Stuck between the Dropbox that was and the iCloud that isn't yet

iCloud promised ubiquitous access -- all our stuff, every where and every when we wanted it. Not sync, Apple very carefully, almost awkwardly explained it, but an idea that was and is just as simple. You create something, it gets stored on the iCloud, and pushed down to all of your iOS and OS X devices. Not a server-side truth store, and critically, not a file system either. Unlike Google, it didn't live in the browser, and unlike iDisk, which came before it, there were and are no folders or hierarchies to get lost in, no Finder or Explorer to trudge through. iCloud, as Apple positioned it, was and is something new and something potentially much, much better.

The problem is, it doesn't work yet.

The architecture is unnecessarily dependent on apps. If I create a document in Text Editor 1, not only do I have to remember the document I created but, if I want to access it again, I also have to remember the app I created it in. If I later switch to a much better Text Editor 2, my document doesn't switch with me. I have to either copy and paste every document from Text Editor 1 into Text Editor 2, or keep a list of which documents are where. That's a non-trivial amount of cognitive overhead. If at some point I move on to Text Editor 3, or delete (or switch devices and don't re-install) Text Editor 1, it gets even worse. I have to track my documents over multiple access points, and perhaps even re-install old apps just to get back to the documents locked inside. It's a mess.

Decoupled, documents that present themselves to any app that supports editing their type, and apps that simply pull any document whose type they support, would be much simpler and better. A smart version of a document picker would remove the cognitive burden from users and let the system do all the heavy lifting. (I used to want Photos/image-picker-like access via a repository, but increasingly I think a flat store with search better fits the future.)

Of course, even if you do manage to keep track of all your documents across all your apps, iCloud's store and push features still haven't proven reliable enough for primetime. Key values seem to work okay, but documents still sound like a bag of hurt, with many developers struggling to implement them, or giving up on them entirely and switching to another solution. And that's on top of the larger problem facing Apple's services -- they're not historical one of the company's strengths, and haven't historically received the attention that Apple's software and hardware have enjoyed.

While Google, Facebook, and Amazon can snap up developers and designers and push out better looking and working apps, it's hard to imagine a plucky startup Apple could buy -- much less a NeXT-level acquisition Apple could make -- to jumpstart their services talent and technology the way they did their local operating system over a decade and a half ago.

Best case, Apple has secretly been working on something as important to the next generation of online services as WebObjects was to the last. Worst case, we're all in for a lot of pain and turbulence as they struggle to figure it out.

And that's in stark contrast to something like Dropbox, which enjoys about as much popularity on iOS as can be afforded a third-party service.

I've used Dropbox for years. My entire OS X documents folder lives in Dropbox. It's the first thing I install whenever I set up a new Mac. It's the closest thing I've found, Google Drive and Microsoft SkyDrive be damned, to truly automagical sync. It has versioning, it has un-delete, it has selective sync, and it's saved my ass more times than I can count. It's also being improved on the API side, making it even easier for developers to integrate. (Dropbox, it turns out, has also been a bag of hurt for developers for years.)

But here's the thing -- for all Dropbox's automagical-ness, it's a relic of the past. It's a file system. It's a hierarchy. It's a folder sync. It's a bunch of encrypted data stored on Amazon's S3 network.

As much as iCloud is the right thing still not realized, Dropbox is the wrong thing done brilliantly well. And at the end of the day, that still amounts to the wrong thing.

Those of us used to, and clinging to, traditional file systems love it, and will continue to love it as it becomes marginalized into obsolescence, as the growing mainstream -- those who aren't power users but are increasingly empowered users -- who won't get it and shouldn't be subjected to it, sweep past it and onto newer, better things.

iCloud could be that better thing, if Apple can nail it. It could be the iPad-style car to the old file system truck. So could something else, including a new version of Dropbox. But nothing and no one is there yet. So, as iPhones and iPads and other appliances bring computing to a broader user base than ever before, the services that bind them remain stuck between the best-ever version of the past, and a still sputtering and stammering future.

Rene Ritchie

Rene Ritchie is one of the most respected Apple analysts in the business, reaching a combined audience of over 40 million readers a month. His YouTube channel, Vector, has over 90 thousand subscribers and 14 million views and his podcasts, including Debug, have been downloaded over 20 million times. He also regularly co-hosts MacBreak Weekly for the TWiT network and co-hosted CES Live! and Talk Mobile. Based in Montreal, Rene is a former director of product marketing, web developer, and graphic designer. He's authored several books and appeared on numerous television and radio segments to discuss Apple and the technology industry. When not working, he likes to cook, grapple, and spend time with his friends and family.

