May 17th, 2008

Dear Web Applications: Where Are My Files?

by

What’s wrong with the “friends connection” programs announced by Facebook, MySpace, and Google? Many people have been trying to explain the principle of data portability as if it were a new concept, but it’s actually not. It’s been on our PCs for years.

Think about the applications you use on your computer — the ones that run LOCALLY on your computer. They all produce files. You’ve got your word processor files, your spreadsheet files, your presentation files, your accounting software files. You create some data with the application then save it to your drive. You can take you take those files and put them on any other computer and open them with any application that supports the file type.

Think .doc, .xls, .jpg, .mp3

Web applications are different, because they don’t run on your computer — they run on the servers of the application provider. You access the application over the web, using your web browser.

So the application isn’t on your computer. And neither is the data you create with the application. That, too, is stored on the servers of the application provider.

Social networking sites like Facebook and MySpace are applications that run on the web. You can use them to create data, just like applications on your computer. You can enter information about yourself in your profile, and you can create connections to your friends profiles.

All the information is stored in your profile — on the Facebook’s or MySpace’s servers.

You can’t actually get at the “file” with your profile data. It’s in a big database, not separated out like the files on your computer.

Here’s “data portability” in a nutshell: I used the Facebook application to enter data. Where’s my file? I want to save it on my computer, and maybe use other applications to open it.

What the Friend Connect programs do is let other applications read SOME of your file on Facebook, MySpace, etc. But these programs don’t let you actually take your file, save it, move it, do what you want with it, like the files on your desktop. And they don’t let other applications fully open your file.

Why won’t Facebook and other web applications give you your “files”? Because you didn’t pay for the software. When you buy Microsoft Office, you get a copy of the software to keep and use as you please, so there are no limits on how you use or store the data you create with the application.

But many web applications aren’t charging for the software. Instead, they want to sell ads, i.e. they want to be media companies. That’s how Google, a software application company, got rich. So that’s what everyone else wants to do.

But to sell ads, Facebook et al need your data. And they need you to keep using the applications. And if you can take your files with you, then maybe you — and all of your friends — will start using another application. OR you’ll keep using Facebook, but you’ll create data with another application that Facebook can’t access.

That’s why Facebook created Facebook Platform for others to build applications — so it can keep all the data.

Ask Nick O’Neill puts it plainly:

While I am a fan of data portability, the reality is that true data portability kills social network sites. If we take data portability to the extreme and I was able to export all of my data and contacts from Facebook, Facebook would be nothing more than a well designed communications platform. Perhaps in the end that’s all they will be but for now, their valuations have been based on their skyrocketing user base.

Want to explain “data portability” to a non-geek Facebook user? Ask them if they’ve saved their Facebook file to their computer. Ask them if they’ve backed up their Facebook data. Ask them where their Facebook data is.

Facebook, MySpace and other social networks want to base their business models on the absence of an application feature so basic it’s been around since the earliest days of PCs.

And the reality is that you don’t have to literally save your web application files to your computer hard drive. You can keep them on the web.

But you should be able to put them on any web server you want. And use them with any compatible application. (See Dare Obasanjo for the difference between data portability and interoperability.)

If cloud computing, web applications, and the web as OS is really going to replace local computing, it needs to have more features, not fewer.

If Facebook et al want to have long-term viable businesses, they need to keep users because their applications are BETTER. Not because users have no choice but to keep using their applications, given the inability to save a file.

Comments (17 Responses so far)

  1. [...] the desktop opens up some interesting ideas. Scott Karp brings one of those up when he asks “Dear Web Applications: Were Are My Files?” AIR is a perfect use case for [...]

  2. [...] As some people are astutely pointing out, humans have been moving their data around for years. [...]

  3. Great post Scott. I love the analogy.

    My only point of disagreement is the reference to the competing claims about data interoperability and data portability. We’ve long claimed that one of the goals of DataPortability is an ecosystem of interoperable applications (see a presentation I gave to Australian journalists in early March[1]). We’ve also long claimed that portability isn’t so much about ‘physical’ moving your (virtual) data but being about to port the context of its use ie, using it in another situation. The economic value of data control is not just where it’s stored under your control, but if in fact you can access it when you want to.

    [1]http://liako.biz/2008/03/my-presentation-at-kickstart-forum/

  4. Having used the same analogy for years, it’s great to see it start to become more commonplace. As the web becomes more of an application platform, why shouldn’t it work the same way?

  5. > it needs to have more features, not fewer

    I agree with everything else you have said but disagree with that statement. More features do not equal better. Apple knows this, Google knows this and more software companies are cottoning on that more is not better.

  6. [...] job of exposing the lock-in strategies inherent in social networking sites when he asks “Where are my files?” He correctly ties this strategy to the free, advertising supported business models being used by [...]

  7. Great article. I never really thought about Facebook like this. I have always wanted to download my Gmail though, because if that ever went down then I seriously don’t know what I’d do.

