Article

← Retour au blog

🇫🇷 Lire en français

22 years on the Web and the world is a billion times bigger!

2015-07-29 #981 · rév. 1 · EN

It would be a truism to point out how fast time flies.

Yet when it applies to one’s own journey, taking a minute to look back at what has happened since discovering the World Wide Web in 1993 is well worth it.

In 1993, after founding two custom PC assembly shops in the south of France, I joined the video game industry in Paris to direct STORM for an American publisher.

Back then, Mosaic was the only Web browser around. Websites were rare, even in the United States.
HTML was rudimentary, and there was no JavaScript. Performance was modest, and nobody was talking about Web Apps or OPAs (One-Page Applications) yet.

Twenty-two years later, we develop software with functional-capable languages: JavaScript and Swift, to name just two.
Hardware has changed dramatically as well: from the isolated PC—or weakly connected, to be more accurate—and the living-room console (the Sony PlayStation or Sega Saturn of that era), we have moved to a highly mobile (and therefore constantly connected) world where the edge case to handle today is a partial loss of network connectivity!

Storage and computing power have been abstracted to such an extent that we barely worry about them anymore: we even say they live (lost?) somewhere up in the cloud…

Microsoft Azure and Hadoop have become commonplace. Big Data algorithms allow us to parallelize data processing, cutting perceived processing times down to a few milliseconds regardless of data volume: the petabyte (10¹⁵ bytes) has become the benchmark unit of 2015, whereas the megabyte (10⁶) was that of 1993.

In 22 years, we have multiplied instantaneous addressable memory capacity by a factor of 10⁹! That is a billion times more…

It is fabulous.

So much so that today, it is becoming feasible to run Big Data workloads client-side right on an iPhone 6…
Indeed, by adapting the technology deployed back in 2006 on Mobilezoo.biz for J2ME handsets to the 2015 iPhone ecosystem, my research into client-side MapReduce algorithms, paired with CloudKit for persistence, now makes it possible to perform semantic text processing across multiple sources—including the Web—on a galaxy of thousands of iPhones.
The immediate benefits: a drastic reduction in infrastructure costs and the pooling of consumed MIPS by distributing them across an app’s user base.

Stay tuned for the rest of the adventure…