“This exciting project is designed to unify different operating system desktop interfaces into a browser only application interface. It helps users to feel comfortable with any application using the interface they are used to. It helps companies to unify their intranet applications into one desktop interface – built on existing interfaces or one which incorporates their own CI. It helps to design the same interface for all types of devices using browsers like PDA and other mobile devices, Notebooks, Desktops, Tablett Computers and any other future devices which may come up.” Read more at x-desktop.org and definitely check out their demo!
X-Desktop: The Browser Based Desktop
About The Author
Adam Scheinberg
Technology Executive • Web Developer • Father • Foodie • Music Snob • OS enthusiast
Follow me on Mastodon @[email protected]
Wow – It’s just like the defunct desktop.com and webos.com, or even dansteinman.com. It’s just a rehash of years old DHTML crap that sucked then, and still sucks. Useless.
Bloody hell! I really shouldn’t be surprised by this. But, to see what they’ve done is an altogether different experience to merely pondering the possibilities.
It seems that the war on having one interface to applications supplied by the operating system is a never ending story. It is driven by money and the big guys. It is not driven by the user. So let them fight. At least it finally seems that there is some common base finally reached in terms of one interface. The browser.
Actually, I think they have it backwards. It would seem like the users are the ones who want one (read – a consistant) interface to applications supplied by the OS, and the ‘big guys’ are the ones trying to shove applications in browsers down our throats. Of course, I could be wrong?
I don’t know about everybody else, but I’m sick and tired of being told that the brwoser has to run everything but the kitchen sink, as it has been my experience that apps running in browsers are more clunky and less responsive than their desktop bretheran, especially the ones written with Java.
Surely things like .NET will allow applications to run on multiple OS’s, and thus inherit the look and feel of each OS. If you like the Windows look – run your .NET applications on a Windows machine; if you like the Apple Aqua look, run your .NET applications on a Mac; etc. They stand a better chance of running reliably and responsively that way.
Ok. A bloody slow themeable window manager that runs inside of Mozilla. What’s the point of that?
I think this is a really cool/inevitable thing. It could allow people to have more choice as to which OS to run, and could therefore be of great assistance to independent OS’s For those who are inclined to complain about the speed, keep in mind that if you really are using this for your desktop you can greatly reduce the number of other resource-heavy processes running (such as by using twm or fvwm instead of KDE, or just using a SVGAlib-based browser instead of X+Window manager). Also, another good speed-increase could be gained by using a browser with a JIT compiler for javascript, if such a thing does/will/could exist.
It’s neat.
But man it’s slow.
What’s with all the abstractions? Why can’t we just have a bloody single consistent UI that doesn’t suck? Next thing you know we’re going to be running this in a java browser in a pc emulator on a mac.
This just reminds me of those confusing sentences which use the same word a lot “I know you know that Erica knows that I don’t know how inconsiste my UI is” Bah
I think its more interesting in what it shows Standards complient web technology can really do.
Imagine now that you make your site act more like a desktop.
Think if Yahoo, or AOL, or your pit of media/web companys set up there login like this. You could totaly change the experance of a visitor to your site.
This is the type of stuff that i think web developers should be experimenting with to see how to use standards to make sites more usefull to the user. There are things that could be very cool to do, but no one does them becuse
1. so few web browers are complient
2. THe speed and complexity, and risk are large
So keep with your vanila website and just let it be
I wouldn’t hold out much hope of Winforms working correctly on other operating systems. Just think, millions of Joe Moron Windows Programmers wizzing up their copy of VS.NET completely oblivious to the fact that the the GUI library they are using is completel uncompatible with the rest of the planet.
No, the best opportunity should before GTK# community to create a plug in for VS.NET so that Joe Moron Windows Programmer can simply point-click and drool his way through his “programming eXPerience” whilst still remaining compatibile with “everyone else”.
I am typing this inside of the browser window opened through the ‘test URL’ button. The window decoration looks like OS X, but the controls are native IE ones. No speed problems in this browser unlike the ones reported by someone using Mozilla (unless the complaint was about redraw, which I totally agree with).
Such a browser-based interface is already offered by Peoplesoft (which I use at work), and I’m sure by many more companies … so what makes this so important?
Does this require Java or IE to run?
they can only produce slow pigs, IMHO
Just want to make sure that no FUD is spread about mozillas ability with this demo.
I am useing phoenix, and its very very fast.
It’s not that bad but the redraw sucks arse !
