500 words into the future
Unapologetically opinionated views on technology, in the office and out
Friday 6 November 2009, 7:32 PM
Bye bye browser?
On one side there’s Flash and Silverlight and the rich internet applications world, which is working on ways of taking the web outside the browser and onto the desktop, where it “lights up” applications and plugs them into a connected world of APIs and services. On the other is the HTML5 working group, and their vision of a browser that can do, well, pretty much anything. With HTML 5 there won’t be any need for applications – it’ll all be web pages running on super-speedy JavaScript engines and with CSS for look and feel.
Here comes the difficult bit.
They’re both right. There are things a well written RIA can do that a web page can’t, and there are things that web page can do that are impossible for a traditional application. With traditional code you need to push new applications to every desktop every time there’s a change. Even .NET’s click-once and AIR’s self updaters don’t make much of a difference – you still need the latest version of the code to get the latest features, and that (with a flagship RIA like Morgan Stanley’s Matrix) can be a hefty chunk to download. At least with a web page, one change and then everyone who uses it can get access to the latest version.
It’s all a trade off. Not every web site suits every user, nor does every RIA have a fully engaged audience. That’s why so much work is going into getting those experiences right, whether its online design tools like Mozilla’s Bespin, or Sketchflow in Microsoft’s Expression or the designer developer workflow between Flash Catalyst and Flash Builder. But a web page and an application are outside the operating system, and if web-centric OSes ever become common, they need to have some way of supporting and interacting with the web. That’s why there’s so much interest in Google’s ChromeOS and Microsoft’s Windows 8. They’re going to be the first real operating systems of the modern web.
Microsoft has already started work on Windows 8, and like Windows 7 before it, the veil of secrecy has risen over Redmond. There’s rumour, but rumour is never trustworthy. But every now and then a snippet of information finds its way outside the walls. One interesting source is Microsoft’s job advertisements. Hiring people for the cutting edge isn’t easy – you can’t just grab an engineer with an index card in a newsagent’s window. That’s why there’s a lot of information in a job posting – it’s got to attract the right person, with the right interests for a minimum three year slog to the next OS.
So we were intrigued when we spotted this advert:
“The web is the center of most consumers’ PC experiences and the platform of a new generation of developers. This is a rare opportunity for Windows to redefine its application model. Upcoming web applications are evolving features of traditional client applications. We will help them: we will have the best platform for standards-based web; we will help web developers take full advantage of the power of Windows client computers; and we will let end-users experience these new applications in ways that a browser cannot. In short, we will blend the best of the web and the rich client by creating a new model for modern web applications to rock on Windows.
We are looking for individuals who can develop system software, using C or C++, to target third party applications written in the latest client-side web technologies. You need to know HTML, CSS, and AJAX.”
We’ve been discussing the future of IE with other journalists for some time. Is it dead, and does Microsoft want to replace it with another way of rendering and displaying HTML that isn’t a browser? We’re not sure, but the advert seems to make it clear that Microsoft has a strong focus on getting rich internet applications on to the Windows desktop. It’s already demonstrated an offline version of Silverlight, but a tool that could take a web site and turn it into, well, a Windows application would be quite fascinating.
It’s easy to imagine clicking on an icon and getting a familiar look and feel. It’s not that easy to imagine that that application is really a web site somewhere, and the HTML, CSS and AJAX that power it are being abstracted away and delivered as Windows code. That’s the stuff dreams (or even nightmares) are made of, a sophisticated piece of software engineering that renders the browser unnecessary.
Which of course leads us to the next question:
Who needs a browser ballot now?
--Simon
PreviousComments on this post
Good read, and good questions raised. :)
It's a tricky one because local machines still haft to boot up into a launch platform whether that be in the traditional sense from a storage device or something newer, like an embedded option say an extended flashrom.
Obviously the latter would benefit something along the lines of google's chrome operating system somewhat more by eliminating as much overhead as possible beforehand, this would also compliment the mobile platform markets.
The downside though with the cloud markets gaining more and more pace with every moment that passes by is that by the time such OS comes about, just how much of ones own data will the end user's have control over anymore? and what happens when connection outage strikes.
You then have the other approach where as the local operating system still exists albeit probably a lot more streamlined, simply picking & choosing and integrating the best of the web technology's that become available over time, but not dependent on them.
The upside to this approach is that the end users will still retain control over there own data somewhat more so than the previous option, especially if a offline mode is available.
The down side though as already expressed could very well be in the form of software update rollout methods but that may only effect the actual operating systems, if the applications are being delivered via the web interfaces.
But it wouldn't surprise me if all party's involved just ended up building a operating system very similar to one another, there by just integrating the best of the web technology's available during that time, and then just a continuation of upgrades or replacement components during the softwares life cycle.
This comment has been deleted at the users request