    @Paul – I think more features does equal better. What makes things worse is when you put too many features in plain view. For instance, on Google there are many hundreds of features that are available right from the search field (Site search, inbound links search, math calculations, translations, unit conversions, advanced searches, etc)…The trick is to create the useful features that 20% of the population or less would want and just hide them from plain view. And even Apple has a tremendous amount of features that I discover everyday(being a Windows user most of my life)…Once again though, many of them aren’t obviously there.

  8. @Breck You can download your email in GMail through the POP3 feature.

    I agree that Apple knows how hide complexity. But even so if you compare feature count of say Microsoft Office with Apple’s iWork I am pretty sure Office would win. However I find iWork to be a better product. It does have a lot of features and it does layer them well in the UI but it does have less features.

    For my purposes though Google Docs is even better and it definitely has far fewer features than Office or iWork.

    Feature count doesn’t take into account suitability to the user. You can’t throw the same features at every user and say it is good for the user.

    All other things being equal more features equals better but not all things are equal. There are constraints to play feature count against; UI, download speed/size, responsiveness, security etc. Add a feature and you impact on other aspects of software. There is a limit to how much the UI can layer and hide complexity.

    This applies to everything in life, not just software. I’m glad my commuter car doesn’t have the water cannon feature of a fire-truck. Adding that feature to my Ford Fiesta wouldn’t make it better.

  9. @Paul

    Yes, more features isn’t necessarily better, but I was mostly being facetious in referring to something as fundamental as file creation as a “feature”

  10. I was with you in the first paragraph, but then you started calling it a “feature”.

    I know people who would say being able to set the IRQ of a device is a “feature”. My dad thinks being able to toggle in your boot loader from switches on the front panel is a “feature”. (How will PCs ever get popular if they lack such a basic feature?) I know mainframe geeks who would say “need to think about saving your own files” is an anti-feature.

    I think it’s perfectly reasonable to say that “save a file” is an implementation detail that will go away. It seems really important for a little while, and then gets abstracted away, like IRQs.

    If you want to convince me, then tell me why getting my data out of Facebook (if I had any there, hypothetically) is important. Or why a competing service/product does what they do but better. Otherwise you sound like an old “in MY day…” geek. And this part of computing used to suck!

    In the 1980’s, all my data was on files on my own disks. I had to back it up by hand, I had to carry it with me, and if you had a different kind of computer or application, it was useless. In the 2000’s, most of my data is on the web. Sites like Google are really good at backups, make it available everywhere I go, for free, in easily-parseable HTML, and if I want my own copy I can write a 10-line program to download it all. I fail to see the problem here, or why a site like Facebook should go out of their way to support another interface.

  11. @Ken

    I agree that the notion of saving files will be abstracted away on the web — but what replaces it needs to be at least as flexible. Imagine if Microsoft locked 0ffice files so they couldn’t be opened in iWork. That’s what Facebook has done. The notion of files won’t go away until you have total unfettered control of your data in the cloud, regardless of where the data physically residesm

  12. Good points on the subject. In order for anything to become mainstream, users need to take interest and understand the concepts. The common public understands social networking at this point and that is why facebook and myspace have the success they do.

    I do find it interesting that you mentioned data portability without including a link to http://dataportability.org/.
    The more people involved in making this a success the better off the web will be for it. And, people will truly be able to own their web identities.

  13. [...] “Dude, where’s my file” – asking the portability question the in context of business reality (Scott Karp) [...]

  14. As many others have pointed out, social networks are hypocritical because they rely on the data portability of email apps to grow their own networks, yet refuse to let that data back out.

  15. Hey, Kendra from Box.net here- thanks for the great post. Just wanted to jump into the conversation- here at Box we also believe that enabling control of your data is critical to member’s rights. We have many features that enable people to do whatever they want with their own data, including (easily!) taking it out of our system if they’re so inclined. (www.box.net/info/openbox)

    So as we watch businesses/ social networking sites take to the web as a platform for talking to their audience (customers), it is important that we be increasingly aware of what data they have access to and why. This comes up again and again in discussions around huge companies like Google- and the other companies that they own/run- because Google knows a lot about us, which is good for them and a bit scary for us. My data should be my data. I should be able to save it, send it, share it and have only those I choose have access to it. Total control of your data in the cloud, if you will- I like that Scott.

    Take good care.

  16. I recently came to the same realization about Gmail — about how data portability just isn’t there. But I decided to forget about keeping control of my data and I made the switch to Gmail anyway. Here’s a full post on my rationale: http://www.michaelmadej.com/2008/04/why-googles-gmail-is-able-to-overcome.html

  17. [...] is taking a big shot at Facebook in the PR war over data portability and social network interoperability. I signed in to Google Friend Connect, implemented on the Go2Web2.0 blog, and saw [...]

Add Your Comment

Subscribe

Receive new posts by email