The Web Is Broken

The Web Is Broken

Summary: I was reading David Berlind's post about Google Browser Sync and was intrigued by his commentary on "Webless Applications" or as I refer to it, the offline problem. The most interesting part was his mention of JavaDB (which he blogged about here). At first glance, JavaDB seems like a very cool concept: a relational database, built on Java, that has the wherewithal to allow you to surf the web and then "sync" those changes when you get back online. It seems like the ultimate solution to the online/offline problem. Except for one thing, it uses the web.

TOPICS: Browser

ferrari_1.jpgI was reading David Berlind's post about Google Browser Sync and was intrigued by his commentary on "Webless Applications" or as I refer to it, the offline problem. The most interesting part was his mention of JavaDB (which he blogged about here). At first glance, JavaDB seems like a very cool concept: a relational database, built on Java, that has the wherewithal to allow you to surf the web and then "sync" those changes when you get back online. It seems like the ultimate solution to the online/offline problem. Except for one thing, it uses the web.

The web is a broken medium. Don't get me wrong, the content on the web is fantastic. The depth and breadth of information on the web is what makes it so great. The actual experience of getting to that content and interacting with it however, is severely lacking. The way I see it, we have been building on a broken system for years. Just because something works does not mean that it's the best tool to use, and that is what has happened with the web. The web is great for reading content, and five years ago, that's all it had to do. Now we have an unheard of level of interaction on the web. The Read/Write web has become a reality, but the web wasn't built to handle the "Write" part.

Recently we've seen the trend towards the interactive web enacted via web applications that work within the browser and provide a decent level of interactivity. The problem as I see it is that people have gotten used to relying on a browser to view the web. This may sound silly, but there are much better options on the horizon. By drawing the content out of the web and repackaging so that it flows into applications built specifically to consume web data, the potential is limitless. Microsoft's WPF and Adobe's Apollo do just that. They serve the best parts of the web to developers on a platter and it's going to make for some fantastic, rich experiences that simply aren't possible within the realm of the web browser.

ferrari_traffic.jpgBuilding web applications for the browser is akin to buying a Ferrari to commute to work - sure you have a lot of horsepower at your disposal, but you aren't taking full advantage of it - you're driving 20-30mph in stop and go traffic. The web browser will remain important for viewing the content on the web. When interacting with the web however, people are going to find the browser too limiting and they will want more than the browser can offer them. They are going to want to take their Ferrari and make it go as fast as it can. It's a shame people are expending so much energy on solutions that will only get us half way there. The web simply wasn't built to allow interaction, and we should all be moving in a direction that will give us the best chance of revolutionizing how we use the web. We don't need something that makes using a web browser offline easier, we need to shift the paradigm of the web entirely.

Topic: Browser

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


Log in or register to join the discussion
  • Offline Browser Apps Are Inherently Inefficient

    There are many problems with David Berlind?s suggestion for using local storage to support the offline use of web apps. A very important point is, once an application goes offline and you are able to do appreciable stuff with it, you by definition have a rich client application. The application will not only need storage capabilities, it will also need business logic on the client machine as well. Therefore all David is suggesting, is creating a new type of rich client application that is architected around the browser.

    Now someone might ask the question, ?What is the point? If the only way to get a web application to work offline is to make it into a rich client application, why constrain yourself to a design around the browser which is inherently limited? Why not just create a regular rich client application that synchs up with a server?? Therefore the above shows that the only way to efficiently solve the offline problem, is by creating an RIA.

    I believe the real reason a lot of people are resistant to RIAs, is because they make OS? (Windows in particular valuable). Quite a number of people are just put off by MS? success, and would like to skew things away from MS? advantage. But that is not the way to build strategies. If the best solution is a certain route, then the best solution is that route, and the free computer market will recognize it and reward it. If Google and others want to effectively compete against MS, they will have little choice but to take the RIA route, and try to outdo MS in services and advertising. Google et al cannot take only the AJAX route. If they try to, they will fail badly.
    P. Douglas
    • The Windows Draw

      Excellent commentary P. I hope Microsoft can leverage their software strengths with WPF and create some really awesome applications. I think that the best RIAs need to be cross platform, making the OS irrelevant, but it remains to be seen whether or not that can happen. The first step may be a MSFT solution that works really well but only runs on Windows.
      • I agree, but ...

        Ryan, thanks. I agree with you to the extent that cross platform is important for making the user experience consistent across platforms. However two or more tiers of user experiences (WPF, WPF/E) from MS is not so bad. The problem is that if we insist on a one tier of user experience (e.g. WPF/E on older Windows, Linux, and Mac PCs) we will wind up getting a Least Common Denominator effect, where RIA technology is artificially held back because of older or weaker technologies. Allowing for 2 tiers would result in WPF pushing the envelope in new innovative applications, and WPF/E still producing dramatically better applications than current day web applications ? but just not as great as their WPF counterparts. E.g. if MS could come up with templates or facilities in its Expression tools to easily create apps that show 3D relationships among elements in applications, that could drive a new breed of applications built on WPF that we just wouldn?t see if we insisted on just one tier of experiences in RIAs.
        P. Douglas
        • this assumes...

          that MS is going to get right the first, second, or third time.

