September 2003 Archives

I picked up an issue of the August issue of Scientific American from a friend. The facing page from the table of contents has a intriguing ad: it has a picture what looks like an extra-fancy bike computer. The ad copy reads:

BIGHA.com: The brain on a BIGHA predicts weather patterns, gauges heart rate and can store it all on a multimedia card. It's sometimes hard to believe it was designed by man.

I had no idea what was really be advertised here, beyond perhaps a fancy bike computer, but when I got the chance, I visited the web site.

I'll save you the time: BIGHA is selling a recumbent bike. It looks nifty, sounds like fun, and costs $3,000.

I don't know much about recumbents, but an FAQ I found on the subject says you can expect to pay $800 or more. Sounds like BIGHA is on the "more" side.

Still, it looks nifty. I'm guessing BIGHA is hoping the target demographic of this article will think so, too, and will be willing to pay the price. And if I had lots more money, I might check it out.

In the meantime, I'll continue to make do with my 15-year-old Trek 520.

Over the years one of my great frustrations has been the inability to marry databases with web sites in any kind of straight-forward way. There are lots of applications that naturally work best as simple databases. But woe unto you if you want to share that data on a web site.

When I was back at CNN, I wanted to build a simple database of changes that I could periodically publish out to an internal web site. What I was looking for was something similar what you can do using Microsoft Office: you can take a spreadsheet, a Word document, or a Powerpoint and save any of these in HTML form.

The closest thing to that I wanted was Filemaker. Filemaker would let you manipulate your data using Filemaker's very nice native Windows or Mac application. You could then publish a dynamic version of your database that would actually let others query your data and view it in various ways. Nifty, but since Filemaker acted as the webserverr, you had to serve the web site off your personal Mac or PC. Not very practical.

(The thing that bugged me the most about Filemaker: I wanted to be able to put URLs in my text fields and have those URLs show up as hyperlinks when those fields were displayed by Filemaker's web application. Simple and obvious, but I could never find a way to do with Filemaker back in '00. I could make a particular field clickable, but not a random URL embedded in a text field.)

I know about LAMP, which says you can build anything you want with Linux, Apache, MySQL, and Perl. But I want to be able to manipulate my data using something more user-friendly than raw SQL queries and command-line interfaces.

In the end, I was never able to find a way to do what I wanted.

Fast forward a few years. I'm now at Earthlink, and once again, I want to publish a database of changes for our SA team. This time I have a couple of choices - I can use Remedy, which is the company choice, but is difficult to modify. (I've played with Remedy before; it's flexible, but difficult to get to do what you want without having someone around who does Remedy for a living.) ELNK also has an internally supported simple web-based ticketing system that I could use, but it doesn't really quite do what I want. I'm back to my lament: if we can get chocolate and peanut butter together, why is it so hard to get databases and web pages to talk to each other?

Enter QuickBase. QuickBase is an Intuit-hosted web database service. Intuit won't sell you the code to run this - at least they don't want to - but they'll host your database for you, and give you a remarkably simple interface to both manipulate your data and share it with others.

It's not fancy. It's not relational. But if you want to share data that really should be a simple database, it's surprisingly easy to get it to do what you want. I was able to modify QuickBase's sample "bugs/features" database into a change control database in just a few hours. And if you include a URL in one of your fields, it's displayed as a hyperlink! Hallelujah!

The only real downside is that it's relatively expensive unless someone else is paying for it. Intuit has pricing plans for small businesses that range between $15 and $50 per month, and that allow for up to 5 users on up to 15 or so databases. For ELNK, I think I'd need the "Corporate Workgroup" plan, which runs $250/month for up to 10 users and 18 databases.

I'd sure like to see a freeware solution to this problem, or at least a reasonably-priced piece of commercial software. But given the the fact that I've never seen anything else that even comes close to what QuickBase can do, I'm glad to see any solution that I might be able to use to solve my problems at work.

