Entries from March 2007 ↓

The news coverage of American Idol (and wow, that’s a powerful statement about America in six words, but) today is all about the unlikely front-runners, who by their charm, pleasantness, or strange hair, command the voting. We’ve started with a singing contest, right? So the best singers are still moving forward?

First, the voting is fairly screwed up to begin with. It would be trivial to set up a one-viewer, one-vote system, but they’re pretty forward about saying “hack our system as badly as you can”. People with the time or inclination to vote only once (or five times) know that they’re being cancelled out by:

• teenagers with more time on their hands,
• people with (sponsor) Cingular’s phone service (who can vote more frequently, more affordably),
• the heads of voting blocks (internet fan clubs, or, on the other side, internet vote-gamers)
• outright hackers (who hold great power, but are the only group mentioned here of whom the AI folks don’t explicitly approve)

So the rational voter is already seriously disenfranchised. You could also get into the lunacy of basing voting results on least-favorite (so your one vote is actually eleven votes to get rid of everyone else), but they run it the way they run it. I seriously doubt that this system could accurately produce a winner on “America’s Tallest Person”: an hour of interviews, commercials, behind-the-scenes footage, and judge commentary, and when the contestants tell you how tall they are, you vote for America’s tallest. I seriously don’t think it would work. But let’s assume that the system is fair and reflects the considered will of the viewers.

It’s an entertainment show. People are voting for the future that they will find most entertaining. Sure, the final winner gets a recording contract. Big deal: that’s like a year from now. Next week, the person who loses will be gone. The votefortheworst.com folks are finding it’s very entertaining to belong to that community, and the Howard Stern fans find it entertaining to do whatever he says to do. The snowball has become something of an avalanche, and people seem to enjoy the fact that crazy-hairstyle-person remains on the show. They are given the choice to keep him or not, and they choose to keep him. (If you think that’s going to translate into a recording career, or people buying albums… wow. It’s going to translate into a couple national anthems at WNBA games, and the iron then cooling off considerably.)

In high school debate, it’s all about values. Debate judges don’t vote based on how well you spoke or whether you had a skull-crushing Jimmy Carter quote that made your opponent stutter: it’s all about what values you represented, how well you supported your value, and how you compared and contrasted your value with your opponent.

American Idol voting is made in a complete vacuum of values. You could get a pretty good measure of “which of these contestants is the best performer” or “do you want to see succeed in the recording industry” or “do you want to see again next week” or “learn more about” or “who do you most personally identify with”, but they never frame it that way. As a result, you get a pretty sloppy mess of “why did so many people vote that way?”, but at the same time, maybe that’s part of why people find that show entertaining.

Back to the point: values and criteria are super important, whether you’re trying to cast a vote for president or make a decision at work. Opposing options almost always have things going for them, but if it’s up to you, figuring out a higher-level “what do I want, why exactly do I want it, and how does this choice help me get it?” can help you make wiser decisions and get to know yourself better.

• Car audio installed. Best Buy didn’t have time for me on Sunday. Boo hoo.
• Got a little caught up on sleep, I guess.
• The Cheetah Pizza guy, who once offered me a free third slice of pizza in my two-slice combo box, actually snuck a third slice in there today. Leftovers, I guess.
• Read a little bit of Code Complete, which the UPS brought.
• Oh, just in time to be announced as the newest addition to the DBA team at work. That’s okay, it’s always good to learn.
• Kids took a nap, so I actually got work done.
• Googled “new business grant minnesota”. Ding!
• Made a delightful pasta primavera, despite the fact that Becky got all the cheese stuck to one of her brocollis.
• Exercised (watching Mr. Show) and cleaned up before 9pm.

But now it’s after 10, and it looks like I get to sleep for 6.5 hours again.

Wow, the internet is lousy with halfway-there GTD apps. It’s disappointing, because as much as you can do this whole thing with paper and pencil, it lends itself supremely to the digital world.

