Image - Begging You by John Squire¹

Earlier this month, Dropbox raised about $250 million in a funding round that values the company at $10 billion. This means that the value of the company has doubled since 2011.

For me personally, the value of Dropbox is constantly increasing. The more I use more than one device, the more useful Dropbox's syncing service becomes.²

But its the problem that they solve for me — which goes some way beyond "putting my stuff on the internet" — that I find more interesting.

Strange artefacts found in a @ft newsroom drawer labelled "stats": pic.twitter.com/Q5htAlsFxG

— Martin Stabe (@martinstabe) August 16, 2013

Harry McCracken wrote an article last year about the "Save" icon, and the peculiarity that the floppy disk is still the universal symbol for 'saving' a document.

This is something I've written about before – actually, I don't think the icon is the real issue, so much as the actual act of 'saving' being obsolete. It isn't the icon that we need to get away from — its what the icon represents. The action of saving your work.

The problem of "save"

When you "save" something, you are putting something from the memory on your computer onto a disk, as a file. Simple enough, but it raises a couple of issues;

Something you have saved is generally considered 'safe'. So something you haven't saved is 'unsafe' – it could be easily lost by a crash, accidentally shutting down (or accidentally clicking "no" on a dialog box that asks if you want to save when you are hurriedly trying to shut down your computer…)

That's why you "save" it.

So far, so simple.

Back in the early personal computer days, hard drives were a luxury³ – floppy drives were an essential — something you saved was a tangible "thing", which you could then stick into a different computers, give to a friend or colleague — even put in an envelope and send it across the world. The physical "container" was the media — you could stick a label on it, share it, copy it, file it away in a box marked "stats"…

Now, floppy drives have been pretty much replaced by hard drives, USB sticks, email attachments, online services (like Dropbox, iCloud, Google Drive etc. etc.) But the basic paradigm for using a PC hasn't really changed too much. You work with something "unsafe" in memory, and then save it somewhere when you're done (clicking on the icon that probably has no meaning to anyone under the age of about 20.)

The workaround to this "unsafe" working file when you have a permanently attached hard drive is 'autosave' – have your application save your work for you on a regular basis, so the worst that could happen is that you lose 5-10 minutes worth of the work you did since the last autosave. At least, thats the theory — I'm sure we have all experienced what happens in practice when your unsaved "Document1" you spent all morning on vanishes without trace – maybe because you were working from a version you opened from an email (so saved somewhere mysterious, and disappearing when the email closes), or maybe because of an issue with a network drive, or some other gremlin that stops autosave from being the catch-all that we wish it was.

The benefit is that — when it works — you don't really need to Save any more — just click on "yes" when you are asked if you want to save when you close down at the end of the day. (Although why you get that question when your work is autosaved isn't clear…)

But - if you are saving as the same file, in the same folder, on the same disk, you lose the old version of whatever you are working on. Even if, for example, you delete everything in your working document and start again with a blank canvas. (Perhaps you — or the organisation you work for — doesn't understand how to use Powerpoint templates properly, for example…) In other words, what you have 'saved' might not actually be 'safe' — and you might not find out about it for some time. And while the "Trash" folder works fairly well for a single user at a single computer, rescuing inadvertently deleted or overwritten files from a multi-computer, multi-user network drive presents a whole new set of challenges.

Combining these last two points introduces a new problem – its possible (actually quite easy) to think that you are wiping work from a new (ie. unsaved) document, and then have your computer – or someone else – wipe it from your 'saved' version.

The new problem of "Sync"

But that isn't the problem that I face on a near-daily basis (which Dropbox – if I remember to save – solves.) Its when I save something on my computer at work that I want to do something with on my computer at home.

Few of us work with a single device, so saving in a single place isn't as useful as it used to be. Phones and tablets tend not to work well with USB sticks, so a physical device (floppy disk, memory card, USB stick etc.) isn't as much help as it used to be. Soon (if not already), we will want all of our applications and data to be available to us everywhere – meaning across different devices. Shortly after wanting them everywhere, we will expect them everywhere.

So "saving" actually becomes a part of a different, broader problem; "syncing." If I save something, I want it to be available on any of my devices that could possibly open it (ie. which has a relevant app.) Whether that is through Dropbox, iCloud, Simperium or any other syncing service is really just a minor detail here; the point is, I want all of my devices to be in sync with one another - I don't want to edit the same document on two devices and find myself with three different versions (an original, a version edited on device 1, and another version of the original edited on device 2 — potentially clashing with the version from device 1.)

Unless you can do everything you want to do on a web-based application (ie. the idea of "saving to disk" is actually happening on a single "drive" somewhere in the cloud), this is a much simpler problem to describe than to solve.

