Blog Post

Not Hot: Offline Web Applications

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Before all the 2008 predictions come out, it’s fun to take a look at predictions from last year to see how widely on or off the mark we were. I didn’t write a predictions post last year, but I did comment on Om’s request for 2007 predictions, saying:

My prediction or maybe just my hope: people will stop talking about hybrid web/desktop apps because internet access will become nearly ubiquitous.

Did my prediction come true? Not exactly. Internet access still isn’t ubiquitous (though we’re seeing baby steps towards getting it onto airplanes, and people love their EVDO). More to the point of my prediction, there’s definitely been talk and action about the offline web application story. But still, offline web applications and tools for creating them don’t seem to have much momentum.

Adobe released an alpha of Apollo for building hybrid web/desktop applications, then renamed it AIR. Google came out with Gears to enable offline access for web apps, and used it to give Google Reader an offline mode. Microsoft introduced a Gears competitor. (Note that Adobe AIR follows a different model from Google Gears; AIR applications require an installation and are closer to the desktop paradigm than Gears applications, which run in a browser).

The offline web app story in 2007 wasn’t just about developer tools, either. Zimbra enabled offline access to its email and other collaboration capabilities; Zoho made its word processor work in offline mode.

Despite these developments, hybrid web/desktop apps aren’t exactly buzzy. A Google Trends graph for Adobe Apollo/AIR (the name changed in the middle of the year) and Google Gears shows no trend towards increased interest, suggesting that the hybrid web/desktop application space, at least for now, isn’t gaining momentum.

Adobe AIR and Google Gears search trends

I suspect many people are happy with the mix of desktop and web applications they currently use. Desktop applications suit content-intensive work such as writing long documents (MS Word), manipulating large image files (PhotoShop), and writing software code (Eclipse). Web applications suit multiperson collaboration and communications. And if you want desktop application richness for communications like email or instant messaging, there are plenty to choose from.

What about you? Are you pining for offline access in your favorite web applications? Or are you satisfied with the mix of on and offline capabilities you have with your current toolbox of web and desktop applications?

25 Responses to “Not Hot: Offline Web Applications”

  1. The bottom line is that still major players like Google are working on desktop application. That’s the ultimate proof that an all online world is still way in the future.
    Another nice example is this profile aggregator i use called 8hands, which is the nicest desktop tool and I would never even consider using it if it was web based.
    In addition to all that, you can never forget programs that work on files from the HD which will take tons of time to make internet available in a HD speed.

  2. Daniel

    i think you are right about missing something but when you get the service working on non-IE and non-Windows platforms, it be legitimate to include it in this discussion. For now it is a beta project. and a limited beta project at that.

  3. You certainly missed XIOS/3, Xcerion’s cloud OS, which makes all XML applications written for it also continue to work offline (even though the Cloud OS is delivered as a service over the Internet). Using the offline capabilities requires no installation or download of the user or developer. The cloud OS boots through the net, but seamlessly continuous to work offline.

    This is the only SaaS that bridges the gap between the online and offline world without requiring desktop modifications or plugins from end users or extra development efforts of the application developers.

    Applications written for the cloud OS becomes comparable in user experience to traditional desktop applications, but still haves all the benefits of being Internet applications.

    How does it work? The cloud OS starts an XML Virtual Machine within the browser that runs and executes the applications locally (within the browser). Any data change creates local transactions in the OS transaction manager, which gets synchronized as soon as the OS gets a network connection again.

  4. The reason for the lack of growth or interest for offline RIA is because most companys are still figuring out how to get buziness value out of more consumer oriented technologies like wikis and blogs. Very few companies, and even less CIOs, are ready to take migrate true business applications to the Web. When they do however, probably within 6 to 12 months, they’ll need offline support. The Web is way too powerful of a computing platform not to leverage for business, but at the same time what happens to the poor service tech or salesperson sitting at a customers kitchen table with no Internet access. Adobe AIR, Google GEARS, and Nexaweb are just a few of the companies that seem to have started to offers solutions. Once good case study is Aflac Japan. You can read it here:

  5. I have to agree with @jccodez: on the new look, Anna. The old one was much easier to read and like. Just minor suggestions that you could consider.