  • While I agree that Apple's services suck and need an overhaul, I don't agree with the frequent imagining people do of a future without file hierarchies. It simply won't work. Have you ever tried to organize thousands of papers on your desk without using a file cabinet or folders? Good luck. The fact is that almost everybody, even "non-power users", do too much in too many different places to not use a hierarchy. I'd rather not mix all of my work files and personal stuff, thanks. So what's the newer, better think you speak of? Without any world-changing proposals, I don't think it exists in the near future. I'd much rather see apple do something along the lines of your global-access repository (where I can manage my own hierarchy, thanks). Of course, I actually want Dropbox to do it since Apple's cross-platform support is nil.
  • I agree about the need for hierarchies. I believe that dismissing them out-of-hand is a temporary fad that we'll recover from before too much longer.
  • I don't get that either. Yes, if you have a dozen documents, it's easier to just see them directly in an app's interface (not sure about more intuitive, but easier). If you have hundreds or thousands, the concept falls apart. People don't organize by app, they organize by project or some other kind of categorization. Sure, you can assign metadata... and IF you've properly assigned it... in theory... you could find stuff just as easily. However, at that point, you've probably put in more cognitive work than just learning how to use a file-system. Ever do a Spotlight search over thousands of documents when you can't come up with specific enough terms? It just doesn't work very well. But, here's the big thing. Apple, historically designed easy ways to do things, while allowing access to more power once the easy was mastered. With this move, all we have is easy, leaving anyone mastering the system frustrated and impaired. Regarding the cloud... the problem here, again, is in it being a bit too simplistic. It's so simplistic, you can mess up your data (if the iCloud doesn't do it for you) and it will efficiently mess it up everywhere in an instant... no history, no backup because the backup is now trashed.
  • I just got paid $6784 working on my laptop using these simple steps leaked on this web page. Make up to $85 per hour doing simple tasks that are so easy to do that you won't forgive yourself if you don't check it out! Weekly payments! Here is what I've been doing
  • File hierarchies are a special case of searching on the name of an object. If you let the object names have slashes in them, they look like directory paths. If you can list all the objects with names starting with /a/b/ you are listing a directory. If you can search on names and content, as with spotlight, there's no point in having any other hierarchy.
  • So true. If you can remember where you put something, you can just as easily remember a tag you've attached to it. And with content searching as well, you don't have a problem. The yellow notepad on ios, so little valued by most of us, proves this superbly! You can easily have thousands of notes in it and rapidly do a multi-tag search. You never lose anything in it.
  • I must be stupid! I want to use many different apps but have all my stuff in one place. I want it filed and sorted the way i like it, where I can see all of my stuff and access it when I need it! I am obviously holding it wrong!
  • No. It's much better to have it spread out everywhere and have to back up everything separately, not be able to delete the app, and ...I forgot the third thing. But it was an awesome point.
  • Before the car, people wanted faster horses. Before the iPhone, people wanted better Treos. It's natural to cling to file systems. I feel the same way. But we're computing dinosaurs. And they're not the future.
  • Ok So where is the preview app for iOS? iOS 7 has a lot of holes to address. Where do I save my PDFs and access them on my Mac and iPad using different apps? iBooks? Seems like there are some half finished white boards in Cupertino.
  • That's part problem, part old thinking. You shouldn't need a preview app, PDF or image files should just present to any app capable of showing them, including PDF Pen, PDF Pro, and iBooks. Because iCloud documents are app-bound, anything made on TextEdit for Mac can't be opened on iOS simply because there's no TextEdit for iOS. And that's silly.
  • Yes, but they are *FILED* with the App. As you said, the interface is the app. That is the problem... it's not future, it's silly. Sorry, Apple blew it on this one. Hopefully they come back to their senses one of these days.
  • Sure, it's old thinking, but it solves a very real problem. I couldn't live without Spotlight these days, but Spotlight would also be unusable without the ability to restrict it to a certain subset of my data. Currently, I use directories to do that. It's a very simple way to narrow down the dataset I'm looking at. If you take away the filesystem, you have to replace that functionality somehow. iPhoto does it with albums, iTunes with playlists etc. etc. They're all just different implementations of the same metaphor. So you do away with the filesystem, you have to come up with a different way to do exactly the same thing for each and every datatype/application. And they're all a pain if I want to access the data from outside the app. Anyone who wants to dump everything in the same folder is free to do that already. Your wanting to take that away from the rest of us who don't want to do that is just silly. Apple's whole iCloud/iOS no-filesystem metaphor is broken. Admittedly, it would work better if it were data-centric, not app-centric (such a ridiculous mistake), but your flat "filesystem" idea still leaves the user with the problem of having to categorise huge numbers of files to make them manageable and reasonably searchable. The filesystem servers exactly the same function as the domain name system, the library classification system, the species classification system. And if things belong in more than one place, there are symlinks or openmeta tags. A filesystem plus Spotlight-like search solves the problem extremely well, and you're looking to replace it with a multitude of incompatible systems for little benefit and a lot of overhead. How do I integrate a hundred apps' hundred different organisational metaphors with my backup/versioning solution that depends on files (so it works with *everything*)? How do I keep my data in sync across different platforms? How do I keep files away from an app that I know will break them (Cloud Outliner, I'm looking at you and the OPML files you destroyed)?
