Caffeine Lab

Recently I’ve been to 2 events organized by companies advertising their APIs, their platform: Facebook and Google.

The Facebook event was pretty exciting, not because of any technical prowess but because of what it enables to do: write applications using an existing, big, friend network (what they call proudly the “social graph”). It’s cool because if I want to create, let’s say, an application to rate and review restaurants, each user can get the reviews from his friends without having to convince them to create an account here and mark me as friend (and duplicate their friend network). It’s also exciting that they allow people to make money on that, either by showing their own ads or by getting paid by users.

The Google event was quite different. While Facebook was launching their platform and focused on this one product, the Google developer day was much bigger with a broad range of topics. The three main topics were:

From these two events it became pretty clear that a lot of companies are fighting today to become the next platform to write web applications. Just like a lot of desktop application were just hacked in a few days in Visual Basic, it seem to be a common believe that more and more web application will just be mashups based on bricks provided by Google, Facebook or Yahoo. If one company controls the basic bricks, this company may just end up being as powerful as Microsoft was at the end of the last century.

You want to use this (non-Google) cool application? Log in to your Google account. Wanna play with this new social web app? Log in to Facebook.

I finished the Google Developer Day chatting with a Googler of the Gears team. He had an interesting comment about the direction software development was going: “When I started in the business I was worried about my future, I was wondering what could become my career in 20 years if I can’t keep up with new technologies that don’t exist yet. Now I’m no longer worried: all the young developers today, who write mashups and ajaxy stuff, have no idea of the difference between a hash and a b-tree and when use one or the other. And we’ll always need engineers who can make the difference.”