New Linux Desktop Environment Built on Firefox

IL-CSIXTY4 writes "'Pyro is a new kind of desktop environment for Linux built on Mozilla Firefox. Its goal is to enable true integration between the Web and modern desktop computing.' This looks like an interesting marriage of the web and the desktop. In Pyro, Web apps run in windows on the desktop, right alongside desktop apps (through compositing). Features expected in a desktop environment, like task/window selection and an Expose-like function, are written in Javascript." "

Comments

Does this mean...

[realdodgeman]

...that I can write application in php+css?

Re:Does this mean...

[Xiph]

you can do that already, You run it with apache and firefox.

Re:Does this mean...

[realdodgeman]

I know, but I meant applications that integrates with your desktop..

Re:Does this mean...

[Anonymous Coward]

Would require either a XPConnect SAPI for PHP or a PHP compiler that emitted javascript (or compatible bytecode).

Basically the answer is yes, Tamarin has the potential to be everything parrot promised. But for now, you're probably better using javascript.

Re:Does this mean...

[tomhudson]

oh great ... that's all we need, a bunch of pyro-mainiacs.

See, you've already melted their server ... didn't anyone teach you not to play with fire?

Re:Does this mean...

[Zonk (troll)]

Probably, but since this desktop runs on xulrunner/firefox it will likely eat up 1GB+ ram by itself, ignore the system themes, render ugly fonts, render ugly buttons (there is a workaround), freeze often when trying to open new windows/apps/etc, and make Win95 seem like a rock solid desktop.

This does sound like an interesting idea, though. Hopefully someone will implement an alternative with Webkit (there is a GTK+ port in progress).

Re:Does this mean...

[DrSkwid]

Apache is horrifically bloated for using as a personal web server.

Re:Does this mean...

[ShieldW0lf]

Didn't everyone agree a long time ago that integrating IE into the OS and using it as a shell was a bad idea?

So what is it that makes this any different?

Re:Does this mean...

[Anonymous Coward]

Aside from the javascript vulns it was a bad idea because IE ran as root and had ActiveX controls. Note that we do not hear such criticisms about konqueror.

The next-gen web browser has almost everything you need for a desktop (there's even work underway to get a subset of OpenGL in canvas) I work ~90% from the command line, the main reason I run X and a DE is to have firefox open so why not run it as the WM?

Re:Does this mean...

[bl8n8r]

> So what is it that makes this any different?

It's entirely possible that the wm could run as the user "joeturd" and not root, as in windows. Still, yeah it's another example of sexy technology winning over common sense and security. Most people really don't care because they do not understand the implications of a cracked box. I've tried explaining "keylogger" to people before and they just look at me like I was talking to them in Estonian. They do not, and don't want to, "get-it"

Re:Does this mean...

[Kalriath]

No, it ran as the currently logged in user. The problem is that Windows insists during setup that the user create an administrator account for everyday use. It's the same scenario. (Well, minus the horrible ActiveX crap - noone really likes that)

Re:Does this mean...

[sarathmenon]

> So what is it that makes this any different? It's entirely possible that the wm could run as the user "joeturd" and not root, as in windows. Still, yeah it's another example of sexy technology winning over common sense and security. Most people really don't care because they do not understand the implications of a cracked box. I've tried explaining "keylogger" to people before and they just look at me like I was talking to them in Estonian. They do not, and don't want to, "get-it"

There is a difference. KDE has had kioslaves for years, where the browser can open more than 20 different protocols, including http,ftp, webdav, ldap etc... When you look at it, the sound design of it wins you with all the advantages it can offer. Everything is a different library, and it takes a good exploit in one to compromise another, something which hasnt yet occurred iirc. Plus the benefits are easily observable, which include opening an mp3 over ssh in amarok, without the app even knowing that such a thing is happening. IE's integration AFAIK was just so that microsoft had an excuse to ship the browser as standard, and had no such advantages.
All said, doing everything via the browser is a bad idea. They should have something like applets which can use the gecko engine to make calls outside, and not have the whole os run off the browser. C++ exists as a system language for a reason - thats speed, performance and stability. I am not trading my desktop for a javascript app, heck I thought vista was a mistake with all its memory requirements!

Re:Does this mean...

[toshok]

IE + win32 is pretty different from what we're doing. That combination was considered bad by the general public for a number of reasons, some of which are:

1. Anti-competitive behavior. Which was a stupid, *stupid* argument to make. I don't want some judge ruling on what my code can or can't do, and neither should you.