  • While it is true that iCloud has some issues, the scope of its implementation has contributed to creating them. People seem to forget that it does work very well for document-based systems (but yes, it's got a few cache problems). The real pain is core data.
    I'd also like to come back to the point you are making about acquisition. We all know that Steve tried to buy Dropbox without success. There is one company whose cloud service I've been using since its beta days, that is a godsend for developers: I have seen and given many talks where Parse is simply defined as "what iCloud should have been", and it really is.
  • The car didn't replace the horse, the engine replaced the horse as a way to move the "cart". Same method of movement, different driving method. Whether it's via an OS file system or an application linked system that pulls in any supported files in storage, the "file system" will always be the basis of any type of interface system because that's how humans organize their thoughts. And search is the same thing as a visual search, it's just faster because the system can do it faster. The problem with the system search is that if you don't quite know how to specify what you are looking for, you will likely end up doing a visual search anyways. When it comes to computer systems, even future ones, every way of categorizing and accessing files is going to come back to the "file cabinet".
  • For the end user, the interface is the app. Everything is just binary in the end, but it's how it's presented to the human being that matters.
  • What about when the end user needs to deal with more than the five Pages documents they have created? Sure, I get it is more simple. But, more simple is only good when it actually covers needs. An App is a tool to work with a document in some manner. It's that end document that is important, not the tool. When I'm looking for a document, I either remember some key component of it, or I remember how I've organized it. I don't think, now what App was used in the process of creating it.
  • Apple walks the fine line between power user and soccer mom. In many instances they choose to keep things simple enough anyone can work them, and then rely/expect a developer to create their own solution for a power user.
  • ...except in this case, on iOS (and OSX, if you want to get into the Mac App Store, a requiremnt to access iCloud), Apple expressly forbids developers from creating comprehensive solutions.
  • I have hundreds of Pages documents on my iPad all filed accordingly, within Pages. For anyone except an extreme "power user" one folder level is more than enough. It's actually the way most people organise their stuff, into folders that are all on more or less one level. I've been looking at people's file systems for many years and rarely do I find anyone who doesn't simply drag some files into a folder ant then leave that folder on the desktop. The thing most techies don't understand is that the people that make elaborate filing systems of many levels of folders and so forth are a rarity, not the norm.
  • Why?
    You highlighted a number of issues in your article, yet this deficient thing is the future? The iPhone is better than a treo. Unless you want to pet a horse, a car is an improvement. Storing files in the related apps IS NOT an improvement outside of simplicity for the absolute beginner. With that kind of thinking, we could also get rid of the virtual keyboard in iOS. Hold the device in one orientation to get the weather, another get's Apple's site in safari, etc. That would be WAY simple for beginning users. Four simple choices; simply tip the device. None of that messy typing stuff that is so yesterday. :)
  • If you really go in depth on how iCloud actually works, it's no different that Dropbox. iCloud itself uses a file system to sync data. There's a special folder just for a particular app in iOS, you put a file into that folder and it automatically uploads and syncs across all your devices. The entire basis of this article of Dropbox is doing the wrong thing is in itself inaccurate. With the Sync API for iOS, Dropbox has actually given us an iCloud with added benefits and total control. There are no headaches any more for managing Syncing which let me tell you is an extremely annoying task. I'd to work 2 months to write up and refine a syncing algorithm for using Dropbox SDK. Had Sync API be available before, that work would've been cut down to 10 days instead of 60. But now that it is, there's no doubt that Dropbox is going the right path with the right choice.
  • I'm not sure what you're proposing as an alternative. File systems are old and somehow hard to understand? You just spent half an article abusing the way iCloud does things, so that's not the answer. DropBox, Drive, and Skydrive work great and feel automagical because they embrace the natural human tendency to organize things into hierarchies. One of the frustrating aspects of IOS is its insistence on pretending there isn't an underlying file system.
  • I think the part being missed here is the "empowered user" that Rene refers to. These are the majority of the users of iOS these days and the main target that Apple is aiming for. These users don't care if they have a file hierarchy or not, they just need to find their PDF or document and view/edit it. The Age of the Geek is upon us, unfortunately everyone is becoming a geek and the true power users will be left behind in support of the masses.
  • "Empowered user" seems a nebulous and fluffy concept. Not sure how the idea of file hierarchies would mystify a person and cause them to feel "unempowered." The ability to navigate, search, and choose which app to open which file is quite empowering. Anything that restricts such activities is apt to feel limiting.
