Wednesday, May 24, 2006

SpanofLife Interface

While lying in bed this morning I thought of the spanoflife interface. Since we only have 3 dimensions to work in, I'll have to go 3d. But, what I realized is that people live their lives in 2D space. Yes, we climb mountains, go into Death Valley, etc., but for the most part we don't gain or lose much elevation, compared to the horizontal distance we cover. So I can use that dimension, height, to represent the passage of time. Each person will have a life line that stretches from the top to the bottom of the screen, and their postiion on the earth will be represented in the X,Y plane.

This is basically the same as some railroad timetables.

This will not be applicable for astronauts.

This post doesn't make much sense.

Tuesday, May 23, 2006


I've been more interested in economics lately. Although my understanding of it is shaky at best, I recently read a few things which are scary to say the least.

First - the impending baby boomer retirement. Now, everyone already knows about the trillions in healthcare costs that this means, and that the younger generations will be shouldering that without enough people in the workforce. But - what about all the 401(k)s? I just read an article in the Economist that goes like this: When everyone retires over the next 10-15 years, everyone is going to be cashing in their 401(k)s. In other words, everyone of retirement age is going to be converting those stocks and bonds into cash to live on. How? By selling them on the market. But....who is going to buy them??

There are not enough younger people in the workforce, in the US, to buy up all of those stocks and bonds. What does this mean? It means that the stock market is headed for a deep dive, because of simple supply and demand. Without sufficient demand, the prices are going to dive, because of the glut of stocks and bonds on the market. This means that, for instance, my 401(k) is going to tank, unless I convert it into some other asset (cash, usually), which of course exacerbates the problem.

The authors of the article basically propose that the only real solution for this is going to be to make it as easy for Chinese and Indian workers to buy these assets as it is for Americans, thereby increasing demand. Alternatively, grow our own workforce by leaps and bounds, by.... oh, I don't know.... granting amnesty to tens of millions of Latinos, who would now be eligible to buy stocks and bonds?

It's fascinating, because it's so basic, and yet such an unexpected result.

So, speaking of cash. Some years ago, the US went off of the gold standard. Up to that time, the US government guaranteed that you could (in theory), go and give them a dollar, and they would give you a little bit of gold (not really, but bear with me). In other words, money was only printed as a substitute for gold, which they maintained in large vaults. This was the gold standard, and it ensured that the government printed no more money than they could back up with gold.

Why gold? While it is pretty, gold is basically in limited supply. What causes a thing to have value? Well, it could be useful, or it could be in limited supply. When you buy stocks, you are betting that the company is going to produce something that people want, thus your stocks are valuable because they allow someone to produce something useful. Gold is easier to understand - it's valuable because there's not much of it. If it was easy to produce gold from, say, dirt, then gold would be as valuable as dirt. Having a gold standard meant that the risk of inflation was low, because the government couldn't just print more money unless they could obtain more gold.

Well, the US went off of that standard. Now we have what is basically paper - it is a currency with no inherent value. It is not backed by anything except the government's assurance that it will pay for things (this note legal tender). So what does that mean? Well, the dollar has been falling in value against other currencies. This means that other nations are starting to not quite believe that the US government will back that currency. (Which is reasonable, since we are running a deficit in the trillions of dollars - $24,000 in debt for every man, woman and child).

And the price of gold has gone through the roof. It was about 200 dollars an ounce a few years ago. It is now $653 an ounce. A lot of that is speculation. But some of it reflects real concern that the world's major currency, the dollar, is heading for the same fate as the peso and the ruble - worthlessness. So people are buying metal dug up out of the ground, only because there's not that much of it, so it is inherently valuable. Scary stuff.

Monday, May 22, 2006

shiny new toys

Well, I'm taking a break from writing about my own attempts to be creative so that I may tell you about several new consumer goods I bought over the weekend.

First - wireless speakers from Home Depot. I plugged the transmitter into my computer, and tuned it to my internet jazz radio station. Then I put one speaker in the sunroom, and one in the kitchen. Now there's instant muzak on demand. They actually sound pretty good to me, since I had really low expectations from what is basically baby-monitor technology.

Second - a Roland Micro Cube guitar amp. I've always been skeptical of digital guitar effects and amps, since what I've heard has sounded like crap. But - this has changed my mind. I went into Guitar Center looking for a distortion pedal. For $125, this is a complete amp that can sound like a jazz amp, a little british combo, a Marshall, or a high-gain Mesa-Boogie. I'm not kidding about the last 2, either - it really does. They've done a freakishly amazing job. It has a 5 inch speaker and runs off of 6 AA batteries, and it gives your head problems. Your eyes just can't believe that the sounds are coming from this little thing that you can pick up and swing around your head.

