Why Elevators Matter

by David Sklar

Related link: http://www.oreillynet.com/pub/wlg/3190

Paul Graham's essay Hackers and Painters is thoughtful and Tim O'Reilly's commentary on it is exuberant, but to what effect?
The people who need to be swayed by Graham's arguments are already swayed.
Hacking is fun, hacking is creative, hacking is not just implementing a spec.
This was made quite clear to me in fifth grade when I read Steven Levy's
fabulous book "Hackers". Software connoisseurs appreciate the fine art of design. They are different than the journeymen with their lunchpails that stuff the code inside the virtual methods handed down from on high.

Most people in the world are not software connoisseurs. They just want it to work. I am not an electrical engineering connoisseur. I don't care if the circuit board inside my computer is laid out in a particularly beautiful or elegant pattern. I just want it to work. I don't care if the elevator in my apartment building is based on a design by a noted elevator designer. I just want to go up and down.

I have, perhaps, now offended aficionados of fine elevators. How can I be so blind to the beauty above my head and below my feet in the elevator shaft? Without elevators there'd be no Empire State Building, no Petronas Towers! Designing a good elevator isn't just rote engineering, it's creative, it's art, it's beautiful!

All of this may be true. I'm just not into elevators.

Web Application development is a very young field. This has lots of exciting and positive implications -- it's bursting with unbridled creativity. Many people are building lots of neat things. But even more people are building lots of shoddy things. Much of this is natural -- boundaries are being sorted out, everyone's figuring out the rules of the game. To those outside this community of developers and connoisseurs, however, the implications are less thrilling.

When my mom* rides the elevator of a web application, she doesn't care about the creativity of the programmer/designer behind the scenes. She wants her airline ticket, or book, or search results, or Happy Flag Day e-card. Maybe there's a really cool macro language on the e-card site** that lets e-card designers whip up cards with embedded video games especially easily, but that's not even on my mom's radar.

Web application development can learn plenty from "software engineering" which in turn can learn a lot from regular old bridge/skyscraper/canal/tunnel engineering. Maybe when the pace of development slows down a little there won't be so much wheel re-inventing. How many thousands (tens of thousands?) internal web content management systems were built in the last ten years? (My personal contribution to this embarrassment: three) People who build bridges have had a few thousand years to develop some standards so perhaps my complaints come a little early. But the thinkers (artists, hackers, makers, whatever you want to call them) and the doers (engineers, builders, ditch-diggers, etc.) both need to realize that their common medium (computer programs) link them inextricably in the minds, hearts, and wallets of the 99.999% of the world that "just wants it to work".

* My own personal mom is actually somewhat computer savvy and is kind of interested in what happens behind the scenes. But the general abstract "mom" is such a universally beloved icon of naive computer usage that I don't want to rock the boat.

** Paul Graham has written about how Lisp was the Secret Weapon to make Viaweb/Yahoo! stores comparatively great. I've never maintained a Yahoo! store or written any RTML, but I've bought plenty of things from Yahoo! stores and elsewhere and haven't felt (as an end-user) a difference.

Are you an artist or a ditch-digger? Both? Neither?


2003-05-15 08:38:10
but you did notice that the system was there and worked. The difference between implementing in Lisp or another language may have been the features available on a given date. Not something you could feel by using the application.

Many applications only exist because someone wanted to explore and try new (for them)things. Not engineer to a spec.

2003-05-16 09:26:17
It's not either-or
You're probably right that I was too exuberant -- I tend to champion things that I think are underappreciated. But I also think that there's something beyond engineering in elevators, too. Elevators sprang from a vision of the possible. An amazing vision that transformed our society. That is a profoundly artistic act.

What's interesting to me is that what was once wonderful and memorable becomes taken for granted. This is true in everything from engineering to painting and philosophy. This was in fact one of the themes of the first extended piece of writing I did, my honors thesis at Harvard. I pointed out that a lot of the confusion about effervescent quasi-mystical passages in Plato's dialogues disappeared if you realized that the logical arguments we now replay with dry precision were once the edge of discovery, inflaming the participants with numinous excitement.

But your overall point is good. We need both art and engineering. And of course, as David Stutz pointed out, craft.

2003-05-16 09:55:49
It ain't easy
With a nod to Three Dog Night, writing efficient code isn't the fastest or easiest way to get an app out there to the people...and generating income for all the non-creatives who have an intere$t in it...

But we are at a junction in history where we could have the most impact upon future technological advances. Or, we can settle for "good enough", and throw it out there. I guess whether you are a coder, or an automobile manufacturer or a pothole filler-if you do the bare minimum, there will be a steady stream of service work down the road...enough to pay for all those well-built/well thought out/long lasting imported goods you have your eye on. Hmmm...

2003-05-16 11:30:45
It's not either-or
Yeah, I don't mean to explicitly demean elevators -- they have, as you point out, transformed our society. (And here in New York, riding elevators like those in the Flatiron Building makes it clear how much they've evolved since their invention.)

I am just trying to be mindful of the folks who think of computer programs the same way that I (most of the time) think about elevators. It's not their passion, so they're not too concerned whether there's a Turk-in-the-box or not as long as they get to their floor. They're not the ones who are going to go to computer program museums, read code for fun or enlightenment, or cuddle up with a biography of Alan Perlis.

In some sense, this is just ranty nitpicking on my part. Your weblog and my weblog and this entire site are not aimed at those folks. I don't think there are columns in Elevator World magazine telling people to tamp down their enthusiasm for Elevators. (Or, more appropriately, columns in Van Gogh Magazine telling people to cool it with the Van Gogh and admire housepainters.) However, there is an inescapable utilitarian aspect to programming that separates it from "purer" visual art forms like painting or sculpture which I think makes "satisfying users" an essential standard by which just about any program is judged. Crab on its Back can be a great painting solely because it creates an emotional response in a viewer (in my case, "Crabs are neat and I like those colors.") I think it's harder (but not impossible) to make the same case for an "artistic" computer program that is entirely divorced from satisfying some user requirement.

These issues are up for grabs, though, because as both you and Paul Graham point out, programming is in its comparative infancy. Will there be a "Museum of Computer Programs" one day? (Will it be a physical or digital museum?) Will elementary school students have "Clever Algorithm Appreciation" class along with "Art Appreciation" and "Music Appreciation"? I'm looking forward to it.