  • It is vague, but refers to the majority of the users who Apple is targeting. The user who has no idea that the magical device in their hand does so much more than send text messages, make phone calls and send/receive emails. Apple isn't looking to keep the power user happy, they are looking to draw in the new user with an easy to understand and use interface. Power users will purchase apps that scratch the itch. The majority of the users (Apples main audience) will just use what's there and be perfectly happy. My mother has no idea what a file structure is and could care less. She just wants to see the gas bill she received via email. Me, I file it away in a folder for that month, with the payment receipt, along with all of the other bills from that month, all from my iPhone or iPad.
  • The author has already stated this in a reply, but bears repeating... The file system is for the computer. What is needed is an interface for the user.... Not the developer who understands the limitations of the computer and the file system.
  • The author has already stated this in a reply, but bears repeating... The file system is for the computer. What is needed is an interface for the user.... Not the developer who understands the limitations of the computer and the file system. Or to put it another way, people want their data... not learn more software.
  • Again the assertion that file systems "are not the future" Tell me, when you organize your own system, do you do it by app or by concept? If you are like, well, anybody, you do it by related concept - imore files in one location, home finances in another location, and so on. It is the least amount of congnitve overhead, because you already make this distinctions in your mind, regardless of the presence of a computer. And, as to the patronizing attitude of "yes, but I'm special, but *others* need it more simple" -- it ain't more simple. Lets take a look at home finance. In a couple of months, everybody is the USA gets to deal with taxes. That may include PDFs of pay stubs, Excel/Numbers spreadsheets of budgeting, and Notes/Word documents of reminders. If you are a business, you can likely add proprietary expense and payroll applications, as well as legal documents. Are you seriously attempting to claim it is "simpler" to hunt for these documents in various different applications, each presented in a separate flat list, versus having an are designated "2013 Finance" on your storage device? Adding metadata on top of those documents - assuming every app will allow the same metadata to be added in the same way, simply recreates every "problem" of a file system you decry, and disintermediates the user from direct access to his/ her data. Bad all around. Nobody is "clinging" to file systems to a avoid change. But you are going to have to come up with a compelling reason why grouping documents by related real-world use is strictly inferior to rigid enforcement off app silos, because "the future" you envision explicitly disallows users from thinking that way. And all of us - not just "dinosaurs" - think that way.
  • Your holding it wrong ! Lol I agree with u apple is telling us our brains don't work right!
  • Rene, I have totally disagree with you. If we continue to dumb down everything we become 1984 all over again. I have interacted with many businesses and schools over the past 27 years with the Mac. Example. A user create's a project with Quark Express. You have the Quark Document, Graphics, Fonts and Text related to the Document. You put all those files into the same folder. Possible under "DropBox", under Mac OS X and Mac OS 9 and earlier. With iCloud forget it. I gave iCloud a try, but like MobleMe Apple just doesn't get it yet! Steve Jobs new "DropBox" was the answer, he just wouldn't pay the price to get it. iCloud for handling personal Calendars and Contacts may work some of the time, but for me Pages in the Cloud is a total disaster. DropBox worked so well I ponied up and paid for it even though I didn't need all the extra space. DropBox Works I can't imagine my friends Law Office under iCloud. It would be more like Documents in a Tornado or Hurricane! I do enjoy most of your writing and your work on MacBreak Weekly, but iCloud. I think your judgement was Clouded! Regards, Joseph J Gudac Jr.
  • Comes on files app in ios 7 and airdrop.
  • Can't agree much, I think Dropbox is perfection; specially if implemented as in the iPad app Notes Plus, there's not much to improve (if anything).
  • Drop box is really great. Hierarchical folder structures are great for easy organising. iCloud is a pain in the butt and I steer clear of it.
  • I love iCloud. It's great at keeping my backups, for restoring, for music. I like that I can download my iTunes music to any iOS device. Same with contacts, Photo Stream. All this is done with little input from me, the user. So it makes it extremely simple to use.
  • Renee, What you want is complete integration, and organized unrestricted cloud. You basically want Outlook on steroids. I'm not saying I could walk into Apple and start coding iOS from the ground up, but what I do know about iOS is it is being pushed and stretched beyond the limits it was created for. To have an OS as the groundwork and apps running over the OS is now in 2013 an archaic design. I think what you are wanting (and this is my best understanding of iOS code & protocol) is beyond the capabilities of iOS, like Widgets or Twitter/Facebook integration without actually building the app into iOS. I predict Evernote integration into Notes but not until a new platflorm is released by Apple to replace iOS, we are all stuck in limbo.
  • "it's a relic of the past. It's a file system. It's a hierarchy" In the immortal words of the Monty Python Cheese Shop sketch, "explain the logic underlying that conclusion, please." Everything on your wish list is accomplished by a file system with document type handlers. Your assertion is therefore...curious. You disclaim it by saying that *you* are also a dinosaur, but, aside from Apple's dictates, where is the evidence of the masses of mammalians who demand closed per-app document silos?
  • If anyone wants to sign up to Dropbox, use this link. It's a referral system, and we both get 500MB extra space for free, no strings attached.
