As I write this, I’m on a plane at 30,000+ feet, headed to Chicago. Seatmates include a couple from Toronto headed home from a cruise to Alaska. The husband and I talk technology a bit, and he mentions that his wife particularly enjoys sending letters as they travel. He and I both smile as we consider the novelty in 2014 of taking a piece of paper, writing thoughts to friends and family, and putting it in an envelope to travel around the world to be warmly received by the recipient.

Both Windows and Mac computers today are centered around the classic files and folders nomenclature we’ve all worked with for decades. From the beginning of the computer, mankind has struggled to insert metaphors from the physical world into our digital environments. The desktop, the briefcase, files that look like paper, folders that look like hanging file folders. Even today as the use of removable media decreases, we hang on to the floppy diskette icon, a symbol that means nothing to pre-teens of today, to command an application to “write” data to physical storage.

Why?

It’s time to stop using metaphors from the physical world – or at least to stop sending “files” to collaborators in order to have them receive work we deign to share with them.

Writing this post involves me eating a bit of crow – but only a bit. Prior to me leaving Microsoft in 2004, I had a rather… heated… conversation with a member of the WinFS team about a topic that is remarkably close to this. WinFS was an attempt to take files as we knew them and treat them as “objects”. In short, WinFS would take the legacy .ppt files as you knew them, and deserialize (decompose) them into a giant central data store within Windows based upon SQL Server, allowing you to search, organize, and move them in an easier manner. But a fundamental question I could never get answered by that team (the core of my heated conversation) was how that data would be shared with people external to your computer. WinFS would always have to serialize the data back out into a .ppt file (or some other “container”) in order to be sent to someone else. The WinFS team sought to convert everything on your system into a URL, as well – so you would have navigated the local file system almost as if your local machine was a Web server rather than using the local file and folder hierarchy that we had all become used to since the earliest versions of Windows or the Mac.

So as I look back on WinFS, some of the ideas were right, but in classic Microsoft form, at best it may have been a bit of premature innovation, and at worst it may have been nerd porn relatively disconnected from actual user scenarios and use cases.

From the dawn of the iPhone, power users have complained that iOS lacked something as simple as a file explorer/file picker. This wasn’t an error on Apple’s part; a significant percentage of Apple’s ease of use (largely aped by Android and Windows (at least with WinRT and Windows Phone applications) is by abstracting away the legacy file and folder bird’s nest of Windows, the Mac, etc.

As we enter the fall cavalcade of consumer devices ahead of the holiday, one truth appears plainly clear; that standalone “cloud storage” as we know it is largely headed for the economic off-ramp. The three main platform players have now put cloud storage in as a platform pillar, not an opportunity to be filled by partners. Apple (iCloud Drive), Google (Google Drive), and Microsoft (OneDrive and OneDrive for Business – their consumer and business offerings, respectively), have all been placed firmly in as a part of their respective platform. Lock-in now isn’t just a part of the device or the OS, it’s about where your files live, as that can help create a platform network effect (AT&T Friends and Family, but in the cloud). I know for me, my entire family is iOS based. I can send a link from iCloud drive files to any member of my family and know they can see the photo I took or the words I wrote.

But that’s just it. Regardless of how my file is stored in Apple’s, Google’s, or Microsoft’s hosted storage, I share it through a link. Every “document” envelope as we knew it in the past is now a URL, with applications on each device capable of opening their file content.

Moreover, today’s worker generally wants their work:

1. Saved automatically
2. Backed up to the cloud automatically (within reason, and protected accordingly)
3. Versioned and revertible
4. Accessible anywhere
5. Coauthoring capable (work with one or more colleagues concurrently without needing to save and exchange a “file”)

As these sorts of features become ubiquitous across productivity tools, the line between a “file” and a “URL” becomes increasingly blurred, and the more, well, the more our computers start acting just like the WinFS team wanted them to over a decade ago.

If you look at the typical user’s desktop, it’s a dumping ground of documents. It’s a mess. So are their favorites/bookmarks, music, videos, and any other “file type” they have.

On the Mac, iTunes (music metadata), iPhoto (face/EXIF, and date info), and now the finder itself (properties and now tags) are a complete mess of metadata. A colleague in the Longhorn Client Product Management Group was responsible for owning the photo experience for WinFS. Even then I think I crushed his spirit by pointing out what a pain in the ass it was going to be to enter in all of the metadata for photos as users returned for trips, in order to make the photos be anything more than a digital shoebox that sits under the bed.

I’m going to tell all the nerds in the world a secret. Ready? Users don’t screw around entering metadata. So anything you build that is metadata-centric that doesn’t populate the metadata for the user is… largely unused.

I mention this because, as we move towards vendor-centered repositories of our documents, it becomes an opportunity for vendors to do much of what WinFS wanted to do, and help users catalog and organize their data; but it has to be done almost automatically for them. I’m somewhat excited about Microsoft’s Delve (nee Oslo) primarily because if it is done right (and if/when Google offers a similar feature), users will be able to discover content across the enterprise that can help them with their job. Written word will in so many ways become a properly archived, searchable, and collaboration-ready tool for businesses (and users themselves, ideally).

Part of the direction I think we need to see is tools that become better about organizing and cataloging our information as we create it, and keeping track of the lineage of written word and digital information. Create a file using a given template? That should be easily visible. Take a trip with family members? Photos should be easily stitched together into a searchable family album.

Power users, of course, want to feel a sense of control over the files and folders on their computing devices (some of them even enjoy filling in metadata fields). These are the same users who complained loudly that iOS didn’t have a Finder or traditional file picker, and who persuaded Microsoft to add a file explorer of sorts to Windows Phone, as Windows 8 and Microsoft’s OneDrive and OneDrive for Business services began blurring out the legacy Windows File Explorer. There’s a good likelihood that next year’s release of Windows 9 could see the legacy Win32 desktop disappear on touch-centric Windows devices (much like Windows Phone 8.x, where Win32 still technically exists, but is kept out of view. I firmly expect this move will (to say it gently) irk Windows power users. These are the same type of users who freaked out when Apple removed the save functionality from Pages/Numbers/Keynote. Yet that approach is now commonplace for the productivity suites of all of the “big 3” productivity players (Microsoft, Google, and Apple), where real-time coauthoring requires an abstraction of the traditional “Save” verb we all became used to since the 1980’s. For Windows to succeed as a novice-approachable touch environment as iOS is, it means jettisoning a visible Win32 and the File Explorer. With this, OneDrive and the simplified file pickers in Windows become the centerpiece of how users will interact with local files.

I’m not saying that files and folders will disappear tomorrow, or that they’ll really ever disappear entirely at all. But increasingly, especially in collaboration-based use cases, the file and folder metaphors will largely move to the wayside, replaced by Web-based experiences and the use of URLs with dedicated platform-specific local, mobile or online apps interacting with them.