First things first: chop off the parts of GTD that won’t work better in the computer world. Ubiquitous capture just can’t happen online, in a desktop app, because you aren’t always there. (If you are close to always there, then you are close to ubiquitous capture, which isn’t ubiquitous at all.) So throw that out. Higher-level reviews are really about life planning and not losing what you considered life-focusing or important at any day or time, so I guess some sort of master list of that stuff could go in there, but you’d really want it journaled, wouldn’t you?

It leaves a to-do list. Really, really amazing to-do lists are trying to pass themselves off as “GTD solutions”, and I’m not buying it. The key to the GTD to-do list is that it has to be multi-dimensional.

Basically, all the tasks have to have a matrix of optional attributes, including:

• project
• context
• topic (high-level project)
• time estimated
• deadline
• urgency
• importance

Some of these dimensions are actually multi-dimensional themselves, like a project tied to two life-goals or a task that can be completed in any of a few contexts. The kids today love organizing that by “tags”, because it’s a freeform way to describe your stuff, but the open-ended tagging systems I’ve used (in stikkit, rememberthemilk, and others) are simply too open-ended. It’s great to be able to describe your tasks in these ways, but you can’t stop there. I haven’t seen in any system a way to pull the tasks that are lacking a context, for example, or the ones that don’t have time estimates, or ones for which the deadline is in the next 7 days, or whatever.

If you think about it, this is the exact same kind of filtering that iTunes does. I think the optimal solution is something along those lines: enter a ton of tasks, projects and ideas (as simply as possible), and then slice and dice through those based on what attributes they need defined (to finish up the data entry) or already have. iTunes playlists can tell you what jazz tracks, four stars or higher, haven’t been listened to in the past four weeks. Address Book can tell you which contacts you have in Ohio with no home phone number defined. A GTD to-do list should offer just as simple filtering.

The best thing about this approach is that it’s not too difficult to set up (if the data entry is simple enough), maintain, or tweak. iTunes newcomers don’t have their custom playlists tweaked out into infinity, but those who do can share their playlist definitions with others. It’s simple and extensible.

The important part of this is getting out of the way of data entry, because “task definition, assign context, tie to project, set importance, enter deadline, estimate time” doesn’t work for everyone and/or every task.

Honestly, I think the problem is pretty simple, and is just being danced around / poorly understood by the people who are doing this work. I like ThinkingRock best so far, because it actually meets you halfway between “anything goes” data entry (because this kind of work really is creative) and a supportive structure. It doesn’t go tag-crazy (and it could, and it will, I’m sure), it doesn’t accommodate time estimates (which I find useful, just because I pick off 15 minute things more often than I plan half-day things), but it reflects what I understand about GTD better than any other solution (downloaded or home-brewed).

They’ve had signs up at the Y saying “get ready for summer” by committing to actually showing up and working out three times a week. They take your $7, and you earn a small gift by accomplishing your goal.

I was a little disappointed to figure out that the gift is worth about $7, all told, because the odds are lottery-esque. My $7 should get me about $500 if this is a totally fair wager.

We saw that there is a giant wall of “participants”, with 63 squares for each day of the challenge. I guess everyone will know if you don’t go.

And now you know: Christie and I signed up this morning and forked over $14. If we actually get these sandals, they will be my most prized possession. (Also, we’ll be pretty much in marathon shape.)

I shouldn’t say that. We decided to join the gym because we realistically evaluated our schedule, and decided we could absolutely do 3 per week, together or independently. (Christie actually gets $20 back from her health care if she goes 12 times in a month, but public shame and sandals are apparently more motivating to us.) So this is just us living up to our initial commitment, just formalized.

They added a handful more of the elliptical machines (my favorite), so it’s easier than ever. We’re really just gonna go.

Wow, that post yesterday was about the geekiest thing I’ve ever written.  4,500 words about a Flash game.  But I’m okay.

I’ve probably lost anyone who wasn’t a die-hard Dan fan to start with, so I feel okay about transmitting this out into the ether.  (People frequently don’t know what to do with ideas: keeping them secret is the only way to keep them totally safe, but sharing them is sometimes the only way to make them happen.)