  • iCloud is not a file browser. The whole point is that the user does not have to interact with it. It just works. Sandboxing is another reason for its design. With OS X moving towards sandboxing, apps are only permitted to open files they create. iOS, which employs sandboxing, does not have the ability to work in a traditional Finder mode anyway. Where Apple does need to improve their iCloud service is to merge iCloud files with the traditional Open / Save feature, so that only the app that creates a file can see it, but you don't have to specifically go to iCloud to see the files in that app Cheers !
  • This argument makes no sense, but maybe that's because you're a journalist and I'm a web developer. I could not operate my development business without a file hierarchy , it would be completely impossible. Hell... A website can't hardly even exist without one.
  • I think this is the core of the "problem" right here. Developers and techies are exactly that small subset of humans that "do" file hierarchies and have no problem with sorting files into folders, keeping things organised, and finding them again when needed. The "problem" comes because almost all other humans have huge problems with this. The majority of people are the "other" way and don't want to have to manage and store their files. The average messy desktop is proof of this fact, as is the ten years or so of Apple's efforts to get around the file system so it's users can actually find stuff and work with files efficiently. The developers making the apps just don't "get" this. It's against their own nature to do so.
  • Those of you wanting a file app, just jailbreak your device and get iFile.
  • Install and run a copy of Copernic Desktop Search on a Windows machine -- or better yet a large network -- to get an idea of how to retrieve and use information in a wide variety of formats despite the fact (yes, despite) that the information is stored in heirarchical folder and file-type fashion. We installed it years ago, with the complicity but not approval of the IT folks, in a large Canadian government department and productivity immediately soared and error rates dropped dramatically as people suddenly started finding and using everything they wanted in a system that had baffled them before. We had thousands of PowerPoints, hundreds of thousands of Word documents, audio and audio-visual files, PDFs, and a host more of things, and suddenly a search could find anything within and across these categories, and on a click all these files opened from within Copernicsearch results. The system always presented the most recent versions first, etc. If Apple is headed in the direction Rene indicates, and they obviously are with iOS and OS now under one man, they could do worse than study Copernic as a model not just for finding, but for organizing and making useful what is found, whether in desktop or cloud information-base.
  • There's a lot of counter-arguments here about which system is best but to me, Apple's iCloud system is superb for pure syncing of data ...and when it comes to universal access of files, you use Dropbox. So all Apple needs to do is create a drop box system on iCloud of its own and all is solved for all levels of user.
  • Check out DropCloud at
    There's no app for iOS yet, but it's a start.
  • iCloud is plain, simple, and just works!
    No need to interact. Just take a picture on the iPhone, and it will appear almost instantly on your iPad, Apple TV, or iPod touch! Make a note and same thing! Put your iPhone on the charger and it backs up your phone without even knowing! Visit a website on the iPhone and you can continue on the iPad! Install an app or buy a song and it will do it automatically on the iPad without even doing anything! Again plain, simple, and just works. I don't think an average consumer wants to deal with file systems or whatever you call those stuff. Just like android where majority of users don't even know that they can change stuff. They just want a cheap phone with touchscreen then that's it. Only the techy wants to customize. The only thing that iCloud should improve on is staying on all the time. There's a blackout every now and then.
  • Those of you who believe file hierarchies will be part of the UI future clearly haven't seen how ordinary users actually work with computers. From a techy point of view, it's painful to watch files being spattered randomly wherever open/save dialogues happen to start, or spread over the desktop, with little (or no) understanding of the difference between files and directories, or even between different types of files. And I'm not talking here about old folk not brought up with computers, I'm basing this on dealing with contemporary university students, suckled on the 'net. File hierarchies do not offer affordances that are a good fit with the average human mind. As more well-matched UI metaphors are developed, they will *naturally* win audiences, and file hierarchies will settle into the background, much like commandline shell: ever-present and available for the subset of users with specific skills and interests, but unnoticed by most. If you're an average power user, you might as well ignore this. File hierarchies won't disappear, and are probably the best way of doing what you do. But any developer of software for ordinary end-users would do well to keep sniffing the air for what's coming. Its exact shape isn't yet distinct, but that it won't look like C:\ or / is obvious.
  • I have such a user here. Problem is, the number of his files has grown over the years. Now he spends a lot of time desperately searching files, he knows that he has, but cannot find them. Often he comes and asks me for help finding his stuff.
  • Exactly, happens all the time. It gets even worse when such messes get passed on to others (eg. through job changes). No system can protect us entirely from messes. But current file system hierarchies encourage them. They'll have to go, and go they will. I'd like to see Apple be the company to get it right.
