I only just found out recently that there’s a Crust in Parramatta. We ordered pizzas from them last night. They were Good. I think Tina is converted.

(This is in no way Movember-related.)

Yes, there’s a good reason that I’m wearing a high-contrast stripy… is it “stripy” or “stripey”? “stripy” looks like a case-insensitive string processing library for Python or something.

…shirt.

Anyone noticed that @7pmproject has more motion blur than most TV? As though the studio is dark and the cameras have a long exposure time…

Wow. My head looks even more bobbly and off-centre than usual today.

In an ideal world, what sort of setup would you have to take one picture per day in as close to exactly the same position as possible? Based on recent experience, you’d at least need

• a fixed camera,
• a chin rest, also fixed, and
• consistent lighting (away from windows, or only taking pictures at night).

That sounds like a lot of hard work and, more importantly, space. I wonder what you could do with it to justify all the effort…

Lock Stock and Two Smoking Spongebobs #manlegs

The Mo is starting to attract comments. Some random dude in the lift told me he tried to grow one for Movember a couple of years ago but it got too itchy, which is… a lot of information to share between floors, I think. A guy who used to be on my project at work did the whole “no that is WRONG you can’t DO THAT you look WRONG” tirade, notably doing so for several seconds before I even realised what he was talking about, which was mildly amusing. So to him and all those like him I say: just ’cause you can’t grow facial hair…

Let no one say I don’t suffer for my art.

(Yes my art is sculpting facial hair.)

Had to re-shoot today’s because I forgot I still had a fragment of toilet paper stuck to a shaving cut.

I’ve never been very good at the tiny-square-of-toilet-paper thing. No matter how long I leave it to heal, the act of pulling it off invariably rips open the wound and it starts gushing again. The cut today wasn’t even really bleeding; I just put it on there in case it opened up again at some point, which it did, as soon as I tried to pull it off, because apparently we’ve bought some brand of toilet paper that welds itself to your face.

There used to be a little stick of stuff that you could dab on wounds to stop them bleeding. I remember trying to buy one a few years ago and not being able to find it. Does anyone know what happened to it (or, indeed, what I’m talking about)?

(Alternate title of today’s update: Things You Didn’t Want to Know About Dave’s Blood and Toilet Paper)

Week 2 begins.

Tina got back from Melbourne last night, so it was the first time she’d seen the mo since early on day 2. She didn’t veto it, which is good. Although she did say I look Mexican.

Still having issues with symmetry. Part of the problem is that shaving is an asymmetrical process if you do the whole thing with the razor in one hand. But then, if I tried to shave half of my face left-handed, the left side of my face would be a jagged bloody mess. I wonder if ambidextrous people are more likely to even out their goatees correctly?

Random almost-related observation… The classic conversation-at-the-driving-range scene in a movie or TV show amost always implies that one of the characters is left-handed, because they’re usually facing each other (and I’m pretty sure nobody would swing with their off-hand by choice).

I noticed this in the third season of Dexter – and Dexter was the one driving left-handed. This struck me as interesting because (a) it’s an ironic contrast with his name (”dexter” being Latin for “right”, from which we get “dextrous”), implying that he’s not what he seems to be; (b) the negative stigma of left-handedness would be a useful thing to associate with his character; and (c) I hadn’t noticed it before, and unfortunately I didn’t pay attention to it afterwards, but I’m about 80% sure he’s portrayed as right-handed everywhere else. On the other hand (pun strictly unintentional), I remember noticing that Brian from the first season was left-handed. But maybe I got that wrong too. Google was no help here.

It’s taken this long to work out what the conclusion of this aside is, but I think it’s that it’s annoying that they could have either made a slightly bigger deal all along about Dexter being left-handed, as a metaphor for him being more “sinister” than he seems on the surface, a possible direction that his name seems to set up but not explore; or, if he’s not left-handed, not randomly and inconsistently implying that he is. The most annoying thing is that if Miguel had been driving left-handed, they still would have been facing each other in that scene, and it would have implied those things about him, while also subtly setting them up as opposites, which would have been awesome.

Sorry, I’m overthinking this. The point is…

Um…

I’ll never have symmetrical facial hair.

Edit: Of course Dexter is right-handed. Ten points to the first person to leave a comment pointing out the completely obvious and umistakable clue that I just noticed.

I was watching a COMP1927 lecture online (yes, this is what I do for fun), and remembered something that I was going to post but hadn’t gotten around to yet.

I’ve been fiddling with Shell sort gap sequences a bit. The idea I was working on was that a good gap sequence would consist mostly of numbers that are mutually prime; it seems like numbers with common factors would lead to sub-sorts that cover a lot of the same ground, for the same sort of reason that hash functions don’t work well if the hash value is modulo a number that’s some kind of factor in the data. (I don’t have any particularly good theory to back this up; it’s just intuition.)

Apparently (or “wikily”, a word we invented recently to mean “according to Wikipedia”) a geometric sequence that increases by a factor of 2.2 makes a good Shell sort sequence. So I tried out a sequence that:

• consists of products of the first n primes (if you write out the prime factors of each number in the sequence, each of the first n primes occurs exactly once); and
• increases by approximately a factor of 2.2.

Finding the best such sequence (the one with a ratio between pairs of elements that’s close to 2.2, by one of a handful of definitions of “close”) is a task in itself. Computing the sequence for n=16 took nearly a week, and the result is:

1, 2, 5, 11, 23, 53, 119, 377, 817, 1739, 3813

or, as products of primes:

1, 2, 5, 11, 23, 53, 7×17, 13×29, 19×43, 37×47, 3×31×41

And, empirically, it does pretty well. Not as well as Ciura’s sequence (1, 4, 10, 23, 57, 132, 301, 701, 1750), but pretty close. I don’t have the test-run data on hand, but I might put it in a follow-up post later.