I wonder what people would think about miniature, two-to-twelve person theaters.  Living-room sized spaces, decked out with high-quality furniture (maybe theather-themed, but probably just couches and chairs) and top-line AV equipment, available for hourly rental.  They’d be in malls, and you’d probably book your two- or three-hour block by appointment.  The business doesn’t get involved with what DVDs you bring, or if you watch TV (live or recorded), sports, etc.

I think it probably has to be more expensive than an average night at the movies, maybe on the order of $30 (per person) for three hours, but I can imagine a market willing to pay that for a completely private, completely personal home-theater experience.

That’s my business idea.  I have dumb business ideas all the time, but this one in particular keeps pestering me.  I think it has potential, just because I personally would really like to relax with HDTV, appleTV, all the new video game systems, and other emerging home entertainment stuff, but I have no budget for even a small slice of that.

I think the next step is math-based: how many of these spaces can be built out in a mall square?  In a strip-mall space?  In unused office space?  And at full (and 90%, and 50%) capacity, how much money could they bring in, after you’ve paid for rent and employees?  How soon could you pay off the projectors, sound systems, and computers?

Played a ton of Desktop Tower Defense yesterday.  If you haven’t seen it, “creeps” enter the field from the top and left of your screen, and try to leave via the bottom and right.  You place little towers to take them out, and stopping them earns you gold, which you can use to lay down more towers or upgrade the ones you have.
I survived all 49 levels earlier this morning, and I’ll share a little of what I know before I retire completely.  I’m not going to walk through each type of enemy, tower, and boss, but there are a couple of things I figured out that might help you.

First-level thinking (and you get past this if you read the “quick start”) is: let’s block their path!  If we make them take a long, circuitous path, then we have more time to sap their health.  There are lots of schools of thought on how to make the most effective labyrinth, but my first instinct was to build a quick, corner-to-corner wall, and make the enemy traverse the length and width of the field two or three times.  You can also create more dense, edge-to-edge walls, but the important thing is that you control their flow.
If you look at the boards of the top scorers, you’ll see that many of them never upgrade their pellet towers.  These cheapest towers in the game don’t really upgrade well, and I’ve had luck leaving them as level 1.  Not everything on the board has to be working for you if you are doing a good job of blocking.  As a matter of fact, you might see that some of the top boards have only six or seven fully upgraded towers, and a bunch of pellet towers.  Since the number of enemies is finite, so is your income, and as much as we’d like to have 20 maxed out squirt towers, there isn’t that much money in the game.  It’s surprising how well one frost tower and four squirt towers can cope.  On the other hand, I’ve seen a few boards with maxed-out pellet towers, but don’t get confused: a level 1 tower of any type is really just a wall once you hit the mid-game.  99% of your damage is going to be done by upgraded towers.
Different configurations have different strengths and weaknesses.  You might find that the second or third flying wave completely overpowers you, or maybe the immune waves go through too fast.  That might mean you’re relying too much on darts or frost towers.  The game definitely encourages balance, once you figure out the basic strategy.

Most of the top scorer boards don’t have a lot of air defense.  Anti-air towers only fire on flying targets, and they don’t do an amazing job of that.  They aren’t a great investment.  Squirters do a lot of air defense if you have them near the middle, and frost towers also work.  The flying boss is nearly impossible without some seriously upgraded anti-air, but if that boss goes through untouched, how much does it hurt you?  2 health.  That’s all.
Most top scorers have a single point of entry into a maze.  That means blocking the entry ways, so when 10 come from the left and 10 come from the top, they end up having to enter a narrow passageway at about the same time.  If you’re not using frost and darts, this doesn’t matter - the other tower types don’t have “splash”, or do damage to nearby enemies as they hit their target.  But a long stream of enemies can quickly become a bunched-up, slowed down, heavily damaged clot when you get them to show up in a long corridor together.

Once you figure that out, you realize that you’re not really building walls to keep your enemy from leaving, but to route them back to your main, upgraded weapons.  With the right layout, an enemy might have to walk in the range of a maxed-out dart tower five or six times before it finally gets a clear look at the exit.

