My college roommate Eric sent me a link to Drew Houston’s keynote presentation from the inaugural Dropbox developer conference (DBX). While I’m a little sad I didn’t get to attend (*ahem* friends who work at Dropbox who didn’t invite me), I was very pleased to find that some of the major themes of the keynote were very much aligned with some of the things I’ve blogged about recently: mainly, putting together a web file system and creating better synchronized application experiences. While I still have questions about the details on the feasibility of the merge-less syncing back-end they were promising, I hope app developers everywhere pay attention: while DBX might not have gotten as much attention as Google I/O or WWDC, what Dropbox announced is equally important in terms of building the types of services that users will want.Leave a Comment
Tag: web file system
Last week, Google unveiled its long-rumored Google Drive product with great fanfare. While the gaggle of tech journalists/bloggers issued predictable comparisons of Google’s new service with online storage/syncing services like Dropbox, I couldn’t help but think that most of the coverage missed the point on why Google Drive was interesting. Yes, its another consumer-facing cloud storage service – but the really interesting aspect of it is not whether or not it’ll “kill Dropbox/Box.net/iCloud/[insert your favorite consumer cloud service here]”, but the fact that this could be the beginning of a true web “file system”.
I’ve blogged before about the strengths of the web as a software development platform and the extent to which web apps are now practically the same thing as the apps that we run on our computers and phones. But, frankly, one of the biggest things holding back the vision of the web as a full-fledged “operating system” is the lack of a web-centric “file system”. I use the quotes because I’m not referring to the underlying NTFS/ExtX/HFS/etc technology that most people think of when they hear “file system”: I’m referring to basic functionalities that we expect in our operating systems and file systems:
- a place to reliably create, read, and edit data
- the ability to search through stored information based on metadata
- a way to associate data with specific applications and services that can operate on them (i.e. opening Photoshop files in Adobe Photoshop, MP3s in iTunes, etc)
- a way to let any application with the right permissions and capabilities to act on that data
Now, a skeptic might point out that the HTML5 specification actually has a lot of local storage/file handling capabilities and that services like Dropbox already provide some of this functionality in the form of APIs that third party apps and services can use – but in both cases, the emphasis is first and foremost on local storage – putting stuff onto or syncing with the storage on your physical machine. As long as that’s true, the web won’t be a fully functioning operating system. Web services will routinely have to rely on local storage (which, by the way, reduces the portability of these apps between different machines), and applications will have to be more silo’d as they each need to manage their own storage (whether its stored on their servers or stored locally on a physical device).
What a vision of the web as operating system needs is a cloud-first storage service (where files are meant to reside on the cloud and where local storage is secondary) which is searchable, editable, and supports file type associations and allows web apps and services to have direct access to that data without having to go through a local client device like a computer or a phone/tablet. And, I think we are beginning to see that with Google Drive.
- The local interface is pretty kludgy: the folder is really just a bunch of bookmark links, emphasizing that this is a web-centric product first and foremost
- It offers many useful operating system-like functionality (like search and revision history) directly on the web where the files are resident
- Google Drive greatly emphasizes how files stored on it have associated viewers and can be accessed by a wide range of apps, including some by Google (i.e. attachments on Gmail, opening/editing on Google Docs, and sharing with Google+) and some by third parties like HelloFax, WeVideo, and LucidChart
Whether or not Google succeeds longer-term at turning Google Drive into a true cloud “file system” will depend greatly on their ability to continue to develop the product and manage the potential conflicts involved with providing storage to web application competitors, but suffice to say, I think we’re at what could be the dawn of the transition from web as a software platform to web as an operating system. This is why I feel the companies that should pay more close attention to this development aren’t necessarily the storage/sync providers like Dropbox and Box.net – at least not for now – but companies like Microsoft and Apple which have a very different vision of how the future of computing should look (much more local software/hardware-centric) and who might not be in as good a position if the web-centric view that Google embodies takes off (as I think and hope it will).2 Comments