Computer Networks and Applications: Open or Closed?
Computing and networking practices move in cycles. In the early days of computing, processing occurred in a few large mainframe computers, with users dialing in using "dumb" terminals. The PC revolution of the 1980s made it possible for simple tasks to be performed locally, on the desktop, while more complex computations took place on minicomputers. As computers became more powerful, workloads shifted increasingly to the desktop. In the late 1990s my office desk featured both a Pentium-based Dell Optiplex, which I used for document processing, email, and web browsing, and a Sun Sparcstation connected to a departmental server, which I used for statistical work. (Now I can't remember the last time I saw a Sparc workstation, except at a museum.) As internet connections at home and work became faster, 3G and 4G cellular networks became more common, and wifi hospots became ubiquitous, documents, data, and processing moved back to servers — as we say today, "to the cloud." Plus ça change, plus c'est la même chose.
Something similar is happening with software and services. The early computer languages and applications were frequently proprietary or "closed." The web encouraged open standards, transparency, and free software; e.g., information would be encoded in HTML that could be rendered by any browser running on any operating system. Apple, however, threw a giant spanner into the works by popularizing the closed ecosystem model associated with iOS. Today, people largely access information via apps, not browsers, apps that are often proprietary, specific to particular operating systems and devices, and part of one company or consortium's closed ecosystem.
Writing in Monday's Wall Street Journal, Christopher Mims says the closed, app-based model is killing the internet.
The Web was intended to expose information. It was so devoted to sharing above all else that it didn’t include any way to pay for things. . . . The Web wasn’t perfect, but it created a commons where people could exchange information and goods. It forced companies to build technology that was explicitly designed to be compatible with competitors’ technology. Microsoft’s Web browser had to faithfully render Apple’s website. If it didn’t, consumers would use another one, such as Firefox or Google’s Chrome, which has since taken over.Today, as apps take over, the Web’s architects are abandoning it. Google’s newest experiment in email nirvana, called Inbox, is available for both Android and Apple’s iOS, but on the Web it doesn’t work in any browser except Chrome. The process of creating new Web standards has slowed to a crawl. Meanwhile, companies with app stores are devoted to making those stores better than — and entirely incompatible with — app stores built by competitors.
(A nongated version is here; see also responses here and here.) The app-based world, Mims claims, is "a system that makes innovation, serendipity and experimentation that much harder for those who build things that rely on the Internet."
I'm personally not a fan of Apple's walled-garden approach. But who cares; commercial computer and internet technologies and platforms were not designed to satisfy my personal tastes, but to earn profits, which they can only achieve by satisfying large numbers of consumers. As I noted in a 2006 article, writers like Mims romanticize, the "free," open, nonprofit (i.e., government-funded), commons-based, peer-oriented Internet that prevailed from the 1960s to the 1990s. Academics like Yale's Yochai Benkler make the same mistake. In truth, there is no single "optimal" model for the Internet, or for any valuable technology. We should allow entrepreneurs to experiment with different approaches and to compete for patronage, and not have academics, journalists, or regulators pick the "correct" model or platform. If the proprietary, app-based approach proves successful, good for it. If consumers prefer more open and transparent approaches, they are free to patronize existing and emerging entrepreneurs and organizations that provide such products and services. Let's let a thousand standards, platforms, technologies, and business models bloom!