I agree that a 'WebOS' is slightly silly, but I believe it to be a marketing name for a much stronger concept - the 'InternetOS'. Today's OSes were largely built for a disconnected world - the 'my machine is an island' scenario - which isn't necessarily the case today.
The pertinent question is: where do today's systems fail us in an always or often-connected world? Maybe you'd work on the networking stack, providing services that made transferring between connections as seamless as possible (hopping from public wi-fi to 3G to WiMax, and prioritizing traffic based on network bandwidth and costs). Maybe we change filesystems to better support synchronization - for example, build in better change tracking so that we can synchronize using the minimum amount of bandwidth. Maybe you'd look for ways to be able to transfer the state of your applications, so you can work on one machine, and then continue working on your phone, then on your hotel TV, then back to the original machine. What about collaboration and communication? How can we build applications that seamlessly combine web information with local information (maybe even blurring the difference)? Not everything belongs in the OS, but the supporting functionality certainly does.
None of these ideas are particularly new, but today's software stack is in need of some pretty serious evolution to get the most out of the promise of a connected world. Javascript based desktops are an interesting prototype - maybe Javascript & HTML would be a good technology for portable applications. I'm not going to dismiss the bigger need because today's 'WebOS' systems are little more than toys...
Why can't we just have an internet FS? Compile all the apps down to a few common byte code formats and allow for namespace control via chrooting and similar mechanisms. No more web browser! We'd need to make symlinks more powerful, to enable opening a connection -- but there'd be little in the way of special UI required.
The technology for that is already there with S3 if you take full advantage of ACLs. It would just be a matter of agreeing on a standard for shared access.
I think as there are no examples of a unique type of application on these platforms, it's hard to see what the point of them is.
More over developing apps is still left to programmers (and a select group of them who are always interested in learning any new platform/language), it doesn't make it easier for Joe Average to do any sort of involved tasks.
Specifically with YouOS (the only one I tried), why do I need to understand applications and installation?
People want to ability to control their computers, not just use them. As Hamming said the purpose of computing is insight. If all you can do is install a new calculator, I don't think I can be bothered. I've got one which works just fine.
Just moving the desktop to the browser isn't enough of a differentiator.
Maybe this model might work for mobile phones, which is unlikely to have the full range of apps that a desktop has.
I disagree with this. An OS is whatever apps run on, and you could write things that most of the next generation of web apps would run on. I believe Parakey was (and perhaps still is) intended to be such a platform.
I think the author is referring to the OS role of controlling applications centrally, the way the f8 platform controls facebook apps. Something that a lot of decentralized web apps run on is not the kind of OS the author is talking about.
Does anyone use any of these services? I've always thought it was best to let the underlying real operating system take care of windowing and let each URL be one application.
I have same feeling. Youos is wonderful but I always feel it waste too much screen space. On the other hand current browser is not very suitable for running multiple web apps also. Maybe we need a better container...
But really, what we are talking about is just a resurgence of thin clients. And why not? They make sense. No updating software, all your files are there for you whereever you go and today we have the bandwidth to handle it (we do vid downloads without batting an eyelid!).
Especially with Google's gears coming along, it will mean we can do everything offline. It's getting close to CMS your home dir and apps.
When I was in high school (15 years ago or so) I said the perfect computer app would be one where you didn't know it was there and just used it to do what ever you wanted, like in TNG. This is getting close!
1. Privacy (and don't tell me it is overrated, it's not)
2. Reliability (yes, I do understand my hard drive can fail, but my OS doesn't delete my e-mail just as GMail did)
3. And the fact that I have to go online to fetch my file
Just my feeling:)
1. It's better to run web application with more screen space, not every app need menu bar, toolbar, etc brought by every instance of browser.
2. For browsing, tab is excellent. For web application, tab is not enough. For example, I can't implement an "always on top" thing using tab, I can't let browser remember every application's location and size. These features are good for productivity.
Current browsers are designed for maximum surfing productivity and not dedicated to provide a perfect web application container.
The pertinent question is: where do today's systems fail us in an always or often-connected world? Maybe you'd work on the networking stack, providing services that made transferring between connections as seamless as possible (hopping from public wi-fi to 3G to WiMax, and prioritizing traffic based on network bandwidth and costs). Maybe we change filesystems to better support synchronization - for example, build in better change tracking so that we can synchronize using the minimum amount of bandwidth. Maybe you'd look for ways to be able to transfer the state of your applications, so you can work on one machine, and then continue working on your phone, then on your hotel TV, then back to the original machine. What about collaboration and communication? How can we build applications that seamlessly combine web information with local information (maybe even blurring the difference)? Not everything belongs in the OS, but the supporting functionality certainly does.
None of these ideas are particularly new, but today's software stack is in need of some pretty serious evolution to get the most out of the promise of a connected world. Javascript based desktops are an interesting prototype - maybe Javascript & HTML would be a good technology for portable applications. I'm not going to dismiss the bigger need because today's 'WebOS' systems are little more than toys...