Plus the interface is great - 2 knobs for all the effects. They realized that hey, this is a practice amp. People don't want to sit there and diddle with 16 knobs just to get a bit of a swooshy sound, so they put 4 effects on one knob.

And the amp modeling. Yes, ok, digital technology really is amazing. It doesn't quite respond to touch in the same way that my little tube amp does, but it's close, and it has a much fuller sound when played clean. I believe the way it works is this:
  • an electric guitar is hooked up to an amp. (say, a Marshall stack).
  • the signal right out of the guitar is recorded (before it hits the amp)
  • the signal out of the amp is recorded
  • the signals are compared, and a mathematical model is constructed of how they differ, ie, here's what you have to do to the raw signal to get it to sound like the output from the amp.
  • That model is encoded onto a hardware chip, and placed into the amp.
So now, instead of your distortion effect coming from a diode clipping, a chip is taking your signal and transforming it in the same way that the Marshall would have transformed it (compressing it, distorting it, speaker cone interacting with air), and it comes out of the amp sounding like a Marshall.

Third - the best of all. The Nelson RainTrain sprinkler. 20 pounds of cast iron, shaped like a tractor with a sprinkler head on top of it. Every time you have to water your lawn, you get to lay out the hose like a little train track, then put the shut-off ramp at one end and the sprinkler on the other. Then you turn the water on, water spins the sprinkler arms, arms turn the worm gear, worm gear turns the (2-speed!) transmission, sprinkler crawls across the lawn, following the hose track that you set out for it, and stops when it hits the ramp.

Great sprinkler to play in, too - you don't have to choose between dry and drenched. Abby loves it, I love it (possibly too much). I mean, it combines trains, tractors, and hydro-power, and mostly incidentally, does something that you're supposed to be doing.

Thursday, May 18, 2006

Parsing Wikipedia

On spanoflife, I set up a MySQL database, and started to set it up according to the ERD I built, then realized I should really prototype some interface ideas. So I went looking for sample data and discovered that it is really hard to find any data in a machine-readable format. Most biographies and timelines are in the format of paragraph text. The richest source of this data seems to be wikipedia. So. The real next step is to build a parser which can take a page on, say, Galileo, and read out the events in his life into database entries. I guess I'll use python just because I have it on hand. I found a parsing library, so I can turn things like:

In 1611, he went to Rome, where he joined the Accademia dei Lincei and observed sunspots. In 1612, opposition arose to the Copernican theories, which Galileo supported. In 1614, from the pulpit of Santa Maria Novella, Father Tommaso Caccini (1574-1648) denounced Galileo's opinions on the motion of the Earth, judging them dangerous and close to heresy.


1611 He went to Rome, where he joined the Accademia dei Lincei and observed sunspots.

And so on and so forth.

Wednesday, May 17, 2006

Player Dulcimer

I have this hammered dulcimer sitting in the corner that I've been meaning to play more. It's a beautiful instrument, and it makes a beautiful tone, but I go through these periods of musical drought where I don't play anything, and then it just gathers dust. Plus, I have noticed that playing it exacerbates my wrist problems, so I can't play it for very long. So.... what to do?

As I was cruising, one of my favorite surplus electronics websites (yes, I have more than one), I came across this page:

Isn't that just the cutest little spring-return solenoid you ever saw?

So I thought, hey, I could put little hammers on the ends of these things, suspend them above the dulcimer in a frame, connect a MIDI-to-parallel interface, plug in a computer filled with Christmas carols and reels, and voila, instanct Celtic charm. This is quite similar to my player pipe organ that I built earlier - the only difference is that these are little bangy things instead of little openy things. I already have the MIDI-to-parallel board, since I have no room for the organ in the new house. All I need are the solenoids (on order), and the computer (well, we have a bunch of those...) , and the frame, which shouldn't be too difficult.

Monday, May 15, 2006


Another thing I'm struggling with is how to integrate the display of time (basically a linear string with events on it) with space (messy, 3-d). I would like a zoom feature here too, so that you could, for instance, view a line representing all events on the European continent, then be able to zoom in and separate by country, then maybe by city, etc.

I don't really want to get all complex with geographical views, since that would involve getting into 3d rendering. For a first pass, it would probably suffice to separate out the timelines side-by-side. Although... it may be neat to have a 2d map at the bottom with the timelines growing out of it. As you zoom, the map zooms too - but now you really are talking about another API, and lining stuff up on the screen. But it would be a very graphical way to show the geographical proximity of various events and timelines.