2. IE had (and has) an awful track record with security. Firefox has a great track record, I think you'd agree?

3. ActiveDesktop was created during the "push-technology" era, and so was mostly used as a way for content providers to stick things on your desktops. Netscape actually had a very similar product (pointcast, I think?) around the same time, which AD was created in response to. Thankfully pretty much everyone has decided that push = bad.

pyro isn't about push technology, and its goal isn't to provide a place for CNN or ESPN to put headlines. It's about looking at what users are doing and seeing that more and more things are being run in the browser, and in the future even more will be. why not afford these web applications the same sort of status on the desktop that native apps enjoy today? Why should there be any difference at all?

Re:Does this mean...

[ShieldW0lf]

The thing is, my browser is the most vulnerable application I run. I connect to the most disreputable sites with the browser and it just interprets what they send me. Sometimes by accident, sometimes because that's what I'm looking for... doesn't matter.

So if my browser gets owned by someone and they're able to make it dance to their tune, what ramifications will having Pyro installed when it comes to consequences?

Re:Does this mean...

[NotFamous]

Check your environment variables. Here's mine:

c:>set
...
MS=bad
OSS=good
...

Yours may be different.

slashdotted after the first comment

[discord5]

Already slashdotted after the first comment, so ... this is what the future web-desktop will be like huh?

Re:slashdotted after the first comment

[AvitarX]

So I've heard for a decade+

Re:slashdotted after the first comment

[jkrise]

Already slashdotted after the first comment, so ... this is what the future web-desktop will be like huh?

Not if the server is within the intranet. Here's the text from the site:

What is Pyro?

Flickr Add-on

Exposé-alike

Window Picker

Pyro is a new kind of desktop environment for Linux built on Mozilla Firefox. Its goal is to enable true integration between the Web and modern desktop computing.

By merging the Web with the desktop, Pyro offers the first big step toward a new future for the Web and the applications built for it.

In Pyro, Web content is no longer confined to the browser's window. Instead, trusted Web sites and extensions are given access to the full range of interactivity and control enjoyed by native applications today.

Imagine...
Rich Web pages running side-by-side with native applications
Single programming environment for the whole desktop
Desktop-wide mashups, killer Web integration
Novel desktop effects

Pyro enables a desktop that tracks the latest in Web technology, and helps mold the future of the integrated Web.
[edit]
NEWS

From Ars Technica

July, 20 2007:
Pyro project offers Firefox-based desktop environment on Ars Technica, by Ryan Paul.
Pyro delivers Web apps to the Linux desktop on DesktopLinux.com.

Check out the slides!

July, 18 2007:
Pyro Announced during GUADEC '07 Conference Keynote Speech.
[edit]
How does Pyro work?

Pyro works fundamentally by drawing your entire computer screen as a Web Page, all from within Firefox. Indeed, at the core Pyro is simply a window manager which renders Web content alongside existing native applications.

By leveraging the trusted Firefox Add-On system, all the capabilities of dynamic HTML, JavaScript, CSS, SVG, and Adobe Flash are available to enable incredible applications, extensions and themes.

Bringing all these Web technologies together with the newest generation of Linux display technology, called window compositing, allows Pyro to integrate native applications as an intrinsic part of the overall Web Desktop, seamlessly merging the two.

Re:slashdotted after the first comment

[Anonymous Coward]

Yes, but they should change their 500 http error page to look like the BSOD that we all know and loathe.

Re:slashdotted after the first comment

[paulatz]

Yes, it is slahdotted, but coral cache is working: pyrodesktop.org.

Re:slashdotted after the first comment

[smallfries]

Maybe I'm being naive, but how would this be different to running a separate browser window for each page, without any navigation controls. You know, like some really nasty site interfaces do before you navigate away from them really quickly...

What is the point? Why does it need a separate window manager? Why on earth does the summary mention compositing when it doesn't appear in the article?

Re:slashdotted after the first comment

[Anonymous Coward]

nope. http://www.pyrodesktop.org/Screenshots
only the wiki part got slashdotted.

Re:slashdotted after the first comment

[ArsonSmith]

Like they always say, "Those who don't learn Windows are doomed to repeat it."

I think that's how that goes right?

Re:slashdotted after the first comment

[Heembo]

It could be worse, we could all be forced to use proprietary locked down mac's with Steve Jobs permission to only use the Safari Browser! SOMEONE HELP! AHHHHHHHHHHHHHHhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh hhhhhh...........

Re:slashdotted after the first comment

[toshok]

I'm not sure I understand your first paragraph well enough to answer.. maybe you should go read the article? Pyro has nothing to do with browser windows you look at webpages with. those will still show up as they normally do, with navigation controls.

