So, I’ll probably post soonish (read: within a week. If not, poke me.) with some photos and whatnot of whatever happens today, but I figured I’d point these links out just in case someone finds them interesting today. (I’ve been told that I should point out that the majority of these items are not links, as that wouldn’t quite be legal. There are some links below to other songs.)
Halloween, like a select few other holidays, needs some music. It’s kinda required. There are some obvious choices:
- Ghost Riders in the Sky – The Johnny Cash version is pretty decent.
- Psycho Killer – Talking Heads
- Clubbed to Death (Kurayamino variation) – Rob Dougan (off the Matrix soundtrack, instrumental)
Then come the clichés. I happen to have a many-years-old CD (from grade school) of these songs, most of which seem to fit:
- Monster Mash – Bobby “Borris” Picket & The Crypt Killers
- The Addams Family Theme – Vic Mizzy
- Ghostbusters – Ray Parker Jr.
- Jaws Theme – Henry Mancini
- The Twilight Zone Theme – Nell Norman
- The Purple People Eater – Sheb Wooley
- Love Potion No. 9 – The Searchers
- Werewolves of London – Warren Zevon
- Alfred Hitchcock Presents Theme – Arthur Fledler
- Toccata and Fugue in D Minor – Bach
They’re actually not that bad either.
And if you don’t mind mashups, I’ve found the Mashup Industries’ Bride of Monster Mashup and Evil Twin of Monster Mashup to be pretty decent. (Free reg req’d.) Some of them aren’t quite child-appropriate, so make sure you listen before you play them for the world. I still have DJ NoNo’s County Sound stuck in my head.
Anyway, I might be playing music for Trick-or-Treaters at a friend’s house, so if anyone has any appropriate suggestions, please feel free. I’m sure I left out a lot of good songs.
(This post is going to be much more mathy than my standard posts, mostly because I thought about this as I was filling out college applications and decided it would be interesting to follow up on the random thought.)
(I apologize in advance for the poor quality renderings, but they should at least be legible)
A bunch of math follows in the full post.
Second math team meet. Can’t really think of any way I’d rather have spent that Thursday night.
Had a good time, obviously. Seniors tied for second place, so that was good. Unfortunately, that’s about the best I can say, because the rest of the team didn’t fare quite as well.
Ah well, it was amusing. And I got to try out my new camera, which is responsible for the I-can’t-believe-it’s-not-VGA panorama above. The full version looks much better, honest!
Beyond the fact that I’m now blogging more often than ever(?!), now I can apparently blog at superhuman speeds. (more…)
So, I was a member of a group that went to homecoming earlier this month. Of course, going to the homecoming dance kinda implies that you’ll, well, dance. So, here’s the gist of what I was told or otherwise understood:
- Dance with your feet at least shoulder-width apart. Probably further.
- Shift your weight from one foot to another, while bending your knees. Bend the knee you’re shifting your weight to more than the other knee. This should happen naturally anyway.
- If the beat is fast enough, or you’re alternating sides too slowly, try to bend your knees more often than you shift your weight.
- Dance at a speed appropriate for the music. If it’s a slow dance, revert to that style of dancing (or step back and let someone else do it). If it’s a really fast dance, move faster.
- Try rotating your torso from the base of your spine slightly. A few degrees is good enough.
- Move your arms. It can be as simple as waving them wildly about, but keep your hands above your waist most of the time or you look like you aren’t really interested in dancing.
- If you can, move your hands around. They don’t have to be coordinated, but you have to hold them slightly closed, with the palm down. Don’t keep your hand open, and don’t make a fist, even though you may look like you’re punching while you’re moving your hands.
- Like your feet, keep your elbows loose and not necessarily close to the rest of your body.
- If you ever wanna make it, nod your head.
- While nodding your head, move it from side to side a bit as well. It doesn’t need to be in any constant direction.
- Don’t bother with “The Rock Concert Instruction Manual” – it’s only for “MegaStars” on stage.
After the first few minutes, most of the above should be relatively easy to sustain without much effort. If after 20 minutes you still can’t do it while looking at something else, you might be trying too hard. The point of dancing is to have fun, after all.
So, imagine you had a way of identifying individual computers when they visited your website. You could link together the sessions of a single user, and identify their actions across those sessions. You could see how long they spent on your website, and why (rather, where) they kept coming back. You could even see if they had multiple accounts on your service (assuming you have a website where users log in), by identifying who logged in, on which computers, when.There’s just one problem. Well, besides the semi-Orwellian sound to it, anyway.
There’s no good way to differentiate a user and a computer. What if it’s a shared computer among a family, and the two people with accounts on your website are brother and sister? Combining their accounts together is unlikely to get any useful data, because it’s a mix of two different users. An extreme example would be a library computer, which is used by hundreds of different people before being used by the same one twice. It’s rather irrelevant if you’re not trying to identify a pattern for a single user, because the user doing the action doesn’t matter when computing aggregate statistics. But that falls apart when you try to compute a “people interested in X are probably interested in Y” relationship (or something similar).
So, how do you handle a multi-user computer for per-user data? The simple solution is to only compare data while people are “logged in” – but that requires them to have some incentive to log in. If you’ve got a business website, requiring people to log in without any reason will just drive users away. (See the New York Times visitors pre- and post-required signup.) Assuming that’s not a good option, the next question is “How likely is it that multiple users will visit the website on the same computer?” For my blog, that’s pretty simple: Not likely at all. For a website like Google or the New York Times, it’s rather likely that multiple different people will see their website from the same computer. I’m not in a position to collect such statistics, but I’d still guess that more than half of computers that access any given website will be from a single user. If that’s true, then useful data could still be gained by computing the most common uses and discarding the uses that are very far away from the normal. The question is how to keep as much useful data as possible while ensuring it’s actually useful data. If anyone has any suggestions on how to do that, please let me know.