The next Web revolution

The Web celebrates its 10th anniversary and it's still a pain to use -- clunky, slow and unresponsive. But thanks to creative small companies like Chicago's 37 Signals, the Web is finally becoming as fun and flexible as your favorite software.

Published August 10, 2005 7:38PM (EDT)

Odes to the World Wide Web inevitably burst with superlatives. The Web is the biggest, the fastest, the most addictive thing ever. The Web will revolutionize this, supplant that. The Web will set you up on the best date you ever had and the sex will be out of this world. Just now, as we celebrate the 10th anniversary of the day the letters "WWW" hit the big time -- Netscape went public on NASDAQ on Aug. 9, 1995 -- the superlatives are flying particularly fast. In the August issue of Wired, founding editor Kevin Kelly predicts that 3,000 years from now, people will regard the development of the Web as a pivotal moment in human history, as important as the advent of American democracy and the world's major religions. "This will be recognized as the largest, most complex, and most surprising event on the planet," he writes.

Boosters like Kelly are no doubt on to something; taking the long view, the Web may well alter the course of history. But let's cast a frank eye on the present day's surf report, shall we? The Web at age 10 is a pain. Precious little online works as well as it should. Compared to the speed of desktop software -- such as your e-mail program, or iTunes -- using many Web sites, even the biggest and most popular, is like swimming through mud. Think about the features you take for granted in iTunes: buttons that respond as soon as you click them (not five seconds later), a search bar that begins to work at the instant you type, playlists that can be rearranged by dragging and dropping. Almost nothing online works as naturally; you wouldn't even dream of managing your music with a Web tool. On the Web, to attempt anything complex -- even to write a blog post -- is to flirt with disaster, or at least annoyance.

Yet I aim not to gripe, but to offer hope. In recent weeks, I've been talking to many clever people who are using creative programming techniques to build a better World Wide Web. The online experience they envision is more responsive than the Web we use today, and it's more useful and fun, too. On this better Web, you can drag and drop items to rearrange them, see a search box fill up while you type a query, and prompt an action as soon as you press a button. The model works, in other words, as intuitively as the best software in our lives. You've likely seen bits of it already. These new techniques power Gmail, Google's fine Web e-mail system, allow you to drag maps in Google Maps, annotate pictures in Flickr, and use your mouse to reorder your movie queue in Netflix.

In addition to better software, I discovered something else about the new Web: Creativity is back. The idea that the Web is a giant get-rich-quick vehicle no longer pervades the business. Instead, recalling the mid-1990s, a host of truly talented people are looking at the Web as a canvas for their creativity. And there's one small company that's emblematic of this effort to build better applications, and, indeed, is pioneering an entire business philosophy designed to make the Web great. The firm is called 37 Signals, and if you've never heard of it, don't worry. You're likely to start using its software any day now.

37 Signals is named after the number of radio waves we've received from space that scientists consider potential signals of intelligent life. Its creators build the kind of applications you didn't know you needed until you use them for the first time, at which point you wonder how you ever did without. Last year the company created Basecamp, a Web-based project-management tool unlike any project-management tool before it. If you've got a many-person task to do -- any big project, from redecorating your house to redesigning your home page, planning your wedding to planning your wake -- Basecamp gives all participants a central spot on the Web in which to plan and discuss the endeavor. The software has been adopted by hundreds of advertising firms, law firms, Web designers and book publishers.

More recently, 37 Signals launched Backpack, a program that does just what its name suggests -- it gives users an easy, casual storage location on the Web, a place to scratch down important notes, draw up to-do lists, and store important files organized around specific tasks (say, all the stuff you need for a business trip). The Wall Street Journal has praised Backpack as the best tool of its kind, and perhaps more important, bloggers have been jumping for joy over it. Lifehacker, a blog that offers tips to help keep your life in order, calls the software "a perfect online replacement (or supplement) to that fancy notebook you've been scribbling in."

Basecamp and Backpack represent the future of software on the Web not just because they're elegant, easy-to-use programs that will likely make your life better. The two applications are also interesting because they were created in a novel way, using a new programming model that allowed 37 Signals to build each program very quickly, and with very few people. Indeed, this method of creating applications -- doing it fast and on a tight budget -- might well be called 37 Signal's animating philosophy, its central mission.