Your second paragraph makes me think you didn't read it either.. "compositing" is mentioned quite clearly in the "How does Pyro work?" section:

Bringing all these Web technologies together with the newest generation of Linux display technology, called window compositing, allows Pyro to integrate native applications as an intrinsic part of the overall Web Desktop, seamlessly merging the two.

Re:slashdotted after the first comment

[omfgnosis]

Apparently using browsers other than Safari deletes your brain...

Re:slashdotted after the first comment

[smallfries]

Perhaps it's not just that I read the article, but also that I understood it. I get the impression that you didn't. The "clear" mention of compositing is:

Bringing all these Web technologies together with the newest generation of Linux display technology, called window compositing, allows Pyro to integrate native applications as an intrinsic part of the overall Web Desktop, seamlessly merging the two.

I'm not sure how much you know about a desktop/windows under X so I'll go from basics. When you render content to a window you update the client window. This is the part inside the standard window manager decoration that draws the borders / close buttons / title bar etc. Inside the client window of a normal browser, the browser draws a bunch of controls (address bar / buttons / tabs etc) and then the content within a child window of the client window. To "run" a web-app alongside other windows seamlessly all you need to do is remove the extra frame from the content window with the controls.

This can be done in javascript on a standard browser. There is no need to write a new window manager, and window managers shouldn't be concerned with the contents of the windows they manage. They should be neutral to the applications. Their aim is to run webapps seamlessly alongside native apps. This is not a problem currently. It doesn't require a solution. If it did then a custom window manager would not be a good way of solving it.

They describe all this in non-technical hype and buzzwords, as quoted above and then throw in that they can solve this using compositing. Compositing is the process whereby a window is rendered to an offscreen buffer in OpenGL so that processing can be applied to the image before display. Normally this is used for drop-shadows / proper transparency and other special effects. It has no connection to what they've described at all. Hence the questions.

Re:slashdotted after the first comment

[toshok]

I'm not sure how much you know about a desktop/windows under X so I'll go from basics.

First off, I'm one of the pyro authors - no need for the remedial (and in some cases incorrect) X lesson :) And no need to explain to me what our aims are.

Compositing is the process whereby a window is rendered to an offscreen buffer in OpenGL so that processing can be applied to the image before display.

"Compositing" has nothing specifically to do with OpenGL.

The OpenGL compositing managers use the X server composite extension to map window contents (through the backing pixmaps the composite extension creates) to textures.

Pyro uses it to map window contents to canvas elements.

And you're still thinking in a desktop centric way. It's not our goal to just run webapps seamlessly alongside native apps. That's just a necessary step along the way to the ultimate goal, which is a entire desktop built from web technology. It's also a (imo brilliant) proof of concept - that the gecko layout engine (as well as javascript and css) are up to the task. If it can run a window manager, why not application $X?

Re:slashdotted after the first comment

[smallfries]

First off, I'm one of the pyro authors

Ahh, whoops. Now I look like a bit of a tit. Sorry for being patronising I didn't realise that from your previous post.

Ok, so serious question. At the moment we have a way of building applications, as code talking to the various X APIs. HTML/css/javascript etc are a different way of building apps. You want to replace the entire desktop with the web style way of doing it - but isn't this just reinventing the wheel. I'm not trying to troll here, but I just don't see what's so cool with replacing one bunch of language / APIs with another in this context. But then again I didn't really see the point of active desktop in IE4. So what is so cool about rewriting the desktop/applications in HTML and friends?

IE4 Anyone?

[Anonymous Coward]

Didn't we learn our lesson with Active Desktop? This is one of the reasons I use Firefox instead of IE. It's not so tied into the OS that when it crashes, it's taking down other apps as well.

Re:IE4 Anyone?

[realdodgeman]

It is not the same. Active Desktop was intigrated with your desktop, Pyro IS your desktop.

Re:IE4 Anyone?

[Joebert]

Sounds like a great way to bypass annoying browser security vulnerabilities.

Re:IE4 Anyone?

[Glytch]

It seems like the opposite to me; an easy way for malware writers to not have to bypass the barriers between the browser and the desktop in the first place.

Re:IE4 Anyone?

[Joebert]

Yeah, but at least we'll get rid of browser vulnerabilities.

Re:IE4 Anyone?

[Eddi3]

Look; the only reason they made this is because a lot of people thought it would be a good idea. Now some of the same people are shooting it down? I don't get it. I guess I must be new here, or something.

Re:IE4 Anyone?