    1. The column width in the new look is too wide (hence harder to read).

    2. The rounded fuzzy rectangles in your old site for posts and comments seemed to contain the content visually – nicely.

    I think your old site was one of the best to read. I honestly think the new look is not as good. In fact, I visit GigaOm maybe 3 times now, vs. 7 per week before. If it is thinner columns maybe I will red it 5 times per week, :-).

    Please do not take my opinion negatively.

  6. Internet access still isn’t ubiquitous – but it is just a matter of time.

    Google Gears is a very minor and largely ineffective solution to a larger problem. It is just a stop-gap solution.

    Offline apps are effective so far because they offer large set of features, advanced functionality (scripting, APIs, etc.), advanced graphics, structure, and data interchange. Adding collaboration to them is a piece of cake. MS Office can collaborate with anyone if MS just turns their MSN Messenger technology properly. They just have not done it yet – but it is extremely easy to do.

    The biggest problem in offline (and online) apps is not collaboration. It is the segmentation of data and functions. Most folks have been taught that they should use a word-processor to create a document, presentation software to create slides, and spreadsheet to do serious math. This has created a culture that seeks .doc, .xls and .ppt files to convey their ideas. The online mimics have been similarly fragmented. – discusses Omni-Functionality – delivered over the web. We are not in the 1990s with Windows 98 anymore. We are on a web that has more space, horsepower, connectivity, information and storage than any individual’s PC.

    Seamless web is on.

  7. Alan,

    “AIR is a cross-OS runtime environment for building Rich Internet Applications. The keyword is Internet hence ruling out offline.”

    Having developed several AIR applications, I have to say I disagree. It IS most certainly a tool for bringing web apps to the desktop

  8. Desktop apps are quite a bit more than the ability to work while unconnected.

    The ability to share data with other local apps is big… desktop notifications are as ripe an area as occasional connectivity… local file access is, of course, a significant reason to go beyond the browser model’s necessities.

    Trend lines might be more useful after general delivery to the real world.


  9. You raise very good points in your post. Concerning the online/offline capability, what is crucial is not to be always connected, but to always have access to the data, whether connected or not. If I loose the network connection during a conference, if I am sitting in a plane, if I am in an hotel with a very slow connection, I want access to my data, I want to be as productive as if I had my office’s or home’s connection. The switch from online to offline should be transparent for the user. Data should just synchronize while online and being localized for offline use.
    The other point you raise is about “hybrid web/desktop application not gaining momentum”. I think this is because every single application is looked at individually. Making a web app working locally is not just about localizing the data and the application, but it should integrate a better user experience. This has been done successfully for a few applications like the Time Reader, eBay desktop… The hybrid application should also integrate all web benefits like social experience. And mail + messenger cannot replace the web social platforms.
    To conclude, I am not sure people are happy, but they found different ways to work around it. They adapt to the situation. And this will remain until the “hybrid” experience is brought to the next level.

  10. @Clay: I didn’t claim to be doing exhaustive research. I’ve noticed there’s a lot less talk of RIAs and hybrid web/desktop and offline web apps this December compared to last. I’ve noticed that Google isn’t doing much with offline. I’ve noticed that announcements of offline access didn’t get all that much attention. The Google Trends graph is just one way to check into that. Compare that to the Facebook trend (up — up — up) and you’ll see a huge difference.

    You say current apps aren’t compelling… is there one you think would be? My not-so-exhaustive research was aimed at hearing what GigaOM readers are wishing for in terms of offline web access.

    @Alan: Adobe AIR lets you repurpose web stuff — Flash and HTML — to the desktop. Is that an “offline web app”? Sure could be, depending on your definition of such. These terms, including RIA, don’t have some god-given meaning.