"We have this big thing about embracing constraints," says Jason Fried, the company's founder. "When you have constraints -- less time, less money -- people care about every dollar they spend. Customers ask us, 'How does Basecamp compare with other project-management tools?' We say it does less. Our products do less, and that's why they're successful. People don't want bloated products, and constraints force us to keep our products small, and to keep them valuable."

Fried founded 37 Signals in Chicago in 1999, which, for the coast-dominated tech industry, is known as a long time ago in a galaxy far, far away. At first, the company worked on Web design. Given its distance from the epicenter of the industry, it brought a novel sensibility to the business of creating Web pages. The company's own site from the time suggests that even during the boom -- a time when nobody in the tech world advocated minimalism, when even the sock puppet had a book deal -- 37 Signals was devoted to less is more. The company's home page was mostly white space, the main feature a list of links to 37 "nuggets of online philosophy and design wisdom." Here's one representative nugget: "What drives us is the knowledge that the everyday person is seeing only the smallest glimpse, 1/100th at best, of the full potential of interactive media today. To most folks, the Web is a scary place. Our mission is to change that perception."

The story of how 37 Signals morphed from a Web design firm that built sites for businesses into a Web software company that builds applications for regular people is reminiscent of the Native American legend about Indian tribes who found a use for every part of the slain buffalo. Here's the quick version: In the course of creating Web pages for businesses, 37 Signals realized it needed a tool that would give its clients an easy way to monitor progress on their designs. The tool it created, a Web-based program meant only for the firm's internal use, was a hit with clients, many of whom wanted to use it for other projects at their offices. So Fried decided to transform the internal program into an application for everyone. In 2003, David Heinemeier Hansson, a programmer who lives in Copenhagen, Denmark, joined the firm to help with the task, and shortly thereafter 37 Signals released Basecamp to the public.

But in creating Basecamp, 37 Signals saw there were still other internal tools it could give to the world. Experienced programmers use a variety of programming languages to build systems for the Web -- you may have heard of Perl, Java, PHP and others -- but for Basecamp, Hansson used a relatively unknown programming language from Japan called Ruby. This turned out to have been an inspired choice, since it allowed him to create all the code he needed to run a Web application -- what programmers call a "development framework" -- from scratch. Hansson saw this framework would not only help 37 Signals build other programs, it would also prove handy for other developers. So last summer, 37 Signals released the framework as an open source software project called Ruby on Rails (about which more in a minute).

Launching a well-liked application like Basecamp and a complete development framework of Ruby on Rails in one year wasn't enough for 37 Signals. In selling Basecamp, the company learned that customers loved the application's to-do list feature. So this January the firm built a stand-alone to-do list site, a service that allows anyone to create a quick list of tasks, for free. The site, called Ta-da List, offers a good peek at the kind of software 37 Signals creates: easy, well-designed, highly functional small apps that ought to come with an addiction-danger warning.

Here's another amazing thing about 37 Signals: Only five people work there. There's no ad-sales department, no marketing team, no H.R. department, no tech support crew (Fried handles all customer questions himself), and no receptionist (there is an office in Chicago, but only Fried and another employee, Ryan Singer, work there; the other three people are in Utah, New York and Denmark). That's what I mean about using every part of the buffalo. The company created all it did in a short time with very little start-up money -- Fried eschews venture capitalists -- and other resources. Instead, it put a premium on its experience, constantly looking for creative new ways to spin what it learned on one project into another one. The M.O. has paid off. Today, 37 Signals owes no money to early investors. Because the company is a private firm, its exact financials are unclear. But the picture is appealing. First of all, the company makes money from its Web applications. To use Basecamp, customers pay a monthly fee of either $24, $49 or $99, depending on the number of projects they manage, and $19 a month for Backpack (there are free versions as well). The firm also does occasional Web design projects and hosts design conferences. Fried says the company is making a profit.

Today, notes Fried, starting a tech company requires very little in fixed costs. Most hardware and software (stuff to host a Web site, for instance) is either free or almost free. Standard business processes like handling accounts and marketing are built into the Web. If you want someone to pay you it's just a matter of setting up a Paypal account. If you'd like to advertise your site, you can buy an ad on Google, or somehow get bloggers to talk about you, raising your profile in Google. (37 Signals, which maintains a popular blog, got a great deal of buzz in the blog world.) "Your main cost is really labor," Fried says, and if you're passionate about what you're doing, if you're willing to go six months without a salary, that's an avoidable thing as well.

