Google is not an OS
There was a time, 1999 probably, when we joked about printing up “Yahoo OS” tshirts. Too bad we never did, there was truth to it. There were a whole suite of hacked together, yet highly tuned and refined technologies, for moving content and personal data between ten of thousands of machines, deployed around the globe. Most of it built by people without PhDs, wow!
[Why does anyone care about all the PhDs at Google. The founders of Google quit their PhDs, just like Jerry and Filo at Yahoo. And in mundane similarity, Filo probably still drives a second hand Toyota Corolla, in that same mould of frugality as Sergey’s automobile choice. The even put two “o”‘s in the damn company name.]
Yahoo has the largest database of personal information in history. Yahoo knows what you search for, what movie theater you visit most, where you travel to, and other stuff you wish they didn’t know. Google is encountering the very different challenges of managing user databases. And when they’ve worked out all of the authentication, security, replication, and integration gotchas .. there’s still the question of what to do with that depth of personal information. That kind of database doesn’t enable automated personality deduction. Yahoo hasn’t managed to come up with anything, and neither has any PhD student anywhere. Computers aren’t that smart.
Of course, I want more storage for my Yahoo mail account. But how does that make an OS? Can I write a program for the Google OS? Google isn’t a walled garden like what Yahoo grew into, but it will never let outsiders install applications.
Now Google has a SOAP API. That’s not an OS either, but it points the way. Web Services, RIA, Semantic Web, the Blogosphere, are coming together to enable an Internet OS, built on amatuer application programming in recombinant growth. We’ll all be programming an OS that is everywhere.