          I'm not holding my breath. I'm not bashing MS, but most software companies don't get it right the first few times around. MS might be able to shove it down peoples throats, but by the time they get it right, its very possible, that they are perfectly good alternatives.
  • Agree

    But corporate customers need to buy into it. They're still sold on the browser client. They're used to it, and many still hold bad memories of thick clients. They need to be shown that thick clients have really evolved, taken the best features of the web client, and integrated them so that corporate customers can get the best of both worlds.

    I think another thing that should be emphasized is that rich clients tend to be less expensive to develop and less bug-prone than web clients. What they need to be convinced of is it can be just as cheap to deploy a smart client app. as it is to deploy a web app. This aspect used to come down in the web app.'s favor.
    Mark Miller
    • Economics is Key

      You're absolutely right Mark, the corporate customers need to be shown how cost effective it can be. That's going to be a big hurdle, and it doesnt' help at all that these products are all still in beta. Hopefully we'll get there.
  • Browser-Editor

    The irony is that the original web-browser was a browser-editor (that also hid the HTML from the user); Berners-Lee's vision was of something more like a Wiki than the static web.

    Unfortunately Mosaic only implemented the read side, and of course that set the standard for IE.

    The issue with moving away from browsers towards separate desktop client apps is that you're back to the client-server deployment problem. Of course we've moved a good way down that road (if MS and Apple can manage to auto-update the world's operating systems surely most firms can deploy a simple app). . .
    • RIA Update Solutions

      MS has the ability to auto update RIAs that persist on your computer for offline use and other advantages, and also stream RIAs to your computer if you want to enjoy them but not have them persist on your computer.
      P. Douglas
    • Interesting

      Thanks for the info about Berners-Lee Jules, I hadn't ever heard that before. Now I'm off to do more research on it. Thanks!
      • Deeper into the memex

        Hi - just grab a bit deeper into the history of the web and you'll find Vannevar Bush's vision of the memex, the "invention idea of the internet". Then you will see that the web was two-way in the beginning. In fact, the memex idea sounds like a weblog..
        • Deeper into the memex - Links

          Sorry, copied the links from the wrong Wiki...

          Vannevar Bush


      • Weaving The Web

        His own 'Weaving The Web' book is worth reading, at least for
        the historical chapters - he's quite good at giving credit to other
        people. He's also probably more negative towards Netscape than
        Microsoft . .

        A big inspiration for him was Ted Landau's Xanadu project,
        along with existing, non-networked hypertext systems, like

        It's quite a quick read and worth it to understand how HTML
        • Thanks

          Thanks JulesLt, I'll check it out. Also, if you read this, drop me an email. You made a comment on one of my other posts that I wanted to ask you about offline.
  • Too Much or Not Enough?

    It sounds to me like you just want a "better experience", right? So what would it take, really? Does the web really need to do everything? Or are we making it do too much already?

    No, the web ain't broke, at least not any more than e-mail is. Though the latter has a lot of good argument that it is. Considering both ends of the web, the server and the client, not any better solution can handle with much better ease the depth of content that is currently available. Fortunately the web is simply a medium, not the end or beggining of the result. Most browsers allow additional tools to be added which solves the real problem you stated in your article. If proprietary solutions like MS's or Adobe's were ever put in replacement of the web, then the flexibility would falter to a complete standstill.

    The web is just fine the way it is.
    • Not A Proprietary Web

      I don't want a proprietary solution for the web, I just don't think building web applications in a browser is the way to go. We need applications, and we should leverage the good parts of the web, but using Javascript to build web apps just isn't the answer. I don't want to change anything about the core of the web, just change the way people interact with the data it contains.
  • Not designed for 'write'?

    The web is based upon a network between colleges and the US government for sending and posting messages. How is that not designed for write/intereaction?
  • anti centralisation

    I know too little about the web to make any meaningful comment on its intricacies.

    - what I do know is that major centralisation is never a good idea in the long term.
    It may be the most economical and so has a high chance of becoming the chosen way, but that doesn't mean it is the best choice.

    Anyone who has consulted a call centre which has now been outsourced to India will instantly understand.

    The saying that you should never put all your eggs in one basket is as true today as it ever was.
  • I have been saying this forever now

    When will people like Berlind get it? The browser is awesome for read-only stuff (within limits) but the hacks and kludges needed to turn it into an application system are intensely insane. If you don't beleive me, look at the source code to Jakarta/Tomcat or any other open source application server, or read the SOAP protocol and see what needs to be done just to turn the conenctionless, stateless, insecure HTTP protocol into essentially replicating what desktop apps do with zero effort.

    Justin James