AWS S3 - What Amazon Got Wrong

TechCrunch (and others -- including Rob Hof of BusinessWeek and Mike at Techdirt) is reporting the launch of AWS S3, Amazon Web Services' new product - Simple Storage Service. Essentially, Amazon wants to make building scalable web apps easy by taking the pain out of building an infrastructure for data.

Unfortunately, they've done two things wrongs so far.

Problem 1. It still requires infrastructure

The first problem is fairly simple to explain. But, it's a problem that's really hard to solve: Transferring all this data requires a lot of infrastructure. Amazon made it easy to store the data, keeping it safe and redundant and reliable, but you still have to get it to and from Amazon. Adding in the traffic of getting the data to and from users, and this still requires some infrastructure.

This isn't an Amazon specific problem, but it is a big problem. Unfortunately, Amazon's pricing model may have made this problem worse.

Problem 2. Data silos still exist

In my opinion, this is where Amazon got it really wrong (and Openomy got it really right). With Amazon's S3, you can specify Access Control, meaning "this file belongs to this Amazon customer and no one else" (as identified by email address). That's nice, but it's then up to the web app to make sure that people who register on their site use their Amazon email. Meh, that's annoying.

The large hurdle is that S3 doesn't allow the intertwining of each user's data from all the different web applications they use. That means that your data is still totally locked out and unavailable to you!

In short, Amazon just moves the problem down a level, to their servers -- organizing files by application, rather than by user.

This means that your data is yet again not made available to you when you want it where you want it. You're still locked in to the application. If the web app doesn't want to give you access to your data, you have no access to it. If the web app shuts down without notice, your data is gone.

This is where Openomy shines. Rather than store files based on applications, Openomy stores files based on the user. This gives the user access to his/her data at any time, ever. As a user, you own your data on Openomy.

By doing so, Openomy completely eradicates Problem 2. There are no more silos of your data. And, in the process, Problem 1 becomes less interesting because the benefit outweighs the harm.

Conclusion: Amazon S3 is very, very interesting. It just happens to fall short in a couple areas. It'll be really fun to watch how this space turns out over the next few months and years.

Tags: web 2.0, Amazon, AWS, S3, TechCrunch, Openomy