The PC is dead - long live the operating system?
The first part of the headline is pretty much taken straight from the media. It’s almost as if mass-hysteria has taken over the world with doom sayers already predicting a clear line of sight to the end of the PC. Supposedly, because of tablets the consumer will shift activity towards pure media consumption and the PC sits unused somewhere in a corner. It is true that tablets are excellent for consumption of media - but for the rest of your digital life, not so much.
There is simply not a single area in my digital life where the tablet would have an advantage over the PC when it comes to creating and editing content. This is as diverse as writing a document, editing a movie, laying down a track, sorting digital images and playing games. But more importantly, it’s speed! But the sheer adoption of tablets does tell us something, namely that the consumption really is moving to a different experience and that more consumers are finding tablets to be a much easier to use.
This should give us pause because we have to understand that iOS and Android are not fundamentally different operating systems than let’s say Mac OS X or Windows (in fact iOS runs on the same kernel and has inherited a lot of the original OS X api’s) - but Apple, for example, decided early on that it wanted to get rid of the filesystem with iOS. Of course, the filesystem still exist but files are really not represented at all - nor is there the idea of a filesystem with its hierarchies, permissions, role based access etc.
Hence a big aspect of the success of the iPad is about simplifying the computing experience for the rest of us. So we have to look at this development in a different light and ask the question, if we can live without files, filesystems, cryptic commands, processes and cpu utilities on a tablet, could we also live without the same on a desktop?
In other words, no longer is the operating system at the center of the user experience but rather apps. I’ve made a suggestion how to utilize Siri a while back - the basic premise was that Siri could be utilized as a means to self-organize documents and provide an interface to search and quick access of these documents. As an example, instead of browsing through the directory hierarchy and looking for something specific, I would ask Siri: “open the picture that I edited the other day”. This is akin to laying another layer on top of the filesystem but it doesn’t quite go as deep as iOS.
The real issue comes down to workflows - we have files and we organize files primarily because we do something with files. On iOS the workflow starts with an app. On desktop operating systems, it usually starts with a file. In other words, on iOS we access content through the funnel that is the application itself and not the underlying operating system. Which brings up a new question: what happens when the same file is accessible by multiple apps? For example, a jpg document.
iOS has solved this problem for the most common file types, such as pictures. Whereby all pictures are stored in a very specific and unique location.
This is quite an elegant concept because it allows every application that knows how to edit jpg’s to access them without the user requiring to traverse the filesystem hierarchy. In addition, we’re not talking about files anymore, we’re really talking about documents and thus the filesystem layer is shielded from the user. Further, it allows the operating system to manage pictures differently than other documents - in fact, the key is that there is no longer a need to treat it just as a file that can be moved or renamed because you can’t.
I think this is applicable today in modern operating systems as well. The key ingredient is that there can’t be any user defined files. In other words, as a user, there is no facility to change the file type of a document or rename it’s ending. In fact, there’s no ‘Finder’ (on OS X) or file browser (on Windows).
That concept requires applications to ‘register’ a filetype globally (that is with the operating system) or receive confirmation that there is already a file type registered. Either way, it’s the operating systems job to create a space within the user’s home directory where these documents are stored.
The operating system can add value to the management of these documents by knowing exactly what the document is for. Similar to iOS’s photo app - there could be basic viewer apps provided for every file type. Let’s say I wanted to browse all word documents, a viewer could be launched that displays these documents in such a way that would let me quickly glance at them. Further, there could be advanced search right built-into the viewer and of course, a delete function. Likewise, the powerpoint presentation would have the ability to create new documents from templates on slideshare, there could be a way to tag the documents or group them together (the filesystem could represent this as a directory).
The new operating system (let’s call it iOS Pro) would have a built-in app that would allow access to all these document types (essentially, the finder replacement) - but each document type (upon launch of the document type application) would restrict the user to the context of that document type and most importantly this would be a power user function because there wouldn’t be a need for the average user to access this app.
Now, here’s the beauty: all these components are available today on every Mac that runs OS X.
Some tweeks would needed to be done, such as filesystem panels could be changed to only allow a single filetype (many already do) and there would not be a possibility to change the location where the document would be stored. In other words, the file system dialog box would be as simple as entering a name for the document. Further, since this is a ‘new’ user experience, it could actually be applied to only certain users. It is not incompatible with how the operating system currently works but simply a layer that mimics much of what iOS does so well. Everyone wins.
Oh.. and yes, in case this all sounds familiar, Microsoft tried something in that direction with Windows RT - but where it has gone wrong is confusing the user endlessly. I think the user shouldn’t have a choice during a session. In other words, he or she is either living on the desktop OR living in windows RT world.