[nnm.one]

JavaFX is the way to go.

We need to integrate the connectivity of the web and web dev. languages with your desktop, and not merge our desktops & web.
Besides, it won't go further than some mashup apps or some half-assed Office suite, web apps will never replace read IDE's or video editing software or just good old irssi :P.

Re:IE4 Anyone?

[digitalaudiorock]

Didn't we learn our lesson with Active Desktop? ...and why would anyone want this anyway? The only real reason MS did this sort of thing was to support their legal argument that IE was a necessary and integral part of the OS. This is just as bad as the awful practice of embedding other applications in the browser by default instead of launching the appropriate applications externally (konqueror for example). Why does everyone want to copy all the worst ideas MS has had for browser functionality?

Re:IE4 Anyone?

[Anonymous Coward]

I wonder if perhaps the people who are shooting it down are not the same people who thought it would be a good idea.

Re:IE4 Anyone?

[TheRealMindChild]

Think of it like this... HTML is the idiots way out of writing GUI code

Re:IE4 Anyone?

[kimvette]

Well, providing the user is prompted with a [Continue] [Cancel] dialog, everything will be just peachy. ;)

Re:IE4 Anyone?

[Anonymous Coward]

Hardly. RAD tools and the stateful nature of desktop apps make them WAY easier to develop than capable web apps. About the only thing easier about web app development is flexible layout.

Active Desktop was about bringing bits of the web to your machine to sit behind your apps, much like would have taken you into the browser to pull up a dashboard, portal page, or what have you. Do remember that around the time Active Desktop showed up "web portals" were all the rage, and at the time MSFT would have done anything to keep that kind of user looking at their desktop apps.

Re:IE4 Anyone?

[Old Benjamin]

One reason that this is good is because it is exactly what M$ fears. The reason M$ fought (fights) Firefox so much is that it will make it all that much easier to switch to Linux, or even Mac, when the browser doesn't change. Instead of switching the platform under the browser, this just puts a platform in the browser. Now it will be even easier to switch.

Re:IE4 Anyone?

[moosesocks]

Not to start up the old KDE vs. Gnome argument, but Nautilus embedded Gecko for quite a few of its earlier releases.

As for me, I'm slowly being converted into an Xfce user. If they can add a bit more built-in functionality than it already has (there are some options they left out that really do need to be configurable), they'll have a winner on their hands.

Re:IE4 Anyone?

[totally bogus dude]

RAD tools and the stateful nature of desktop apps make them WAY easier to develop than capable web apps.

Exactly... that's why only idiots do it!

Re:IE4 Anyone?

[SanityInAnarchy]

There's no practical reason that IDEs and irssi couldn't both be web apps. They'd need some sort of backend that you control, but the interface would be plenty responsive enough.

In fact, if you're willing to limit the IDE to a JavaScript-only IDE, and drop Perl and plugin support from irssi (or replace 'em with a JavaScript API), you could do it all with clientside scripts.

You are right about video editing software, and I don't think Flash is anywhere near ready to do that, even if it was a good idea. And the same is true of games. And you may even be right about JavaFX being better. But the fact is, your typical desktop business type doesn't need a more responsive interface than what web browsers can do.

Re:IE4 Anyone?

[Elf_h34d3r]

Don't you mean [Cancel] [Continue]?

Silly backwards Firefox logic....

If only...

[jkrise]

the apps include a very simple word processor and a spreadsheet that could work from a server hosted within the company intranet... this would be a very useful project indeed. Basic features would do - no need for all that fancy schmancy stuff.

Re:If only...

[realdodgeman]

Common. It is version 0.1. If you want a word processor, make your own in php with Open WYSIWYG (http://www.openwebware.com/products/openwysiwyg/) .

Re:If only...

[ImaLamer]

Basic features for the office? You have a job in IT don't you?

When you are doing (database/data) analysis and reporting on spreadsheets you certainly don't want to use the Microsoft Works style answer to 'spreadsheeting'.

Re:If only...

[misleb]

the apps include a very simple word processor and a spreadsheet that could work from a server hosted within the company intranet... this would be a very useful project indeed. Basic features would do - no need for all that fancy schmancy stuff.

Because hardly anybody WANTs a simple speadsheet and word processor. Particular businesses with an intranet. People LIKE the fancy-schmancy stuff. Putting it on the intranet doesn't make a simple spreadsheet program any more attractive than if it was on your desktop. Simple (often free) spreadsheet and word processor programs have been around, like, forever, but people just go out and buy MS Office anyway. That is just the way it is.

