Submitted by “Personal computing is currently in a state of transition. While traditionally users have interacted mostly with desktop applications, more and more of them are using web applications. But the latter often fit awkwardly into the document-centric interface of web browsers. And they are surrounded with controls – like back and forward buttons and a location bar – that have nothing to do with interacting with the application itself. Mozilla Labs is launching a series of experiments to bridge the divide in the user experience between web applications and desktop apps and to explore new usability models as the line between traditional desktop and new web applications continues to blur.”
[Source http://blogs.msdn.com/msmossyblog/archive/2007/10/27/mozilla-adobe-… ]
I like the idea of this project. No proprietary run times involved, no for-profit politics. Any normal existing website works. No special development environments needed. Prism just shows you how much difference there is between the open web and the closed web.
Edited 2007-10-30 21:04
Prism is just Firefox with the chrome replaced by DE integration, so it should also work with any proprietary web framework that Firefox can support.
I’d like to see some inter-modal integration between Prism and Firefox. In other words, the user should be able to tear off a Firefox tab as a Prism window or activate the Firefox chrome in a Prism window.
Also, KDE should consider this approach for Konqueror in a future release of KDE4.
I’m still skeptical about projects like GNOME Online Desktop that seek to integrate the web into the desktop at the functional level, but integrating them at the shell level is a great idea. Web apps and desktop apps have yet to really converge in terms of functionality, but they clearly coexist in terms of task management, and they should be treated as equals in this sense.
I hope this spurs some more discussion about launch applets amongst the DE projects, since it’s not clear that the latest implementations (e.g. SLAB and Kickoff) have been well-received.
I heard there was likely to be a menu option, Tools > “Convert to Prism App/Desktop App/Something else”
Kroc, you surely know what you are talking about, right? I have to laugh, like AIR and similar things are trying to go to the desktop. They nearly copy Rebol/View, which had its desktop integration since when, 1999? But that does not matter in regards to your post.
What matter is – I just yesterday gave a read to CSS 2 documentation, because one of my friends wanted me to fix something in his box model. Look, I might not be the best developer, but while CSS model is very powerfull in layouting stuff on screen (or maybe even other media), it is overly complex layout engine. And we are still nearly talking about layout itself. Or – can you give me powerfull data grid without touching Javascript?
The thing is, when you say “no special development environment needed”, you should think twice, what in fact is needed. Html, css, javascript, ajax techniques to make it acceptable. What other things on the server? – what a mess.
Making the long story short – today’s world is trying to use web model for things it was never intended to be used for in the first place. And historically, web model was not redesigned, it was mostly patched, to fit our needs. Or how can you explain me, that in 2007, you applaud things, which are mostly going to appear in FF 3.0?
So much for an “open web” vs the “closed web” …. and look – I am a realist – this model will win, not because it is technically superior, but because it is widely spread and accepted, with seamless deployment. But the “open web” is still “not quite there” 🙂
Cheers,
Petr
So Mozilla is just putting shortcuts to web applications on to the desktop? I’ve been doing that for years. I’m not understanding what is so great about this. The other issue is having each web app open in a separate window. The reason I use firefox is so I can have all the apps on a different tab in one firefox window. I find it easier to keep organized that way.
That and offline storage to keep persistant sessions. It’s nothing amazing or remotely innovative. I’m not sure what their target is.
It’s a simple solution; that’s what’s so great. People want their websites in separate chromeless windows so that if one crashes you don’t lose all your other apps. You shouldn’t need to open a browser just to access GMail, because browsing the web and using GMail can be considered separate things now that “web-apps” are common. The popularity of single-site apps is proof of that. WebMail.app, GMail.app, MailPlane.app and so on.
Mozilla are right on the ball with this. Prism will also allow web apps to be extended just as with Firefox extensions. Add standard CSS to redesign the site a bit to integrate better with the OS, or take on a whole new look. Prism lets you remix your favourite websites. AIR & Silverlight are closed systems for the control of the 1st party – not you. They may use a bit of XML here and there, but that’s just a ruse to push adoption.
Every time you click on the gmail icon on the desktop you are using a browser to access it. I don’t know where you are coming from with that statement that we shouldn’t need a browser to access it. I’m still not seeing what the point is. Its a short cut. I double click on the shortcut to the web app and presto, the browser takes me there. Prism is no different.
It won’t seem different to me and you because we’re technical. We look at a web page and we can see Gecko, WebKit or Trident. Regular users do not. You see a widget and you can tell if it’s native or CSS, regular users do not.