  • "…in every user interface study we’ve ever done [], [we found] it’s pretty easy to learn how to use these things till you hit the file system and then the learning curve goes vertical. So you ask yourself, why is the file system the face of the OS? Wouldn’t it be better if there was a better way to find stuff? Now, email, there’s always been a better way to find stuff. You don’t keep your e-mail on your file system, right? The app manages it. And that was the breakthrough, as an example, in iTunes. You don’t keep your music in the file system, that would be crazy. You keep it in this app that knows about music and knows how to find things in lots of different ways. Same with photos: we’ve got an app that knows all about photos. And these apps manage their own file storage. And eventually, the file system management is just gonna be an app for pros and consumer’s aren’t gonna need to use it." -Steve Jobs
  • Why would it be crazy to "keep your music in the file system"?
  • Conspicuously absent from this discussion is Apple's recent acquisition of Maya Sytems intellectual property. I hate the file system hierarchy, having to remember which files are where, wondering if I can delete a file in one place and lose it completely (I'm obviously not a power user). I think a system based on relationships (to people, projects, subjects, etc.) is more intuitive than the hierarchical file system.
  • This looks good. Looks as though you can tag anything, including emails, for a complete solution. I wonder how well it integrates with apps. Anyway, this is also contrary to Jobs' app centric philosophy quoted above. Maybe Apple is departing from that.
  • I'm referring to the Maya Systems ("IamOrganized) app.
  • Edit
  • I think a combination of tagging, auto-classification (like in DevonTHINK) and a probability-based scoring system with heavier weights given to time and frequency should do well. i.e. documents that were created/modified/read recently, and those that are frequently accessed. Kind of like email. :) The file system still remains, but ideally no one should be looking at it in its raw form. Managing a hierarchical file structure is a major pain.
  • Apple has been app centric for years. Where M$ made a "My Pictures" which was basically a folder with some extra photo related widgets, Apple made an app (iPhoto) and obfuscated the underlying file system. iOS takes this concept to its logical conclusion by concealing the file system entirely. Jives well with the apps in a sandbox concept of operating as well. Things will get interesting when OS X and iOS inevitably merge into a single OS down the road.
  • One wonders where we might be today if Steve Jobs had been successful in acquiring Dropbox in its relatively early days when he tried. He told them he saw Dropbox as just a 'feature' at the time - I wonder if he would feel differently now. Like I often wonder what would have happened if Google had not gone with Android and instead worked with Apple... I suspect we might have something altogether more useful and amazing today. Given that Tim himself has said that iCloud is a vital part of Apple's future, it does seem strange that it hasn't had more visible attention and resources put into it - perhaps there's a secret bunker full of guys on it now. I have a secret 'Best Case' scenario I often think about too... that Apple has secretly been working on something really significant, even paradigm shifting, for the next iteration of our favourite OS and devices. We're close to OS X 10.9 now and they're about to run out of cats... they must have some idea where they're going after that...
  • I agree with Caballera, I think in the majority of use cases iCloud works perfectly well. It seems like you've taken a corner case (the ability to transfer documents between multiple applications), and used it as the foundation for a more general criticism of iCloud. In reality, how often do you move documents between apps? Personally, not often, and in the cases I do need to do that, I can just move the file out of the cloud.
  • Great article, and I'm glad that at least conceptually, you came down on the side of iCloud instead of DropBox. iCloud is indeed the way forward IMO. If I have one quibble with the article it's that like many others it describes how difficult (supposedly almost impossible) it is to implement iCloud with documents, when I think the facts indicate that this "bag of hurt" is actually a gross exaggeration of the true state of affairs. the fact is, there are quite a few apps, (I have several), that implement iCloud quite flawlessly. It's certainly true that the model itself can be frustrating to the end user if they switch apps, just as described in the article. This is indeed a conceptual flaw with the system. The implementation of iCloud (as it currently exists), within an iOS app though, is far from impossible. The very fact that many apps do implement it and implement it well is the clear evidence that it is possible. It's just hard to do, and this is the source of my main complaint about iCloud. What's absolutely annoying and offensive to me is that many apps today simply refuse to implement iCloud in the app even though many of their users want them to rather desperately. There is a sort of micro-community of hipster software designers that are telling us that they actually know better than us and that they will implement DropBox, but they won't implement iCloud simply because it's "too hard" for them and more importantly because they simply "don't like it." The implication being that if you want to be "cool" you also have to like DropBox and "not like iCloud." This is pathetic to say the least. I get that it's "hard to code," but my response is simply "step up and do what your users want anyway." Why should I as a customer buy software from some developer that is saying something is "too hard" for them to implement? Others can do it, why can't they? The reality is that there is this stupid little game of "who's cool" going on and if you don't happen to believe in DropBox you're SOL in terms of many major apps, and of course resigned to the ranks of the "non-cool" merely for wanting iCloud. I think it's heavily ironic that these "uber-cool" programmers are actually hanging onto the past, while simultaneously dissing people who want iCloud as "useless seniors" (yes, I've heard those words come out of some of these people's mouths), when in fact these customers want the future, and the developers are the ones stuck in the past.
