iOS lacks a document filing system

Since the beginning of 2010 when the iPad was released, there has been no end of debates over whether it is suitable for creating content, or whether it is primarily a “content consumption” (ugh) device (as if the choices were thus limited…). I am resolutely of the opinion that the iPad is an easel that very much supports serious creative endeavors given the right environment.

I unfortunately had (as you may have noticed) to qualify that last statement. Besides a few colleagues at work, two examples of iPad-using people that I base this statement on are the Macalope and Harry McCracken. And these examples have something in common: in all three cases, once the work is done, the documents are sent, handled, stored, etc. by either a corporate server, or a publishing CMS, or some other similar infrastructure. Here the iPad only needs to make a good job of storing the document for the time necessary to complete it; once done and sent, the document can even be removed from the device.

Let us contrast that with another situation. My father is a high school teacher; for the last 25+ years he has been working using computers, preparing teaching notes, transparent slides to project, diagrams, tests and their answers, student average note calculation documents, etc. on his Macs (and before that on an Apple ][e). He shares some of these with his colleagues (and back) and sometimes prints on school printers, so he is not working in complete isolation, but he cannot rely on a supporting infrastructure and has to ensure and organize storage of these teaching material documents himself. He will often need to update these when it’s time to teach the same subject one year later, because the test needs to be changed so that it’s not the exact same as last year, because the curriculum is changing this year, because the actual class experience of using them the previous year led him to think of improvements to make the explanation clearer, because this year he’s teaching a class with a different option so they have less hours of his course (but the same curriculum…), etc. Can you imagine him using solely an iPad, or even solely an imaginary iOS 5 notebook, to do so? I can’t. Let us enumerate the reasons:

  • Sure, one can manage documents in, say, Pages. But can one manage hundreds of them? Even with search this is at best a chore, and it’s easy to feel lost as there is no spatial organization; and search could return irrelevant results and/or not find the intended document because of e.g. synonyms.
  • If one remembers a document, but not the app which was used to create it, it’s hard to find it again, as the system-wide search in iOS cannot search in third-party apps (at least it couldn’t when this feature was released in iPhone OS 3.0, and I am not aware of this having changed), so one has to search each and every app where this document could have been made.
  • In some cases, for a project for instance, it is necessary to group documents created by different apps: sometimes there is no single app that can manage all the different media for a single project. On iOS these documents can only exist segregated into their own apps with no way to logically group them.
  • If there is a screwup, as far as I am aware it is not possible to restore a single document from backup, in fact it does not seem possible to restore a single app from backup, only full device restores, which may not be practical as it likely means losing work done elsewhere.

iOS needs a document filing system, badly.

The worst thing is, with the exception of file transfer in iTunes (which pretty much only shifts the issue to the computer, with some more overhead), the situation is the exact same as it was in iPhone OS 2.0 when third-party apps first became possible. iCloud solves exactly none of these problems: it is great to simplify working between your different devices, but it brings nothing to the single-device case. This has nothing to do with the hardware limitations of any iOS device, this is entirely the doing of the iOS software; in fact, while this is acceptable for the iPhone, I feel this gap already limits the potential of the iPad unnecessarily; and regardless of how you think it will happen (my take, which I will elaborate in a later post: Mac OS X is the new Classic), it is clear Apple has Big Plans for iOS, but it is hard to take iOS seriously for any device used for work if Apple hasn’t even shipped a first version of a document filing system, which is quite a design task and will require multiple iterations to get right for most people.

Now you may be wondering: does it really matter for working on iOS to depend a corporate, publishing, design studio, etc. infrastructure? Most people working on computers already work in the context of such an infrastructure. I think that yes, it does matter. Even if we admit that people working outside such an infrastructure are the exception rather than the rule, there are many of them, enough to prop up a competing platform (potentially the Mac) that would cater to their needs. Plus, sometimes such an infrastructure (e.g. in small businesses) may be unreliable, so it is a good idea to have a fallback. Moreover, it’s not really a good idea for Apple to make iOS dependent on such an infrastructure, as then Apple will not be able to control aspects of the experience it likely cares about, and will not be able to define, for instance, the modern notion of how to encapsulate user creations (I can imagine Apple getting past the concept of documents themselves and introducing something new), or how document typing information is represented. Whereas if iOS devices had a document filing system worthy of its name, but could also still be used in such an infrastructure as they can today, then Apple could define the rules and external infrastructure would follow the lead. Currently, iOS devices are more akin to terminals when it comes to working on them; not quite VT-100 or Chromebooks, but you get the idea.

When I see the absence of a user-visible traditional file system in iOS being lauded as some sort of brilliant new move, I’m scratching my head. It is a bold move, for sure, and not having something does represent accomplished work in the sense that it is a design decision, but honestly not having this feature is the easy part, creating a worthwhile replacement is the hard part, one that Apple has not shown even an interest in tackling. Moreover, the absence of a user-visible filesystem is nothing new. Indeed, back in the 80’s when computer GUIs were developed, two philosophies emerged for dealing with documents: a document-centric approach, where documents are at the center and applications are but tools which can each be used for a specific task on these documents, and an application-centric approach, where applications are the focus and documents only make sense within their context. The Apple Lisa, for instance, was document-centric: users would tear down from a stationery to create a document, which could then be operated on by tools. By contrast, the Macintosh (and everything it then inspired) was mostly application-centric. In this context, iOS merely is purely application-centric. Precedents of such systems exist, and include game consoles with memory cards for instance.

And was it really necessary to forego the filesystem in its entirety in the first place? Admittedly, it has become more and more complicated over the years, with documents being diluted by an ever increasing number of non-document files visible to the user, especially after the Internet and Web came to be. And, okay, even the Macintosh Finder at the origin did represent applications and system files along with user documents, and thus was not really a document filing system. However, was it really necessary to throw out the baby with the bathwater? It would have been feasible for iOS to feature a clean filesystem view with most everything invisible and various enhancements (like virtual folders and virtual filenames) so that it would only feature documents (in fact, I think the Mac OS X Finder in 2001 should have shown only the inside of the home folder, with applications launched from a Launchpad-like mechanism, but I guess a few things like the need to support Classic prevented that anyway). But maybe filesystems as users know them had truly become fatally tainted, and maybe it was indeed necessary to take a clean break from the past; in the end it doesn’t really matter either way, however it is not a good thing to forego something and put no successor for so long.

In the end, I am afraid Apple is not taking this aspect of the computing experience seriously, and is neglecting it. They ought to take it seriously, because it will matter, I think it will matter a lot in fact.

I explored a related aspect of document management in a followup — February 21, 2012

~ Reactions ~

Jesper (who, unbeknownst to me, had already touched some of these points, such as the specific notion of a document filing system) expands on the matter, also theorizing why the iOS group makes iOS be that way.

Unfortunately my knowledge of Magyar is exactly zero (and Google Translate is a bit hit and miss), but I’m sure Benke Zsolt is saying very interesting things.

I am honored that Lukas Mathis would link to me, but if I am mentioning it as a reaction it is because of the slightly overstated, but pretty good comparison he added.