This is about removing the mental divide between web and normal application to regular users. That’s a big big thing – just look at the huge investment that Microsoft, Sun and Adobe are putting into this area.
“It won’t seem different to me and you because we’re technical. We look at a web page and we can see Gecko, WebKit or Trident. Regular users do not. You see a widget and you can tell if it’s native or CSS, regular users do not.
This is about removing the mental divide between web and normal application to regular users. That’s a big big thing – just look at the huge investment that Microsoft, Sun and Adobe are putting into this area”.
Does the average Joe care? I don’t think so. My wife has all her tabs open in Opera: Facebook, Gmail, Google Calendar, etc…She doesn’t even looks at the window title, I’m sure. I don’t imagine her having desktop icons for these web sites, LOL 🙂
This is about removing the mental divide between web and normal application to regular users.
Interesting. Actually, I suspect it is even valuable to knowledgeable users. For example, I routinely startup the little RealPlayer application to listen to internet radio stations. I could use a browser, but for such a simple task that seems like an overkill. Also, I like to keep the player separate and visible so I can adjust the volume or select a different station. I don’t want to have to go flipping through my browser tabs looking for the right one. Other internet applications may lend themselves to similar arguments.
Comparing JavaFX, AIR or Silverlight with simply hiding the menubar and toolbar in Mozilla is so ridiculous, so grotesque, that it sounds like an april fool’s day prank.
Why? Same result on the screen, ne?
JavaFX, AIR and Silverlight reek of over-engineering and E,E,E. Mozilla’s solution is highly compatible and elegant. If there’s no 64-bit AIR runtime for your flavour of Linux, then what?
XHTML+CSS+AJAX is elegant how? It’s arguably the worst mess of a development platform ever created, *especially* for doing an application of any real complexity.
XHTML – Document structure
CSS – Presentation
Javascript (or AJAX as some people like to call it) – Scripting
Seems like quite a neat division of labor to me.
In theory, perhaps, in practice I find it anything but, especially when you factor in having to deal with things like every browser having major quirks in dealing with various aspects of the specs, let alone that the state of Javascript debugging is laughable at best. Most importantly, none of these languages were designed to do what’s being forced on them. They were designed for relatively simple things, not large scale applications and it shows.
Edited 2007-10-31 17:05 UTC
¿*Same* result on screen? You are either joking, or you have never used a web app.
AJAX apps are amazing as far as web pages go, but the user interface is slow, VERY limited as to what it can do, and alien compared with native applications. Oh, and it also feels alien compared to other AJAX apps, come to think of it, because web apps are all different — like we were back to the time of DOS.
And development of an AJAX app is a nightmare, with such a mishmash of different technologies stirred in such a haphazard way that you need a team of 70 experts to write a hello-world app.
Sure, if there is not a Silverlight, JafaFX or AIR plugin for your platform, you’re f–ked, but then you are, too, if there is no AJAX compliant web browser with full Javascript. And yeah, AIR and Silverlight (and JavaFX to a lesser extent) are proprietary technologies. However, JafaFX and Silverlight, at least, are publicly documented and will exist in free software implementations.
If these three competitors are really undesirable (and I only see AIR today as undesirable), then we better come up with something better than Prism.
Yeah I actually read that document and it is written by the typical usability goon who thinks they solved the universe when they are just catching up to the rest of us.
Prism creates local shortcuts that open firefox at a particular URL with a particular title bar name and particular UI elements removed.
Okay. So now when I click a link in my Prism gmail it will open in a new window with a totally different GUI because naturally the rest of the internet requires the use of those removed UI elements? And when I run a different Prism website I’ll get another subset of firefox UI elements? So basically Prism will, in practice, bombard the user with an undefined number of different undefined subsets of firefox user interfaces.
Amazing. The flawed “remove functionality to improve usability” ideology is strong with this one.
Yeah, this is ridiculous, not to mention the mess in the task bar…
I also like this idea, and I think we should maintain the sharp distinction between “web” apps and “desktop” apps, while allowing for some cross-pollination. The standards available now are sufficient for doing the kind of work I feel is really sensical over the world-wide web. Anything heavier, and we’re talking remote X11 connections, and the like.