This might be obvious, but upgrading takes time: specifically, downtime.  You have to upgrade sometime, but there are ways to optimize.  Dart towers are useless during flying waves, and frost towers are pointless during immune waves.  Beyond that, I try to upgrade one piece at a time.  Since you only have a handful of really useful towers, you don’t want half of them down for an upgrade at the same time.  Besides, having money for two upgrades at the same time means you didn’t upgrade when you could have done the first one, so you screwed up once already.  That money doesn’t earn interest.

I only rarely sent a wave in early.  I am pretty sure that you have to survive the game to have those bonus points count, and I don’t see the advantage.  There are levels that seem to end “early”, and you might jump ahead there, but it doesn’t happen a lot.  (Again, your enemies are your source of income, so jumping ahead is not really a serious disadvantage if your maze can take it, and if you can build out that fast.)

Another thing you might notice is that the top scores are littered with half-filled boards, and a smiley face or a signature at the bottom.  I imagine that these players have sold off their towers as they beat the last guy.  Selling towers is important at the beginning of the game: you get a pretty good price for your used towers, and you definitely don’t want to leave gaps where you’re planning future development.  Stick in a 5-gold pellet tower, reclaim your 3 gold when it’s time to sell, and don’t lament the 2 gold you lost.

Another strategy I never implemented: juggling.  If you’ve ever opened up a hole by accident while upgrading, you know that the enemy is always recalculating the optimal (shortest, not “least damage”) path to leave.  If you have him running down three long hallways and something changes in the maze so he’ll only have to run down two, he’ll backtrack to go through the hole you just made.  (If you haven’t played, note that it’s illegal to put down a tower that would actually prevent the enemy from leaving.)  It wouldn’t be hard to create a couple of key points in your maze where the sale of a tower here and the erection of another one somewhere else would result in a particularly nasty enemy backtracking as many times as you had gold to do it.  I imagine the game essentially comes down to this on “hard” and “challenging”.

Anyway, the strategy of this game is amazingly deep, and the “live action” aspect of it means that even the best strategies can be unsuccessfully implemented.  The masterstroke of the developers was to show off the boards of the best players: you can gain a ton of useful strategy information from reading these boards and seeing how people play it.  If you haven’t played, give it a shot and let me know how you like it.

I’m the best blogger ever. Most of my posts are “hey, let’s see if this configuration tweak worked.”

Since you’re asking, I’m trying to get tags working again. They aren’t, or haven’t.

update: TextMate, brilliant as it is, must have undone some of the blogging.rb changes I made (and I only remember making one, but I could be wrong). Either way, it appears that I’ve fixed my problem, and can post again from TextMate, with tags, at my convenience.

Label: Akon, good work on that Gwen Stefani track.
Akon: Hey, thanks.
Label: We’ve been singing “woohoo, yeehoo” around the office for days. It’s like a spring anthem for us.
Akon: That is what I like to hear.
Label: It’s probably the best thing you’ve ever done.
Akon: Well, it’s just “woohoo, yeehoo”… as a statement, I kind of liked my last single more.
Label: You mean the one with Eminem about aggressive sexual congress with a clubgoer?
Akon: No, the one with Snoop, about… sexual congress with a stripper.
Label: Right. No, “woohoo, yeehoo” is definitely a more profound artistic statement than either of those.
Akon: I disagree with that - I am successful and adored for my ribald tales of sexual congress.
Label: We are almost certain that you will be remembered years from now, even after death, as the “woohoo, yeehoo” guy. We’re shooting a video this week, and…
Akon: I am extremely busy now. And during February.

I got totally ready an hour ago to tell you about my day, but then I decided that my post wouldn’t be totally ready to go until I’d worked out Show Menu Items, something I’d always wanted to do in Quicksilver but never figured out. I had three settings to fix, so that explains what happened to the hour.

More later, I guess. I haven’t forgotten about you guys. Make sure you’re hooked up to the Twitter feed (Tony is!) and the del.icio.us feed (Cratchit is!).