Open vs. Closed – A late-night thought experiment
Intel and AMD have been running each other ragged, and racking up legal settlements that surely have an effect on consumers, as the debate rages on as to whether open or closed is the correct religion, so to speak, by which to run a company. Back in the day, when Intel was selling its 286 chip to Microsoft, Microsoft demanded an additional supplier. So Intel went out and licensed the x86 to AMD, a rather “open” move that AMD has since exploited for processor after processor. Reflecting on AMD’s innovation (or lack thereof), scholars have published such works as “Congratulations, It’s a Clone!” Intel will be dishing out $1.25 billion to AMD in the coming years, the cost of settling all antitrust suits, but to what avail? The singular open move Intel made with its coveted x86 technology has cost them decades of clones. Sure, they’re still the worldwide leader in x86, but with AMD climbing into a 20% market share, haven’t they already paid their dues? These days, Intel is significantly less liberal with their licenses. They’ve signed up for the “closed” religion and they look unlikely to turn back.
On the flip side, RISC architectures that garnered any respect did so by massive licensing. Early 90′s licensing of Sun’s SPARC and MIPS’ technologies were seemingly the only way to enter the market, supposedly to get software developers to bite. Which draws a comparison to the exploding role of developers today in the mobile market. While RISC went open, and Intel paid out in significant sales loss in its one open move, where are we in the battle between open and closed?
From the marketing perspective, history teaches us that its better to get your users hooked and make them pay later, than have them dish it out up front. Gillette and his disposable razors were sold at a loss; Coca-Cola started shoving cokes through the Berlin wall as soon as they punched the first hole – all for free. And all to get consumers hooked. Though these architectures generally run a course below the radar of the average consumer, even the Pentium had its limelight, when a floating-point bug hit the internet and Intel learned the meaning of “No press is bad press.” How much are these settlements costing me? And would an open architecture fuel innovation? Or stifle it?
In the summer of this year, Google reprimanded a hacker by the name of Cyanogen. His modifications to the Android OS were fast, and incredibly popular. Though Google’s OS is entirely open-source, Google’s legal department went after him for the release of Google’s Gmail, and other popular Google built and branded apps native to the OS. Yet these apps are also available on the Android market free of charge. At what point do we draw the line, even in the most open of architectures, between mine and yours?
That is to say – how open is good, how successful is closed, and how can we counter-balance the two to fuel adoption and keep costs low?