Frankly I think the idea of taking the entire desktop online is a naive one. While making it possible is an interesting exercise, I think the justification for putting the whole shebang online is pretty weak. “Imagine if you could pick up any computer and work like it was your home/work/whatever desktop.” Well there are too many factors involved, too many variations in hardware and software to make “rich client applications over the internet” into an ergonomic experience. It is more conducive to my getting my work done to sit down at my desk, where I have my filing drawers, and my coworkers to ask than having to shoehorn my work environment into a 640×480 airport computer terminal. The “everywhere computer” is here. It’s called a laptop, and increasingly, cell phones.
I actually think it’s better to require users to track where they put their files, and it makes it easier to focus, organize, and execute their work. Instead of “the file’s out in space somewhere, work on it later,” it’s more effective to think: “copy file to flash drive, get home, and work on it from the flash drive after I put the kids to bed.” It’s more realistic to us as human beings to have something representing our work, than some vanilla “internet.” We’re not computers. Having transparency and abstraction is not always desirable. Sometimes its better just to get your hands dirty, and figure out along the way the best solution to the problems.
Another issue is, it seems like all of these web programming environments are overengineering a simple problem. It’s not as “rich” as having a full suite of internet applications at my disposal, but AJAX seems to be fitting the bill quite nicely in the applications that would require. Maybe the issue is not so much user interaction with these applications as it is maintenance and development of them, and at that I’m at a loss of what to say; I’m not a developer.
That said, I’d like to reiterate my support for Project Prism. I think they have the *right* idea. Maintain a clear distinction between what’s “out there,” and what has a set location, while making it more comfortable to rely on internet services on our desktop computer.
Maybe I miss something, and maybe there are other applications that would be best served over the internet, but I think we technophiles don’t want to admit that hands-on work is what sustains us as people, and having to work on or with something makes us more competent and aware of the context of the whole.
EDIT: Kroc, you beat me to it, it’s all over-engineered.
Edited 2007-10-30 22:22
I can see the point of having something like Flash, or Silverlight to do work, *in theory,* but I doubt it’s actually turns out as useful as we think, because icons, windows, and boxes are what we all get our work done in, all of which are handled fine with javascript. There are some cases where we need more than those, and that’s where silverlight would come in, to say, make a powerpoint replacement, but look at the majority of your desktop applications, and tell me a) which ones you would want to log into a website to access, and b) of those, how many require more than the simple interface paradigms we’ve had for more than 20 years? Would those more complicated applications be usable with yet another layer of abstraction on top of the virtual machines that already run, for instance, .Net applications?
Seriously, this is just Java, reinvented for 2007. See the inroads Java applications have made on the desktop? Anybody? I’m so sick of adding runtimes environments for this or that, when the tools we HAVE work just as well.
Maybe I’m a fuddy-duddy at the ripe old age of 22, but I’ve toyed with these same kinds of ideas for a few years now, and I’ve found it really isn’t conducive to happier, more productive humans to put everything on the internet, just because it’s THEFUTURE!!
Edited 2007-10-30 22:44
“Welcome to the FuturePast”
http://www.penny-arcade.com/comic/2007/09/24
“[Users] are surrounded with controls – like back and forward buttons and a location bar – that have nothing to do with interacting with the application itself.”
I wonder what definition of “interacting with the application” the author is using here
The author means that you don’t need any of the chrome when using web application like GMail.
Some webapps don’t like back buttons; for example.
I have a webapp used daily here. I see it as a collectino of ASPX pages using cookies and other developer magic in a stateless medium to create the allusion of a single compiled set of program processes; that’s just me though. I’m always aware of when I’m in a browser or where my program is running from.
Because it’s a webapp, the browser controls all remain visible (I’ve not shown anyone the trusty F11 key in IE yet). The app has all it’s own links for navigation so it can cleanly cancel processes or whatever it needs to do. If you hit the back button, you end up leaving a sesson open while your “back” session opens as a new connection. The webapp also chokes on the back button if it had form data or similar cached information because it doesn’t repost the form data by default. In both cases, using the browser’s controls breaks the webapp.
Since the webapp has it’s own control buttons coded within the pages and breaks when you use the browser forward/backward controls, this may be an example related too “interacting with the application.”
I would like the idea, but after playing for 5 min, I’ve found, that I do not worry as much about missing “desktop integration” as about missing “browser integration”. Really, there are a lot of extensions, like spell-checkers, translators, customizing scripts, notebooks, download managers, themes, etc. What if we hit a link? Should we navigate or open a new window?
I didn’t expect it this way. Nowhere in the document do they talk about compiling the web application code to make it faster. It will be as slow as Yahoo!Mail Beta.