It doesnt require IE nor Java, im running it on a java-less box with Mozilla running linux. I’m typing this from it right now. It’s not a speed demon, in fact when i type fast it causes my mp3 to slow down, sorta strange. Neat demo, but thats about it, i dont see the day where i’d use something like this for real.
This would be much faster if it was offered on the companies network rather than over the web. This is a very cool idea and I think it is worth the effort.
The correct method is to make things simpler. That is why they said we should have apps work like the internet. How often are you confused about web pages compared to normal apps?
These people are just making the normal desktop appear in the browser. To which there is no point. This is going backwards for the interface. We need simpler apps not more complicated ones.
I can’t believe this
If anyone wants to know how resizing windows in MacOSX feels like, set the theme to Auqa and try to resize a window:)
I’d like to see a real application used with that interface but I don’t really see how usefull this is…
This would be much faster if it was offered on the companies network rather than over the web. This is a very cool idea and I think it is worth the effort.
How so? All of the data is sent to your computer before you start using it.
XWT — XML Windowing Toolkit — xwt.org
Unfortunately the demos are offline atm due to a hardware failure, but the XWT project is 10x more useful that this.
Think real, responsive, remote applications. It is really cool and about 3 months away from it’s widgets set being mature enough to be a complete application DE.
Why is the philosophy behind UNIX better than the one behind Windows? Because in a normal UNIX environment, you just need the kernel and the init process – just about everything else is a userland app that can crash without taking much else down with it. It’s the simplicity. One thing does what it’s supposed to do. Other than Emacs, you dont really have one-thing-does-everything apps. The load and responsibility is distributed among several processes.
Surely, if you are a normal person, you dont use your fork for eating and picking your nose, do you? So why do we have to have a single app (the browser) that does EVERYTHING?
First, there’s the point-of-failure issue. If the browser crashes, everything you have running will also crash.
Secondly, it’s the simplicity issue. People can and will be confused by an application that does everything except taking the dog out for a walk.
People should use whatever gets their work done and can play a few games if they need. Choose simplicity and power over bloat.
I dare you to challenge my opinion.
Agree with most others, seems pretty cool, but I’m having trouble seeing the point in this. Is this something that people are going to find useful?
I dare you to challenge my opinion.
Firstly, how tasks are carved up is an implementation issue.
Secondly, one app framework is comparable to one API.
Where was that challenge? I didn’t see one.
“What’s with all the abstractions? Why can’t we just have a bloody
single consistent UI that doesn’t suck?”
Because there are genuine disagreements about the best way to
implement a GUI.
You could set up a committee and it would discuss for twenty years but
never come up with anything good.
“Surely things like .NET will allow applications to run on multiple OS’s, and thus inherit the look and feel of each OS. If you like the Windows look – run your .NET applications on a Windows machine; if you like the Apple Aqua look, run your .NET applications on a Mac; etc. They stand a better chance of running reliably and responsively that way.”
Java is doing it for years! Java 1.4.2 release for instance, which will be released in couple of months, will include GTK look and feel and Windows XP look and feel.
Thus, .NET does not offer WORA. Java does.
“The correct method is to make things simpler. That is why they said we should have apps work like the
internet. How often are you confused about web pages compared to
normal apps?”
I agree that simplicity is desirable, especially for beginners, who
IMO should not see more than 30 controls on the screen at once.
But web sites are often confusing too, especially those of music
companies and other large commercial companies.
This project has failed to properly identify the problems it is trying to solve and the goals it wants to achieve. While it is capable of being an interesting project it currently is not presenting us with direction. I found the website descriptions to be incomprehensible. Let’s look at this.
It seems that the war on having one interface to applications supplied by the operating system is a never ending story. It is driven by money and the big guys. It is not driven by the user. So let them fight.
He begins by outlining a problem then states he’s not interested in solving it. He should explain the engineering goals instead of making useless political commentary.
Note that what he means by “interface” is unclear. Are there two interfaces to the applications to that come with Windows XP? If I may be specious, should all applications have the same arrangement of UI elements, labels, etc., i.e., should my calculator app be indistinguishable from my e-mail app by having the same interface?
At least it finally seems that there is some common base finally reached in terms of one interface. The browser.
Here he states there is a common base among “interfaces” called the browser.
If he was talking about the interface to the resources that a browser accesses, then yes, all browsers share common protocols such as HTTP. If he’s talking about the human-computer interface of the browsers then in this sense it can be said that all browsers have some degree of commonality in their UI. But this project does not address the UI differences or protocol differences between different web browsers.