The firm has also benefited from its use of Ruby on Rails. A good way to understand the value of Ruby on Rails is to think of it as something like Lego blocks of code -- discrete pieces of programming bits that can be put together in ingenious ways to create Web programs easily. Most Web applications, even ones radically different from each other, need to do a basic set of tasks to accomplish anything. For instance, every Web program needs to have a way of getting input from the page a user is looking at, and of doing something interesting with that input (say, adding it to a database). Ruby on Rails has all of these functions built into it; in order to build basic steps, says Hansson, programmers just need to use prefab parts already built in, not spend their time writing rudimentary code.

There's another advantage to programming with Ruby on Rails: It makes AJAX easy to use. AJAX is an intriguing -- if much-hyped -- programming method that makes Web pages work more like desktop programs. On most Web sites, pressing a link or a button is a two- or three-second affair -- every time you press something, you've got to wait for the entire page to reload before you can do anything else. With AJAX, pressing a button causes an action to occur on the page more-or-less immediately, without having to reload. The name, which was coined by Jesse James Garrett, a designer at the San Francisco Web design firm Adaptive Path, is an acronym for Asynchronous JavaScript + XML; it essentially describes a way for a Web page to talk to a Web server in the background, without having to inconvenience the user in the process. AJAX is the magic in slick-looking applications like Google Maps, in which a map of the world can be moved with the click of a mouse. In older mapping programs, such as Yahoo's, you'd have to wait for a page to reload each time you wanted to look at another part of the map.

These two features -- the reusability of its code, and built-in Ajax -- have helped to make Ruby on Rails increasingly popular among Web developers. Robot Co-op, a small Seattle company behind the popular sites 43 Things and 43 Places, developed its sites in Ruby on Rails. So did Odeo, the new podcasting company founded by Noah Glass and Evan Williams, one of the creators of Blogger.

"When I talk to developers about Ruby on Rails, they're like, 'This is the language I would have designed,'" says Jeff Veen, a pioneering Web programmer and one of the co-founders of Adaptive Path. Adaptive Path usually creates sites for other companies, but -- following something of the model 37 Signals used in producing Basecamp -- Veen has recently put together a small team of developers to create a Ruby on Rails application that the company plans to release to the outside world (the program, a tool to help bloggers measure traffic and other stats on their site, will be out by the end of the year, Veen says). Several other developers also attest that Ruby on Rails makes programming Web apps so easy that good ideas for Web programs are now within reach. "I've had some ideas for applications running around in my head for a while now," says Rael Dornfest, chief technology officer of the tech book publishing firm O'Reilly. "Until now, they would have been prohibitively difficult to create in terms of time and structure. What Ruby on Rails has allowed me to do is express ideas in code more easily than it would have been without the framework."

Expressing ideas in code is an apt description for what many new Web developers seem to be doing these days. What stands out about 37 Signals -- as well as Adaptive Path, Odeo, the Robot Co-op, and a host of other successful Web firms -- is the passion it has for new ideas. These people, you get the sense, truly understand the flexibility of the Web and are delighted by the power they possess to make it better.

Of course, they're not looking to do it for free. But there's no expectation of riches, either -- or, more interestingly, there's a sense that riches can actually damage the quality of the software. "The big point is personal satisfaction and enjoyment," says Josh Petersen, one of the founders of Robot Co-op (which has received investment money from "I've worked at big companies and big development teams and I don't find it enjoyable. Now, one of my greatest joys is sitting at the same table with everyone else here, and getting to use an Apple computer at work."

Every day, it seems, you hear stories about how Americans will have an increasingly difficult time competing in the global marketplace. Talking to someone like Fried -- or Petersen, or other new Web entrepreneurs -- prompts optimism. The Web is 10 years old. It's basically untouched. With so many people now free to build their good ideas onto it, is it any wonder Kevin Kelly thinks they'll remember us fondly in 3,000 years?

By Farhad Manjoo

Farhad Manjoo is a Salon staff writer and the author of True Enough: Learning to Live in a Post-Fact Society.

MORE FROM Farhad Manjoo

Related Topics ------------------------------------------