I believe in storing all information on central servers, being Google’s or on your own secured server. But on the desktop, nothing beats compiled desktop applications in terms of speed, CPU and RAM usage. The problem with desktop applications is that they store information locally (ie: duplicity between your office and your home). They should sync every minute the user information with a remote server. This way, the user would still use a snappy application (ie: Thunderbird, Opera) and have his information kept in a central place. Thus being able to access the same info in the office, at home, at a friend’s, on a trip, etc…These days, browsers already store bookmarks in a central place, you just have to log in to retrieve your bookmarks off the remote server.
I’m definitely not ready to stop using a fast compiled C++ application for a web application that has tons of HTML, CSS and Ajax code under the hood, even if in the title bar it says “Mozilla Prism” instead of “Mozilla Firefox”. For me speed and low CPU usage are more important. I can always add a third-party application like Bacula to sync my info with my remote server through SSH.
Firefox will be moving to Tamarin (Hopefully within the Firefox 4 timeframe) the ECMAScript Virtual Machine donated to OSS by Adobe (and used in Flash for Action Script). A Javascript VM is going to make things fast, like 10-100x faster. There’s even movement to eventually support Python as a scripting language in-page and in-chrome.
Maybe I should patent it ?
It is called: “Web-based Web browser”
😉
This isn’t new, this is exactly the same principle that Apple Dashboard and Opera widgets work on – the browser as a standalone, thin app platform. Opera’s never quite made the jump to having a completely stand-alone runtime on the desktop, but they’ve already implemented an app engine on the Wii, and widgets are now becoming available on the Opera Mobile browser.
‘Web applications’ are created because webbrowsers are the only real standard across multiple platforms for offering remote applications. They are not a good thing, they are an act of desperation by a generation of html+javascript+php programmer.
I will be much happier when there are no longer any ‘web applications’, but instead we can get back to creating ‘remote applications’ with local interfaces.
The technology for such has been around and in use for decades(longer than the web).but instead of using it. Developers and users are going stuck in the hypertext document orientated web browsering program.
I don’t understand why some people don’t seem to realize that implementation is just as important as functionality. That’s one reason why firefox is preferred over other browsers (including the original mozilla suite)
sounds to me like google gears, without the offline part…
I installed it and played around with it. I am not really sure what they are trying to do with this. The browser is a very functional part of the web apps, and the prism is basically the browser stripped down to nothing. It doesn’t take very many clicks to go to the site from my bookmarks list or if I send a shortcut to the desktop, basically gets me the same thing. I’m not sure how your computer works, but my browser only takes a few seconds to load, and it usually stays open all day long on my computer. And there’s no tabs! Firefox has gotten me spoiled with tabs. I’m just not real sure what they are trying to accomplish here. Adding shortcuts to the Programs menu, and desktop? Why would I want the extra clutter?
Does anyone know if Mozilla at least provides chrome such as a status bar so you can be sure of the domain of the app you’re using, links you click, and whether you’re using a secure connection? The lack of such info could revive past security issues with chromeless browser windows.
I’m a bit skeptical on this … I have to say, that I often click the back button when using GMail or Facebook. Sure, they’re apps, but they are still document-centric apps (look at this “sheet” of info, that “sheet” of info) with a workflow that lends itself to tabs (I’m going to open the profiles of these friends, these emails).
This seems like it might make using the web more comfortable for someone who’s never used the web before, that’s it. But nowadays, most people are more familiar with the web than with the desktop metaphor.
It seems like some apps are just better suited for the browser, and others for the desktop. Maybe something like drag-and-drop downloading/uploading, or inter-website transfer (with warning dialogs) would be interesting, but you always have to take into account that web-desktop integration always opens new security risks. We had the best integration from a *feature* standpoint with ActiveX, remember?
But, this doesn’t seem too bad … there’s no extra work on the part of web programmers, so sure, give the user an option to create shortcuts to internet apps without the browser interface … who knows, maybe some will prefer this. However, some websites used to forcefully remove all the toolbars years ago, and most people hated this.
What really needs to happen is for Mozilla to release an IDE (probably Eclipse-based) that can rival Visual Studio for creating web apps, or C# guys are going to run with Silverlight.
Edited 2007-10-31 18:00
“Come marvel as we try out some new buzzwords. We’ll stick with the ones that get blogged the most.”
High fluff, low content…
Come to read up more on it, there’s not really much to this, other than a neat marketing campaign. Codewise, I could probably bang something like this out (with C# and .NET) by myself in a few weeks…