This exciting project is designed to unify different operating system desktop interfaces into a browser only application interface.
Here he is talking about “desktop interfaces”, so in other words, if the Windows XP, KDE, and Mac OS environments were the “desktop interfaces”, then he would be claiming the browser is a common base to all these environments, i.e., all these things are built on top of the browser. This is not true, so he has used “interface” to mean different things in each of the two previous sentences.
What is a “browser only application interface”? No really, what is it? Running the demo, it seems this means it runs inside a browser. Of course the graphical user interface to the browser is already a seperate graphical user interface from what’s running inside the browser which seems to be a imitation of the graphical user interface of an OS. And while it can imitate the look of the window managers of these OS’s, it’s not an integration, because no native OS application actually runs inside the windows of the window manager. It does not unify the various imitation window managers, it’s merely offers a collection of imitation window managers.
It helps users to feel comfortable with any application using the interface they are used to.
Is he talking about being comfortable a web-page like interface, or one of the imitation window managers? The idea behind unification is that everyone uses that same thing. What this seems to allow is every user to use a slightly different imitiation of a window manager. Where’s the unification?
It helps companies to unify their intranet applications into one desktop interface – built on existing interfaces or one which incorporates their own CI.
Again he uses “unify” together with “one desktop interface”, when his demo presents 11 different window managers. He is actually talking about unifying something else but he changed the topic without telling us. Here he narrows the discussion to “intranet applications” late in the game. Also confusingly there is some discussion about desktop interfaces built on top of other things, or is that applications being built on top of “existing interfaces”, or is it applications built on top of desktop interfaces?
It helps to design the same interface for all types of devices using browsers like PDA and other mobile devices, Notebooks, Desktops, Tablett Computers and any other future devices which may come up.
Here he says “same interface for all types of devices…”. So really, he could be talking about my specious calculator indistinguishable from e-mail application example. Could he be talking about an application programming interface?
I ask you, ultimately, what’s inside the imitation windows? Web pages, HTML, XML, Javascript, etc. In the end, the application programming interface is made up of your basic web technologies. Web technologies are inherently designed to be displayed on different types of devices. I hope the project is not claiming credit for this.
Let’s look at the “design goals”.
* Open Source
* 100% Browser based
* Supporting of all operating systems providing a DOM2 / Javascript Support Browser
* No Plugins required
* Simple Object Interface
* Customizable desktop & window skins
* Extensible by user defined functions
At this point we still don’t have a problem to solve. The “goals” don’t provide any clues either. Acheiving the goals does not imply there will be any functionality except a desktop with skinnable windows. Nothing said about any applications.
Let’s try and acheive all these goals:
<html>
<body>
<script type=”text/javascript”>
<!–
//Add your user defined functions here
// –>
</script>
</body>
</html>
The above outline of an HTML page is open source, 100% browser based, runs under browsers that support DOM2/Javascript, does not require plugins, can be accessed via the browsers DOM object interface, and can be extended with user defined functions. We’ve achieved all but one of the goals, and it does absolutely nothing. Now you’re saying this “thing” we’re building has to have a customizable desktop and skinnable windows. Well my operating system already does that.
I really feel like doing a review like JWZ’s (much misunderstood) piece on Linux media players. But instead I’ll be generous and quickly rewrite his PR for him:
X-Desktop allows you to present a virtual desktop inside any modern web browser supporting DOM2/Javascript. X-Desktop allows you to present your browser-based intranet applications inside the windows of a virtual desktop instead of requiring the user to open one browser page for each application. You can customize the virtual desktop to allow easy access to your intranet applications and web-resources just as you would customize the application menus and desktop icons of a desktop machine. The difference here is, you can offer an custom multi-application environment without having to customize each desktop machine because the virtual deskop is delivered to the browser.
As an additional benefit, the virtual desktop can imitate the window management look and feel of various desktop environments so that users may use a GUI style that suits them regardless of the host operating system. Users can also do additional customization such as selecting window colors and skins.
What I’ve done is I looked at his project, and tried to reverse engineer a purpose for it. I tried to imagine what it is useful for, and who could benefit from it. I have made no mention of unifying desktop interfaces, unifying applications, presenting the same interface to two things, or one thing being built on top of something else.
I recommend people go and check this out (and the demos provided).
http://www.xwt.org/
The idea is similar, but the interface is delivered through an applet (ActiveX for Win32, Java for everything else). I was stunned at how fast it was, and how good the architecture seems to be.