I was playing around with outlook to see if i could use it as a testbed for spanoflife stuff (since it already has a nice built in data model and API for dealing with dates, events, and tasks), and I entered an appointment for Feb 2, 1874 just to see if it could handle dates far in the past. It accepted it and promptly put up a message telling me my appointment was 6714 weeks overdue.

In fact, through exhaustive study, the earliest date you can enter is 4/1/1601. An odd starting point, maybe someone has a sense of humor.

As always, Wikipedia astounds, with a page dedicated to 1601:

I think I will have to scrape or link to Wikipedia for this kind of event content.

This does bring up a point about date handling. I want to be able to represent the entire lifespan of the earth on a timeline, which of course is only fuzzily known. I have entries in the tables for start and stop times, but those are going to have to be supplanted by "approximate" dates of some kind. I want to use the modern calendar, obviously, but it has to be able to zoom out to things like "10,000 years ago".

Friday, May 12, 2006

The zen of drywall finishing

The first mudding tools I bought were from the paint aisle, and were not well-suited - the knives too narrow, the mud too thick. I went and got a proper mud pan, wide knives and lightweight compound, and it was actually kind of enjoyable. I'm sure I'd feel different if I was doing a whole room, but at least now I see how you could get pretty efficient at finishing drywall.

Thursday, May 11, 2006


I don't know if I will get around to working on this for a while, especially since it involves computer work. The idea is to provide a web-based "timeline" interface for interesting spans of time. This could be your life or other lives, or it could be spans of history. You would be able to see key events on each span laid out, along with images, etc. Bringing two or more spans together would let you compare your own life with, say, Alexander the Great's. Or see the timelines of the Founding Fathers all together, so you could see that in fact, Benjamin Franklin and John Adams were of two different generations. Or overlay what's happened in the past 100 years with what happened in 1700-1800. Or layout your kids's pictures in a linear format.

The idea I want to get across is that everybody's span is pretty short, so you need to get the stuff done that you want to get done before the Reaper comes. But in a positive way. From a practical standpoint, there's a lot of work to be done. I think I realized this when I just laid out the years from 1974 to 2054 in Excel, and put my age next to each year, and filled in some things up to my current year. I saw that
a) my life is about 3/8 over
b) stuff takes time to accomplish, but not as much time as you think, if you ever get started
c) it was pretty neat to see that all laid out.

Tuesday, May 09, 2006

Motion on projects

I have obtained 3 craptacular women's bikes from some dude in Lawrenceville to serve as the basis for Frankencumbent, the recumbent, front-wheel drive, probably unridable bike. I can't wait to attack them with my new $10 angle-grinder.

I've also registered for another project. Things sure are different from the last time I signed up. Domain name plus 6 months of PHP hosting for $28.

Monday, May 08, 2006

Exercise Go-Kart - Google Search

Here are the results from my google searching on the exercise bike:

Searched for:
rowing bicycle

First result:

Which has ALL manner of wacky bikes that use your upper body. Neat animation half-way down the page. I don't see any that combine standard bike pedaling and upper-body, though. But a lot of rowing bikes.

And the gold mine:

People have been doing this for a LONG time, it seems. I still haven't seen anything exactly like my vision, but plenty of things that come close- mostly 2-wheel, interestingly. And some really entertaining stuff.

Last but not least, here is the exact thing I was thinking of (in a bike format, but the same idea) - there are also other patents from the page above that are trikes, with more or less the same thing - double handles that serve as extra propulsion and steering:

Exercise Go-Kart

Ok - this is recumbent bicycle part II. I thought, since recumbents are so bad at climbing hills because you don't have the weight of your body to assist, why not get the rest of your body into the act? I've been doing weights at the gym to work on my upper body, what if I could do the same thing outside while moving somewhere?

The basic idea is that you harness your upper body to drive the bike, in addition to your legs. I think this calls for 2 things:
  • Recumbent sitting position, so your upper body is upright as it would be at a gym.
  • 3 or 4 wheels so balance is no longer an issue. If you're going to be flailing all your limbs, I don't think balance is going to happen.
I envision either a tadpole trike (2 wheels at front, 1 at rear), or a 4-wheeler, with the rider seated in the middle. Standard bike pedals in front. 2 handles extend upwards from the bike to the rider's hands, at about chest height. (Think ski poles). By moving these handles back and forth, the rider adds power to their pedaling. The actual motion could be varied - one possible motion would be a ski-pole like motion such as you see on an elliptiglider at the gym. Another would be more of a rowing motion, whereby the handles are both forward or both backward at the same time.

