts and switches and linkages in there, and the mechanism for how it fires the darts is pretty ingenious, but there was nothing obviously wrong. Since there was nothing mechanically out of place, I broke out a multimeter and started testing electrical connections. That led directly to the battery compartment – there wasn’t any voltage on the terminals at all! Rotating the batteries around in space, there were certain orientations where a voltage would show up, but mostly it was dead. After popping out the batteries, I had a theory as to why.

For some reason, the little indentation for the battery contact is facing outwards, instead of inwards toward the battery. The battery has a similar protruding bit that’s actually sitting inside the contact instead of making a connection with it. The solution was pretty simple – a bit of copper wire bent and stuck in between the contact and the battery to make sure that there was a solid connection. You could probably get away with using a paperclip in there.

After that little fix, my Stampede always fires no matter how I’m holding it, and I have a reliable toy to battle opposing teams at work or just let my co-workers know that it’s time for the stand-up.

P.S.: If you want to build great software in a relaxed workplace and you can take the occasional Nerf dart in the back of the head, let me know.

Fortunately, it was pretty easy to bring XBList back to life. Not nearly as easy as I’d like – I’ve been poking along on a really major update to XBList that’ll be launched as XBList 4 at some point which will make handling this sort of thing much cleaner. XBList 4 is far from release, though, and I haven’t been putting that much time into it relative to my other projects. Still, XBList 3.3 is at least functional again, with some notable exceptions. Since some things work differently enough that XBList will need more major modifications, I’ve had to turn off several features temporarily, probably until XBList 4 is released. Until then, XBList will no longer show messages waiting for you, or let you manage your friends list (add/remove/confirm) from the app. On a more permanent note, the new Xbox.com simply doesn’t expose as much information as before. This means that I’m no longer able to show when friends are joinable or away. I guess they didn’t think that information was interesting for the website, or maybe they’ll be removing it from the dashboard too – regardless, I can’t get at it anymore. Otherwise, I think I like the new Xbox.com design – it’s much cleaner and simpler, though the huge non-personalized avatars-having-fun banners on the top of a lot of pages is a real waste.

I hate to release a new version without fixing some stuff and adding a new feature, and XBList 3.3 is no exception. I’ve fixed the broken Halo emblem support and added support for Halo: Reach emblems and linking directly to the Reach service record. I’ve removed the Halo 2 stuff since Halo 2 can’t even be played online anymore. I also fixed XBList so it can run under .NET 4.0 which should make it easier to install. Anyway, grab XBList 3.3 and get back to stalking your gamer friends!

Note that I’m also going to be shutting down my forums soon – they’ve always been kinda lame, and I’ve decided I prefer to just get email.

There are a number of ways out there for you to figure out how the cost of installation and purchase balance out with the cost of operation over time. You can always make your own Excel spreadsheet to figure it out, or you can use calculators like this one from energy.gov. However, all the web payback calculators I’ve seen have had clunky 90s interfaces, don’t take into account all the variables, and most importantly, don’t let you compare multiple types of heaters at the same time. So, like any good software developer, I built my own.

My water heater calculator is based on the same calculations used on the Federal Energy Management Program site, with the addition of inputs for your hot and cold water temperature. It’s also more flexible about how you enter your water usage. But the best part is that you can enter as many different water heaters as you want and they’ll all be graphed against each other, taking into account the lifetime of the unit. Get multiple bids, try different models, compare gas and electric. By displaying them as a graph of total cost over time, you can see where each heater breaks even with each other, and how much savings you’re getting by the end.

As a bonus, the calculator will also calculate how much you may be able to claim as part of the Energy Star Federal Tax Credit program. It’s smart enough to know the rules about the credits (gas heaters >e; 0.82 efficiency only, 30% of total cost up to $1500), and you can choose not to use the rebate if you’ve already used it up this year or don’t plan on applying it to your heater.

You can get started with the calculator by filling in values for your water usage and resource costs, or accept the defaults. Then add as many heaters as you like, entering in the cost for purchase and installation, the Energy Factor (which should be in the documentation for the heater), and the estimated lifetime of the heater. The more accurate you can make the numbers, the better your cost projection will be. Then check out the graph to see what your total expenditure will be after every year. If you’re comparing a new heater with the option of keeping your existing heater, just set the Cost to $0 and reduce the lifetime to how long you expect your existing heater to last.

Hopefully this little tool will be helpful to anyone else looking to replace their water heater. I filled it out for a combination of several electric, gas tanked, gas tankless, and heat-pump based water heaters, and it gave me a much better picture of what was worth it and what wasn’t. In the end, even though the graphs told me that the increased efficiency of a gas tankless heater wouldn’t ever pay back the cost difference versus an electric tank water heater, we ended up going with one. The promise of infinite hot water (long showers after a hike!) and no chance of burst water heaters outweighed the additional cost. But at least we were well-informed!

We decided to use the open-source Arduino microcontroller platform to drive LEDs inside the domes – that way we could have a bunch of independently-controlled lights and set their behavior with software. We chose the Boarduino Arduino clone from Adafruit Industries because it’s cheap, easy to assemble, and much smaller than the full-size Arduinos. Soldering it together only took an hour or so.

After that we connected a total of 11 superbright LEDs (ordered from DigiKey) to the Boarduino. Since the Boarduino only has 6 PWM pins (which can be used to “fade” LEDs in and out), we put 5 really bright LEDs on their own PWM pins (for the big domes) and wired the remaining LEDs (slightly less blindingly bright ones) in parallel to the 6th pin. The LEDs are unbelievably bright – even after covering them in an anti-static bag they are tough to look at directly.

At this point we had to sketch up some software to actually control the lights. Eva wanted a random, organic pulsing, so I started by having each light animate through 360 degrees and used trigonometric functions to create a smooth curve of lighting and fading. We tried a whole bunch of different speeds, patterns, brightnesses, and randomization (some different tests: 1 2 3 4) before settling on the final code. The code is a bit messy because of all the things that got changed around. I ended up using 1 – abs(sin(θ)) as the main brightness function, which gave the lights a sort of “breathing” effect.

The 0-1 values from that function got converted into a brightness from 0-255 for the PWM output. Actually, the brightnesses were always between a set minimum and maximum brightness, so they never quite go all the way out. Each cycle the speed of the fade gets randomly modified, so the lights never line up in any pattern – it’s pretty hypnotic to stare at.

After this Eva had the unenviable task of stuffing the whole works into the beast. She built little foam stoppers for each LED, and pushed one up into each dome. Then she carefully crammed all the wires inside, and the Boarduino, a switch, and the 9V battery. It ended up being way too cramped, resulting in a lot of broken wires, resoldering, and hot glue burns. Lesson learned – the next glowing beast will be bigger, with more open access to the inside.

The end result is really captivating. Eva ended up displaying it at Gallery Madeira in Tacoma, WA along with some of her other creatures. Since we both put a lot of personal attention the two of us put into the Glowback, and the fact that due to all the hairy wiring inside it’s sort of “high maintenance”, we decided to keep it for ourselves instead of offering it for sale. However, Eva’s not done with the idea of lit beasts containing microcontrollers.

Eva’s has written up a post on the Glowback from her perspective on her own blog – I suggest checking it out to get more detail on the concept and lineage of the piece.

I didn’t even use the nubbin, I just used its button. However, on my new ThinkPad, the center button switched the nubbin to scrolling mode, and turning that off in the driver just made the button do nothing! However, I recently stumbled upon the solution. If you completely uninstall the UltraNav driver, the middle button becomes a normal middle mouse button again, and the nubbin and trackpad still work. Tab management is easy again!