Have an account? Log in to leave your comments!
journal: think
Desktop 2.0 - The Operating System Lives On
Everything is done through the web browser. This is why I think that web apps on their own are just a fad and will never reach the level that the hype surrounding them suggests.
Over at Apple Matters there is an article proclaiming that The Operating System Is Dead. It’s not the first and it is by no means the last, but like all articles of the type it is very narrow minded and essentially just bollocks. I’ve been meaning to write this article for a long time and pretty much the gist of it is that web apps are a nice concept but a bad idea. Let me explain....
Web apps are incredibly powerful things. You can access your data from anywhere, using any computer. All you need is a web browser and an internet connection. This however brings you two problems. First off, if you don’t have an internet connection then you are screwed. Secondly, you have to use a web browser. Not only do different web browsers display the same page differently, but if you say, accidentally closed a window, you have lost all your work. Everything is done through the web browser. This is why I think that web apps on their own are just a fad and will never reach the level that the hype surrounding them suggests.
Then you get the problems with data transfer. Transferring data around a computer system, from the processor to the ram or the disk is ALWAYS going to be faster than transferring data back and forth over a network connection. As such apps that use a lot of data will never make good web apps. Graphics apps, audio apps, video apps, all of them are very unsuitable for being made into web apps, at least beyond the basics of say, Paint. I can open a 2 hour long video file in quick time and scrub to the end straight away, with a web app I have to wait for the full 2 hours to download before I can do that.
So does this mean that applications will stay the same and web apps will just die out? No, the power will be in combining web apps with desktop apps. The most common example of this is email. Many people have two ways to access their email: through their email client or through web mail. For those who have a desktop email client set up they can download from multiple accounts and manage their email easily. If you don’t have access to this you can just log into your web mail and check up on it while away from your computer.
There are lots of applications making use of this. .Mac is a brilliant example of a service like this. There are a large number of applications that integrate with .Mac, providing them with an easy platform to web apps. For example, one click and you can upload your holiday photos from iPhoto to .Mac or even flickr. You can then view them, or even manipulate them online if you are away from your computer. This is the way I see applications going.
Instead of the internet making operating systems obsolete, it will in fact solidify their place on our computers (well, the fact that a computer needs an operating system to run will do that but you get the idea). You’ll be able to work on your computer as it is now and then upload your work to share with friends. Then, if you are away from your computer for whatever reason you can just hop on a computer with an internet connection and carry on working on what you had.
I see web apps as becoming a way to get around you not being able to access your work directly, not as a means of replacing those applications. This is the way that both Apple and Microsoft are going, with their integration with web services and I think that Google knows that they can’t compete unless they have an operating system base to work from (which could be why they appear to be getting closer to Apple recently, but that’s just guess work). So despite what people like to say, the desktop app is going nowhere. The only thing that’s going to change is how you access your work when your computer isn’t there.
|
|
4 | 3395 |
| comments | views |
thinkback
well i think you’re both wrong. what’s gonna happen is people will stop expecting the kernel and the interface to be one lump. there will be demand for interface portability, driven by all the new virtualization stuff, the sheer speed of these machines nowadays, and the fact that nobody runs on weird CPUs anymore. so if you want mac look and feel atop a linux kernel, windows atop mac, whatever, it all works, and it’ll be much easier to build apps where there’s one set of common, openly-established services that work on any computer (and i don’t mean “the windows API"). users can choose their favorite interface layer when they log in, administrators can choose services layer and backend based on what they want to do with most of the computers, and apple and microsoft will both be small fish in a big pool instead of their current locked-in arrangements.
I remember an older OS6 or so capability where you could take your OS environment status with you via a floppy disc as well as any data or apps spoecivic to you and not on the computer.
Further if the only limitations are accidentally losing everything by closing a browser window, perhaps all that is needed is a host local application tailored for the task of dealing with remote apps.
As for network bandwidth, that is increasing faster than I would have predicted and even rural folks willing to spend bucks can now just barely access broadband. In the next 2 years or so, portable broadband through 802.11n and 4G will become pervasive. As well as Internet 2 for fixed sites.
Apple could lead on the first and IS leading on the second.
Jerry
To hibiscus, the idea is a nice concept but unfortunately the idea of being able to run which ever UI on which ever kernel would require a whole lot of changes to the kernels and UIs of today. For example, you could run the OS X UI on Linux but you wouldn’t get spotlight because that relies on the kernel. The whole idea of mixing and matching would basically require a lot of work for which Microsoft and Apple would see no benefit (and I doubt the Linux world would either).
You would also have to get people to use these common, openly established services which is not an easy task. Programmers are inherently protective of what they use and in many cases may have lots of code invested in a language/API. It would require all applications now to be completely re-written which is another major hurdle. The most feasible concept is that of desktop apps integrating with web apps as this helps satisfy everyone and solves several problems.
To Jerry: Desktop 2.0 is pretty much the host local application. But instead of having one host application dealing with all web applications you have one host application to one web application. This allows you to work on the data locally with the full power of your machine and also get the portability of being able to work on it on the web.
As for the network bandwidth idea, yes, it is getting a lot faster and very quickly too. But you still get hit by the problem that I can transfer data between my ram and processor at a much faster rate than between say, my computer here in the UK and some server in New York. I guess what I’m trying to say is that there are some people that are saying that web apps will make desktop apps obsolete, yet they are ignorant to the fact that many types of apps do not scale well to web apps.









1.
I disagree about Google needing an operating system. Their main strength is that their services are mostly cross-platform, so people who work on both Macs and Windows can use them. I would imagine that most Mac users have to use Windows machines every now and then out of necessity.