If you combine this "sync" issue with an additional "save" issue (ie. I haven't saved my work on Device 1 before I then edited the same file on Device 2 — or worse, the work I did on Device 1 and didn't save, I then need to access on Device 2) then you are pretty much guaranteed to face this kind of problem. (And the more devices you have, the worse the problem gets.) Throw in an interrupted network connection (for example, a mobile device starting to save something just as it loses signal) and you've got an additional layer of complexity.

So I don't think its really possible to work in a world where several devices work together on the same files and have "save" as an option (at least, not unless you actively shut down everything you were doing before finishing with one machine and moving to another.) You need to have something that does it for you in the background. Save shouldn't be an action that you do – it should be an ongoing process that your computer looks after.

Similarly, "sync" works much better as a background process that you don't need to think about than an active process where you have to press a button (or plug in a cable…)

A fresh start

The thing is, when you stop to think about this kind of stuff, its seem so obvious that the way mobile devices work is the best way to go. Everything happening in the background, automatically, without any fuss, on a device that (almost) never needs rebooting.

Which is fine, except when it isn't — sometimes, an application won't save something unless you explicitly choose to save it, which can be a problem if your work is saved in memory, then when you switch away to a different app and your "working" app gets shut down in the background and your work gets lost. Sometimes using two different applications with two different 'save' paradigms can be confusing (ie. autosave on one leads you to forget about saving on the other.)

But the broader problem is that we seem to want our 'computers' to work the same way that they always have done — with "Save" and "Save As" depending on whether we want to start a new file or save our work, with Autosave chugging along in the background, doing its best to watch our backs.

At the moment, we are in the middle of a transition between a 'desktop PC' world and a 'mobile first' world – and this kind of issue is one that, sooner or later, will need to be resolved.

The Microsoft world seems to be heading for a 'one operating system for any device' idea – although, for now at least, mobile phones don't seem to be a part of that vision. The danger here is that we move to a different mental model of 'computing' at different speeds – when everyone used Windows, Microsoft led the way. But when Windows is just one of the alternative, suddenly they need to strike a balance – change quickly enough to keep pace with the competition, but not so quickly as to alienate their comfortable, loyal users.

A recent episode of the Cubed podcast talks about Microsoft's leadership position changing over time – Ben Evans talks about the 'old' PC market, where a dozen or so PC makers ran low margin businesses, outsourcing their industries' innovation to Intel and Microsoft. Now (something I've noted before), their position is very different – the platform wars have moved away from just the PC and into Mobile devices, where Microsoft (and Intel's) advantages haven't given them nearly as much of a benefit as might have been expected – potentially undermining their position of dominance on the desktop…

Meanwhile, Apple seem to prefer the idea of iOS being a 'vision of the future', with Mac OSX gradually adapting some of its ways of working – a much slower march towards a single user interface idea working across different kinds of devices, not even necessarily converging on the "One OS", but a march where Apple controls the pace.

Perhaps the wild card here is Google – who would prefer to see the 'single platform' moving from the device/operating system and into the web browser (where platform agnosticism has been the idea since the inception of the web.) But there is one area where I think "live" collaborative working falls down.

Collaborative computing

If you're a software developer, you might be thinking about development tools that work very well – not because of their automatic, 'actionless' saving, but because of the additional actions. Because the world of software development has spent decades dealing with a) computers, and b) massive, complicated, interlinked collections of documents where one careless edit can cause cascading problems that can destroy a project.

Software developers use version control software to keep track of changes. Rather than have entire teams working on the same code at the same time, chunks of code are "checked out", edited and "checked in" again – new branches are made to test new code, which are then merged back together when they are working. I like to think that some day, anyone working together on computers will take this kind of thing for granted – that the days of dragging Powerpoint slides from one version of a presentation to another will come to an end.

But this is a blind spot that seems to date back to the earliest personal computers. When writing about the godfather of the modern computer Douglas Engelbart last year, I came across a story about a meeting between the inventor of the graphical user interface and the man who would make it a commercial success;

"Steve [Jobs] said, 'All the computing power you need will be on your desk top.' I told him, 'But that's like having an exotic office without a telephone or door.'" Jobs ignored Engelbart. And Engelbart was baffled. "We'd been using electronic mail since 1970 [over the government-backed ARPA network, predecessor to the Internet]. But both Apple and Microsoft Corp. ignored the network. You have to ask 'Why?' "

In other words, "Personal Computers" were designed by Apple and Microsoft with an assumption that they would be used by one person, alone, with removable media. Neither assumption is valid any more, but the design patterns based on those assumptions remain.

⁴