Author: Peter Jones

I have a lot of things I’m working on, but very little work done. I am a very inefficient simple machine.
I do have some visualizations now for my box solver, including how to visualize the answer when done. I think it will be even better looking when I add in a visualization of each part going in. Oohhh, or I could add a stepping series if I go with the randomization method. As long as I manualy increment the seed and I keep the seed when the series is done. And by manually, I mean I set the seed incrementally.

Last night I almost ordered a new brew pot and burner, but the website was being a butt. That purchase has already been vetted and is happening. I might look local because I’d really like to get brewing for Thanksgiving.
The Wif keeps on being amused when I call six-dollar teddy Nigel. Especially as I usually follow up with, because I keep making plans for him.
Among with sensors I have a could LiPos showing up on Tuesday, so that will mean it’s time to build a WifBit.

I think my initial array of pieces should also have each of the translations, and a key for removing it’s siblings from the pool when used.
Probably I should do this in Processing so I can easily see what I’m doing and get a visual framework for the output. A counter for the number of steps would be nice as well. Figure a line after every line that iterates a counter. That versus genetic would be fun to see.

Of course, rather than get working on the challenge I thought of a few more solver options. Then again, it’s a lot easier thinking of options wherever I am than trying to implement them on pen-and-paper.
Same array of pieces, same fail situation, this time rather than methodically, use a methodically seeded RNG. Than again, I could also just start with the array of pieces being already jumbled.
Learn something like Unity or use Gary’s Mod to implement the puzzle with a funnel on the top and an agitator on the bottom.
Do the same, but in real life.
And of course, learn how to make a genetic algorithm thing.
I like the idea of the RNG, and the shaker. The shaker in real life sounds like the most fun. The algorithm is probably the best education experience.
All because I don’t want to have to solve it by hand.

One of my preferred coffee shops, Ada, has monthly puzzle problem handouts (with the answer giving a discount or whatever, as I consider the following to not be in the spirit, this a moot point). Every time I see one of these, I pick of up and while I might not figure out the answer, I like to figure out how to figure out the answer. The last time I was there, the day before my birthday, I picked up, read as bought, a little wood puzzle. This has me thinking, a good way to practice doing things is by doing them. Duh. Five thousand hours is the magic number for mastering whatever after all.
So it’s a, if my memory serves, five by five cube, with a dozen plus pieces in a couple different rectangular sizes, that all fit together.
I figure this could be brute forced. Start with an array of pieces that contains a coding for the equivalent pieces. Then add a three dimensional array for the cube itself. Starting with biggest single dimension pieces first, put them in the matrix at a given orientation. If a piece dimension orientation results in over running the cube, skip the rest of the same piece/orientation combinations. Add another piece. One more matrix of pieces and orientation in the order placed and done.
I should write this.

I had an aggravating Friday at work, did some dishes and killed some zombies Saturday, and worked 6+ hours on Sunday. My work schedule is starting to be a vampire. It consumes twelve plus hours per day and the two more hours of commuting.
I’m just feeling sorry for myself. I need to make time to make shit.
Tonight, I start backing up files for the computer reinstall.
I’m doing this shit.