  • @Gazoobee: you're mistaken here. The fact that a framework technology works for some apps and not for others can be a consequence of problems either with the framework or the apps. The framework can have variants and edge cases that affect situations met some types of apps and not others. This is the case with iCloud: it's near-universally accepted amongst developers to be buggy, and (just as important) often impossible to debug (essentially, Apple hasn't yet given us the mature tools or documentation). Roughly-speaking, iCloud offers developers 3 services. One of these (key-value storage) works reasonably well. One (document storage) works in some situations but not in others. And I haven't yet come across anyone who's comfortable that the 3rd (Core Data storage) is ready to use in production apps. iCloud's problems are Apple's. They've never had a particularly good reputation for online services, and iCloud is their opportunity to get it right. Conceptually, they're half-way there (and far ahead of anyone else). But the implementation is still very much half-baked, albeit serviceable for some situations. Some of the very best developers in the business have put in hundreds of hours, often over a year, trying and failing to get iCloud to work reliably with their customers' data. When Apple gets it right, I assure you, app developers will use it for all it's worth.
  • I disagree. You say "I'm mistaken," but then only point to some general principles of app design that *may* indicate that I *might* be mistaken in the case of one particular app or another. So it seems to me that you are arguing from a position of bias yourself. I'm not disagreeing with the notion that iCloud has problems or that it might be difficult to code for, (which is all you seem to have proven with your statements). I agree with that. What I'm saying is that several apps that I use on a regular basis have no problems with iCloud whatsoever and have near perfect implementations of iCloud support. Sure, *some* apps doing *some* specific things might run into problems. And yes, (technically), I'm wrong to argue from that basis that the mere existence of these working apps necessarily means that "all apps can implement iCloud features without problems." But I didn't really say exactly that and I find that your rebuttal is essentially just hypothetical. The apps I'm talking about, and the iCloud features I'm referring to are basically simple document storage/retrieval and updating (present in every note-taking app for example). These are not "edge cases." This is simple storage and syncing of documents which is expertly done in some apps yet others not implemented at all in others, who merely cite "problems." This is why my feeling is (and I think the evidence supports the assumption), that the reason iCloud isn't being implemented in many of these apps has a lot more to do with "what's cool" and "what developers like/think" than it really has to do with the problems of coding for iCloud itself. I also work in the industry and talk to lots of developers myself so I've had many of them tell me quite directly that "iCloud sucks" or "we only like to use DropBox" and "Apple should implement a DropBox." Of course I'm not prepared to argue it before the Supreme Court, but I think it's pretty clear that this developers bias exists. The comments section on this very article are an excellent illustration of this bias at work.
  • Truly, you're just mistaken about this. I want iCloud to work, and very much hope that Apple can get their act together, but for now it's a liability that any developer is brave to take on. iCloud is barely documented, not technically ready for many apps to use, and introduces huge amounts of technical risk. It not infrequently *loses customers' data*, and is nearly impossible to debug when this happens (and Apple offers little support). I haven't come across a single developer who disagrees with this. I mean, really, not one. I'm wondering what source of info you have that Apple developers worldwide haven't? The fact that some apps appear to work OK means nothing unless you have their data-loss stats. Even developers of apps that use iCloud frequently admit that it's not reliable, and that there's little they can do to help when things go wrong (eg. and iOS devs in particular took iCloud up with great enthusiasm when it was introduced (in preference often to Dropbox because the latter didn't then offer a sync API), but are increasingly pausing to wait for more functional iterations from Apple. Two of the very best development shops in the business, OmniGroup and Bare Bones, cannot get it working reliably. Both wanted to use iCloud from early on, tried, and failed. OmniGroup have had to develop their own sync server as an alternative, while Bare Bones' updates sound increasingly despairing ( It's not a question of it being 'hard' to do, or of its having some problems. It's half-baked, and it involves more very risky gambles with dev time than many businesses can afford to take on. The ball is very much in Apple's court. If they can make it work and document it properly, devs will use it in droves. Until then, they'll continue to invest their time in things they know they can make work.
  • iCloud's lack of a folder structure means large numbers of files become unmanageable: the more files you have, the (exponentially) more difficult it is to find any particular file. Someone or something doing the management of file organization is necessary. But the way iOS does it now — having content (pretty much) accessible only within the app that created it — is unworkable because users want to access that content with different apps all the time. Apple's ideal is that content will be organized on the user's behalf behind the scenes. Doing this in a more open way is what's next — enabling any app that works with text files being able to open any text file, any photo app able to open pics, etc. But then we're back to the original problem: how can large numbers of files be managed without a folder hierarchy? Exactly how iPhoto and iTunes do it: Albums, playlists & smart playlists. This is essentially something we're all already familiar with: tagging. And this, I'm guessing, is what will come with iOS 7. Whether there's some system-based "iDocs" app or not — and almost certainly not — users will finally have the means to find, access and (if they want to but it won't be necessary to) organize