The handles are hooked into the drive train via a standard crank mechanism. For the rowing motion, I would think you'd want to gear the handles down, so the row motion is about half the speed of the pedaling motion, otherwise you'll look like a silent movie of a gandy dancer railroad car.

The main challenge is going to be the steering. Since you're not on 2 wheels, you can't have lean steering. Perhaps the handles could move side-to-side as well for steering - I'm not sure how well that would work - could you provide power from your hands and steer at the same time?

Google sucks my ambition

One of the problems with a completely connected world is that in a few keystrokes, you can discover how none of your ideas are new. Google has given everyone the ability to enter their idea into a box and see if it's been implemented. I know this should be a good thing, letting us all build on the knowledge of others. But sometimes you just want to go off on a tangent without knowing that there's an entire club devoted to the idea you just came up with. So I'm going to pursue my ideas here in a 2-part format- first I'll post the idea and explanation, then in a second part I'll post my google search results for it.

Drywall sucks

I put up a little bit of drywall over the weekend, and it's a messy thing. Just like anything, by the time I got reasonably good at it, the project was over. It is a weird thing, though- you are surrounded by thousands of pounds of talcum powder in a modern house. Our other house had plaster walls, which are so much more refined (but even more messy and require a lot more skill to implement). So what does drywall have going for it?
  • Cheap - it's made of powdered rock, which we have plenty of, and it's not being used for anything else.
  • Flat - or at least, the paper covering is. So really the flat, finishable surface is simply paper.
  • Dense / non-resonant - it provides some measure of sound deadening - contrast to walls surfaced with a thin wood - the whole room acts like a drum.
  • Relatively workable - it can be cut with a utility knife. Once it's up, though, it's toast. If you want it down, it pretty much has to be destroyed.
What sucks about it?
  • Messy. Any cuts to it result in dust everywhere
  • Fragile. It cannot tolerate flexing at all, and has to be protected at all edges by corner bead. Even light bumps from furniture can result in breakage.
  • Heavy.
About the only substitute I could think of would be some kind of board manufactured out of recycled paper. Basically, the exterior would remain a sheet of paper, just like drywall, but the inside would be made of little pressed paper fibers. That would seem to fulfill the cheap, dense and workable aspects, while cutting down on the messy and fragile aspect. Don't know about the heavy part, though. Probably even worse than standard drywall in moisture situations, too.

Friday, May 05, 2006

Recumbent Bicycles

What with $3 gas and everything, I have a renewed interest in biking. I got our bikes out of storage, but it's just so normal. And it makes my hands hurt after a while. My dad and I built a recumbent bike out of electrical conduit back in high school, and I rode that thing everywhere ,even though it had to have weighed more than 50 lbs. (Going uphill fast really not an option). I've been looking at various homebuilt options, and am especially intrigued by some that allow you to hook 2 of them together to build a little bike-car that can also tote kids and groceries. I suppose my vision of the whole family pedaling to the grocery store or out to eat probably looks better in my head than it would work out in real life, but here are the contenders:

The $18 Recumbent. 2 thriftstore bikes, 1 hacksaw, front wheel drive. Probably not going to get Trisha on one of these. (or any of the front-wheel drives). But I do like the fact that it's got a short chain.

Like the above, only with more welding involved.

And then the bike car options. There are also other "quadracycle" things available, but they seem to be more suitable for a beach boardwalk - fun, but you're not going to get anywhere very quickly.

Both of those can also have electric motors attached, which would cut down on the pain of hills.

Text Based 3D Modeling

I've been playing around with the Google free SketchUp program, and it's pretty nifty, although it takes some practice, and only really seems suited to architecture. Like all other 3D programs, it starts killing my wrists after a while, and I wonder if it might not be sensible to build some sort of text interface to a 3D modeling program. Sort of like POVRay (in which I built a really bitchin', animated version of a steam engine, long ago). But of course POVRay is not realtime. I want something like:

rectangle(10,10) upperJoist;
upperJoist.extend(z, 20);

which would create a block 10x10x20. As far as positioning and putting things together, I think you could take a cue from SketchUp and do intelligent things with endpoints, faces, etc. Hmmm... maybe I just want to download POVRay again.

Update: Such a thing kind of exists - Visual for Python lets you do almost what I say in the post above:

floor = box (pos=(0,0,0), length=4, height=0.5, width=4,

ball = sphere (pos=(0,4,0), radius=1,

ball.velocity = vector(0,-1,0)