    @jccodez: kryptonite on your eyes? I like the new look, what about it bothers you?

  11. What I want to see is essentially a representational file structure of documents – working, previous, trashed, etc… – that is on my local machine and is also mapped on some online app. For instance, I work in Text Edit all the time, because I’m a low-fi jerkoff. It would be awesome for me if that app would save a local copy as well as a remote copy that I can access and work on when I’m away from my local machine. Right now I have to jump through a couple hoops to do that, and the payoff isn’t exciting enough for me to do it.

    “I would like that, but I’m not willing to work for it.”

    My prediction is that we’ll begin to hear this more and more. We have gone from “I don’t know how” (usability factors) in the late 90’s and early 00’s, to “I like/don’t like it” (experience design) today. We are evolving as consumers of information, and we are rapidly approaching the same level of expectations for our software that we have for all other service and goods providers.

  12. It’s weird that you’re referring to Adobe AIR as platform for Offline Web Applications. AIR is a cross-OS runtime environment for building Rich Internet Applications. The keyword is Internet hence ruling out offline.

    “I debated whether to include it here, but since it was part of last year’s end-of-year buzz about RIAs, offline web apps, and hybrid web/desktop apps I figured it’s worth tracking as part of the conversation.”

    RIA does not mean offline.
    Hybrid web/desktop does not mean offline.

  13. Beyond office-suite type applications for individual users, and offline access thereof, I think the real promise of hybrid apps will be in facilitating interpersonal and interprocess communication in business.

    In particular, what will elevate this category will be multi-user, collaborative systems that enhance such business productivity. For hybrid apps to do this, however, and gain real adoption, it’s critical that the design of the web app reflect the appearance and functionality of a desktop app. Until now, that has not been accomplished, but we’re very close. I think 2008 will be the year.

    Anne, if you’ll be at MacWorld in January, I’d like to introduce you to a company that will be introducing such a hybrid app there. It was first shown at the Apple Worldwide Developers Conference and is specifically designed for small business management in companies of five or more employees. It’s a visually elegant and technically innovative solution for the Mac OS that allows small business owners and employees to quickly become more productive.


  14. They’re not hot because there hasn’t been a compelling app yet. Most of the AIR and Google Gears apps don’t provide new and compelling value.

    And in all fairness, I wouldn’t generalize the popularity based on Google trends data on two keywords – not exactly exhaustive research.

  15. @RJ: true, Adobe AIR is aiming in a fundamentally different direction than simply making browser apps work when you don’t have an Internet connection. I debated whether to include it here, but since it was part of last year’s end-of-year buzz about RIAs, offline web apps, and hybrid web/desktop apps I figured it’s worth tracking as part of the conversation.

    @Rana: FF3 with offline web apps will be interesting to watch. Now that Gmail offers IMAP, I wonder if people are waiting for offline Gmail.

  16. I think combination of FF3 + gmail offline + still limited wireless access
    will trigger the interest..

    Only when popular tools start using the technology it starts getting noticed.

  17. I have to agree with you – a lot of the web-to-desktop application offerings haven’t been so hot this year. If I can get internet everywhere…then what’s the big deal?

    The “big deal” comes when you can provide web-empowerment to your desktop apps. It’s not about bringing web-apps offline so much as bringing the web out of the browser. Why the heck should IE or FireFox reign supreme over all my web-enabled content? That’s the problem Adobe AIR is really aimed at solving.

    A great example of a good RIA that’s going to be even better on the desktop is Buzzword. It’s a very good word processor that’s much easier and more fun to use than other word processors, and it stores all the files on the web, making sharing and access much easier.

    Web/desktop RIA’s certainly won’t save the world. They’re not the best thing that have ever happened to technology, and we’re making tons of mistakes as we learn how to apply them. Still, they have a lot to offer, and I hope 2008 shows this industry maturing and solidifying it’s identity.