Here's a splendid quote from Henry David Thoreau:

New ideas come into this world somewhat like falling meteors, with a flash and an explosion, and perhaps somebody’s castle-roof perforated.

I've never heard of that quote before.

In the past few months, I've been using an analogy of the Oort Cloud to describe the relationship between ideas about change and change itself. The Oort Cloud lies far outside the range of the planets: it extends out to as much as three light years away from the sun, and contains as many as a trillion proto-comets.

Once in a while these objects bump against each other, and one of them starts the long journey into towards the Sun, where it a comet.

I've been lobbying for improved change control procedures at work, and I've been using the Oort cloud by analogy to support my argument. Ideas are like objects in the Oort Cloud: there are more of them than you can keep track of in their formational stage. It's the ones that come inbound that you need to worry about.

A few blocks down from the hotel I've been staying out when I'm working out in Pasadena is an In-N-Out Burger stand. I usually ride my bike past it after work, and it always has a line of cars that stretches out to the street. I've been meaning to go there to see what the fuss is about.

Well, this weekend after I went to to see the bike races with my friends, I commented on the In-N-Out phenomenon. Since I'd never been to In-N-Out, we went there after the races.

I guess I'm just out of it. In-N-Out only has a few items on the menu: shakes, two or three choices or burgers, and fresh-cut fries. I had a burger with grilled onions and catsup.

I do like hamburgers. Doing a hamburger well is surprisingly tricky; I haven't really mastered the art at home. The best burgers are thick, and perfectly cooked. When I find a good one, it's usually done by the local variant of a diner; good burgers are hard to mass-produce. In our neighborhood in Atlanta, if I want a good burger, I go to the Galaxy Diner.

I guess I just don't get it. The In-N-Out burger was better than McDonald's, but not by a whole lot. The fries were ok, but perhaps not even as good as McDonald's. The the Coke was a coke. (I didn't try the shake.)

It reminded me of my reaction to Atlanta's Varsity restaurant. The Varsity is an Atlanta tradition, but to me, it's not one worth repeating.

I suspect I'm just the victim of not having grown up with it. When I was young, going to McDonald's was a treat.

The restaurant I missed when I moved away from Southern CA in '89 was El Pollo Loco. Their staple is flame-broiled chicken. (What I remember most is the corn or flour tortiallas and salsa they serve with every meal.) After I moved away in '89, I'd never been able to get back to one until I started working for Earthlink in Pasadena in June. Now they are my default choice for lunch, and they're still as good as I remember. Call me crazy, but I'll take El Pollo Loco over In-N-Out Burger any day.

One of the Summer Olympic events I've always enjoyed is bicycle track races. Track bikes are unfamiliar beasts - they have only one gear, no brakes, and they don't freewheel: unlike your average bike, you can't stop pedaling if the bike is moving. They don't have brakes: you slow down by pedaling slower. Unfortunately, Olympic coverage of track races is usually relegated to the after-midnight show.

Track races take place on - you guessed it - a track - typically an oval race trace 200-250m long, with steeply sloped sides all the way around. A typical trace race is a sprint: two riders, three laps. But they don't just start out full tilt. It's no uncommon for the two riders to go very slowly through first two laps, almost coming to a stand still, and then for one rider to make a break to out race the other rider. Lots of strategy.

This weekend an old friend invited me to go with him to the Encino Velodrome on Saturday.

The Encino Velodrome isn't much to look at; it's a concrete track set among a bunch of youth baseball fields. This weekend was the "Far West Cycling Championships." As the track announcer said several times, the Far West was a big deal in the 60's and 70's; this weekend it was attended by perhaps 200 people.

It was great fun. The racers ranged in age from perhaps 12 or 13 to 50. The big event was the Far West sprint championship; the winner was a 20-something Malaysian rider named Josiah Ng. Even more inspiring was the third place rider who was a gentleman in his 50s.

Great fun. Atlanta apparently has a velodrome, so I'll have to get the family out to see some races.