content within the flat file structure using smart search. You remember you last accessed file XYZ last month? Search for content modified in last 30-60 days, then save a smart album with this criteria. And so on. Just as it it now. [By the way, this doesn't mean sandboxing will go away. Again, it'll be just as it is now: apps will be able to perform actions only in prescribed ways and combinations. The only difference will be that the API will grow even smarter. More actions will be available for devs to enable in their apps, with Apple eventually even allowing some Apple-prescribed cross-app actions and system functions (i.e. Applescript/Automator -like functionality). As long as your app follows the prescribed methods, it'll be allowed to do its thing.] The bottom-line problem with all of this — and why many of us are holding tightly to hierarchical folders — is that search alone can *never* be smart enough, I believe, to enable a flat file structure to be easy to use. What if the only thing you remember about the file is that somewhere in the first page of data it included "West" and "1986"? What about multiple versions of a file with only a few bytes difference between them? Instances like these will almost never arise with personal photos & music, so the flat structure of iPhoto & iTunes didn't need to be designed to handle them. Removing the hierarchical structure for all files is unworkable, however, because doing so destroys the inherent relationship between different files, e.g. "where" one file is relative to another. The desktop/filecabinet/folder metaphor ensures that even a user who never organizes his files at least puts file XYZ **somewhere**. And that little bit of location info — be it "on the desktop" or "next to file ABC" or "that email Tom sent" or whatever — makes all the difference.
  • I think that search *can* be smart enough, one day; you never know the limits of technology. What I think is that that location info becomes instead an intrinsic property of the file, because most locations are representative of the files they contain. E.g. the Downloads folder contains downloads; my Projects folder contains projects. Next to file ABC means the file is related to ABC, and I think that if data becomes semantic enough (which I'm sure will happen in the next half century) such that it can contain this kind of information, search based organization can work. (btw, I elaborate in a blog post! It would be really cool if you could do me a favor and read it.
  • I need folders and a hirarchy, otherwise I'd need seperate accounts/users for the different jobs I work. I can't mix several jobs in one flat-storage. This is why I don't use iCloud, because I NEED a hirarchy and folders. I'd Love to see Application-Folders (stackable!), where each Application get's it's folder and you can access it through Finder AND the Application in question. That way it would be easy to transfer a document from on e App to the other and when you opened an App, the documents/folders in it's folder would be visible in the iCloud-Dialog.
  • Haven't read all the replies yet, but let me add my $0.02 on the concept of hierarchies. File manipulation requires search. Search is an "ordering" of information. "Order" defines hierarchies, end of the story. The question is not whether hierarchies are needed (they do - it is a logical necessity). The question is whether the definition of said hierarchies is done by the user or by the machine. As others have noted, a file structure is simply a user-made ordering (hierarchy) of information. Google search or spotlight is a machine-made ordering of information. What the author I guess tries to convey is the idea that search will eventually be a machine task only. I think I agree with this. The "cloud" part of iCloud is not what matters here. iCloud as search is what matters. Search in the broader sense of ordering information. In this regard, as the original author states, iCloud is far from being currently successful in replacing the cognitive overhead required.
  • The more I think about it, the more I agree with you. Most of the other replies do defend the presence of a file hierarchy done by the user and initially I was in that boat too. But if you think about it, Search does make a lot of sense. Instead of doing "get this pdf named Chopin's First Ballade in the sheet music folder in this other folder" a query like "get my copy of Chopin's First Ballade" would make a lot more sense. (hence I got bored and wrote a blog post about it!
  • I'm a bit tired of the "traditional hierarchy vs search" debate. First off all, a hierarchy is not "traditional". It's a concept used in the real world and everyone can understand it. A paperclip is inside a box, inside my drawer, inside my desk, inside my office. It's a great and easy to understand metaphor. Why get rid of that? Secondly, it's not "search or hierarchy". The search technology nowadays is great and it works really well with folders. Both have their strengths and weaknesses. If you want to maintain thousands of documents with projects files, I could theoretically tag them using OSX Mavericks. And yes, tags can be more advanced by adding multiple tags to files and then saving them as a "search folder". It's great for advanced stuff like "tag all my documents about animals wherever they are".
    But you still want to group them in a container. It reduces clutter because storing them in a flat filesystem will only make them get lost. It's the same as in iTunes: most of my files are "lost". I don't know what I have anymore because it's so much. When I create a smart playlist, a lot of songs will match and voila, they are back. But there are dozens of files which may not get matched and stay hidden. In that case I would have preferred to browse in a hierarchy and find my manually organized files back. It's just too much work to organize files manually beyond placing them in folders. You are screwed when you forget to apply the correct metadata. Your stuff gets lost in a huge flat root folder so to speak. Lastly when it comes to files, you still want your stuff to be interchangeable with Linux, Windows and whathaveyou. A database is not, a file system does. Searching, metadata are FRIENDS of a hierarchal system. They should complement eachother.