Heroes of Software Engineering - Drew Houston, Arash Ferdowsi and the Dropbox engineering team

Posted by s.hettrick on 18 November 2013 - 1:26pm

MailBoxes.jpgBy Ian Cottam, IT Services Research Lead, The University of Manchester.

This is the eighth - and likely the last - in my short series about heroes of software engineering, so I will write about something different. Readers may have noticed that several of my previous heroes are now in their 70s or even 80s, so today instead features some of the current generation of innovators. These are the people behind Dropbox, including founders Arash Ferdowsi and Drew Houston and the men and women of the Dropbox software engineering team.

Think Sync!

I first started using Houston and Ferdowsi’s software in early 2009, and I was immediately impressed. The key for me was the way their design concentrated on the synchronisation problem of files across devices. In addition, you can also store your files on your own machines, with the exception of smartphones and other mobile devices.

Dropbox works so reliably and efficiently that I often switch between devices from one meeting to the next, safe in the knowledge that my notes and file changes are always there and bang up to date. The ease of sharing folders with collaborators around the world is also a killer app. People who opt to pay for the so-called Packrat feature (unlimited undeletes) can also use the service as a backup system, but that is secondary to the advantages that come with sync.

Dropbox is simple to use, with all of its considerable complexity hidden under the hood. This is classic software systems design, sharing much of in common with the Apple design philosophy, although Apple’s own iCloud may be now be too implicit for my tastes. Dropbox also manages user feedback in an effective manner, with its forum full of requests for new features. Yet the engineers at Dropbox ignore most of them, and quite wisely in my humble opinion. Feature creep is one of software's worst blights, and it is very hard to get rid of once you, or your software, catches it.

For example, Dropbox leaves encryption up to the users, who can choose from a wide range of compatible third-party encryption systems instead. (My preference, and the one we are trialling at Manchester, is the excellent nCrypted Cloud). With that in mind, I can only say that Dropbox was not only one of the first of its breed, but to me it is also far and away the best.