2007-01-12

Poisoning the WebOS well

The term 'WebOS', or web operating system, has been useful to gain attention to the idea of creating a framework for rapid web application development and making those applications available from any computer with a web browser and Internet access. This is a noble and important goal. Making web applications can otherwise be quite difficult. We should take advantage of the web to simplify creation, distribution, and interaction of our applications.

The windowing glitz of the current so-called WebOS offerings has bloggers questioning the point. Windows within windows just seem silly. They solve a problem of maintaining the state of an entire collection of applications in an environment, but they they recreate a paradigm that users where there is already a significant amount of frustration: the desktop. I'm not ruling out the possibility of using this paradigm, but it distracts from the more fundamental problems of rapid application development and distributed access.

It would be great if this web-based desktop-like, or webtop, functionality could be isolated from the rest of the WebOS discussion. As long as these webtop offerings are being called WebOSes, as long as they are being reviewed only on their webtop merits, the WebOS well will continue to be poisoned. There really isn't much that can be done in the short term. Eventually, some new terminology will emerge to differentiate web application frameworks that provide glitzy windowing and ones that allow the rich array of web services to be used easily and transparently by even the most novice of developers. Sometimes they might be the same framework.

Read more on this blog about what a WebOS could be, then tell me what I should be calling this sort of web application development framework.

5 comments:

  1. Jadon - Looking for a bit of clarification. Is a WebOS essentially a server local to the browser that lets a web app (client) to operate offline?

    ReplyDelete
  2. What I consider a WebOS is something that turns web services into a commodity and simplifies development of web applications. Adding a local server to allow web applications to work off-line is just one part of that. Have you read my earlier post, defining a WebOS API? Stan does a good job of rounding up the current meaning and points to the other best sources right now: ZDNet and Read/Write Web.

    ReplyDelete
  3. So after reading up on those links - I don't see much differentiation between the definition of a "WebOS"/"WebTop"/"InternetOS" from the Java VM . Why not then rename JavaVM into JavaOS or instead of WebOS, call it WebVM? I understand that offline ability is a part of it, and even some kind of local storage, otherwise what happens when the browser is closed? Would it still be "running" as a service/task?

    From what tasks it sounds like it needs to do, "WebVM" or "InternetVM" seems to be a better description, IMHO. :)

    ReplyDelete
  4. I don't know that WebOS is the right name. As I guess you read, there is a university research project called WebOS. This is a useful effort and I think they deserve to keep the general meaning of what a web operating system is.

    Using the name "WebOS" for a specific class of web application development framework is simply a matter of product positioning. If you called a WebOS a VM, then you'd be building on top of the impressions and expectations of Java. Java doesn't deliver a set of standard template applications, let alone a "killer" application.

    Sticking to the Java terminology, you could call a WebOS a "web development kit". That would give you a bit more implication of a more complete offering. It would imply the required tools and libraries to build applications. It would also imply some template applications. Those applications, however, are not generally expected to be very rich or practical. They certainly aren't expected to be production-worthy killer applications.

    More importantly, the problem is, the Java well has also been poisoned. From a developer's viewpoint, it also cannot be assumed that you have a standard set of resources available in all Java environments. Competing vendors have been able to introduce incompatibilities and operation is not always transparent for consumers.

    I have a bit of preference for the terminology "web application framework", but this still doesn't go far enough to imply how big of an impact a WebOS can have on web application development. Frameworks are typically more limited in scope. I'm looking at something that would have frameworks running on top of it and includes a device driver model.

    I don't think that all of the webtops out there should be considered WebOSes. Working off-line and providing local storage is just the beginning. Services provided by OpenID, Amazon Web Services, Upcoming, Google Apps for Your Domain, and many more should also be provided.

    I'll keep considering the term "WebVM", but I'm hopeful some better term will come along for what I'm arguing should be made and what I'm predicting will be change computing as a whole.

    ReplyDelete
  5. I've been defining WebOS as: Replicating desktop computing functionalities online.

    This would include everything from google docs and zoho to del.icio.us (bookmarks used to be offline) to webtops. APIs and Frameworks are as much a part of this as the goofy stuff. I envision the 'ultimate webtop' as more of a Netvibes on steroids.

    ReplyDelete