-matthew

Re:If only...

[Miseph]

*Whoosh* Guess that one just went over everyone's heads.

Let me give you guys a hint: GOOGLE HAS ALREADY MADE AND LAUNCHED A WEB-BASED OFFICE SUITE. I'd be surprised if Google had absolutely no part in developing, and if they aren't yet then I'd be surprised if they still don't get involved now that it's gotten some press coverage.

Re:If only...

[Anonymous Coward]

Doesn't look like it supports paragraphs properly (inserts br only in Firefox, no headings/normal etc) which makes it pretty useless for any real word processing, even as basic as we are talking about here.

Somehow familliar

[Kazuma-san]

This strangely reminds me of Microsofts Desktop in Windows 95/98 and the resulting law suits. I am no programmer, but wouldn't the performance of a desktop system written to support java script etc. be lower than that of a regular written desktop? So, in the worst case it would slow down the whole system.

Re:Somehow familliar

[haakondahl]

This strangely reminds me of Microsofts Desktop in Windows 95/98 and the resulting law suits. I am no programmer, but wouldn't the performance of a desktop system written to support java script etc. be lower than that of a regular written desktop? So, in the worst case it would slow down the whole system. Regarding speed, I don't think it's an issue.

First, all computers wait at the same speed, and presumably the point here is to accomplish something heavily dependent on the network. Even the best network (in my experience) winds up being the limiting factor.

Second, the applications are not likely to depend on the speed of the processor for much, in the user's experience. Now obviously, if we're using bloated software like Word to accomplish what notepad could do, we'll feel the hit. On the other hand, I'm consistently frustrated by the sloth of OO apps. So if FIrefox offers an equally slow solution that is better integrated, I say it's a winner.

Of course, I haven't RTFA, as it is FSD'ed.

Re:Somehow familliar

[suv4x4]

First, all computers wait at the same speed, and presumably the point here is to accomplish something heavily dependent on the network. Even the best network (in my experience) winds up being the limiting factor.

"In theory, practice and theory are the same. In practice, they're not"
Winnie The Pooh

Modern computers don't make everything "wait" for something to happen. They multitask. Even modern browsers (Opera, IE, Safari) multitask. Firefox doesn't.

For Firefox, loading of several files over the network is a Very Important Thing, and it'll just hang in mid-action waiting for the network to say something. That's pretty bad.

JavaScript has no concept of threads. Also it has no concept of security, apart from the "100% trusted" or "100% not-trusted" sandbox.

It'll be very funny to watch this project fail into obscurity, for those interested. I'm not.

Re:Somehow familliar

[Anonymous Coward]

This isn't where we are now, it's where we're all going; see the MS, Adobe, lazlo and JavaFx frameworks for the proof.

Here's a clue

Re:Somehow familliar

[suv4x4]

This isn't where we are now, it's where we're all going; see the MS, Adobe, lazlo and JavaFx frameworks for the proof.

Here's a clue

This is not where we're going. With multi-core processors on the table.

I've read the article in your "clue". For me, this clue just shows the Firefox team is clueless, and that's why I've given up any hope on it.

Lazlo is just a failed open-source version of Flex-like framework. They open-sourced it since they couldn't sell it.

Microsoft .NET and JavaFX are full-blown languages with great idea of what a thread is, and they are multi-core ready from the onset.

Adobe's AIR is just a way to publish your web work as a desktop gadget. They recognize the lack of threads as a serious problem and actually just a month ago Adobe released a new version of Flash Player 9 that has split the engine into multiple threads that takes advantage of multi-core CPU-s. Don't be surprized if the next Flash Player adds multi-thread support in the user space as well (i.e. ablity to spawn threads in ActionScript).

So this is where we are ACTUALLY going. Firefox will see their mistakes one day, but it'll be too late.

Re:Somehow familliar

[Courageous]

Lazlo is just a failed open-source version of Flex-like framework. They open-sourced it since they couldn't sell it.

You are, of course, aware that Adobe is open sourcing Flex, also?

Anyway, I like Flex.

Joe.

Re:Somehow familliar

[MrDrBob]

An application written in JavaScript has to be parsed and interpreted through a VM, which takes both CPU time and memory. Native applications (typically written in C on Linux) have no need to be parsed, interpreted, or run through a VM, and so will consume less CPU time and memory. If your window manager (a pretty important part of your desktop) is in JavaScript, things are going to render more slowly, and we all know that Firefox isn't the most efficient browser out there. (Interesting how the people who regularly spite Firefox about memory leakage don't seem to be apparent in this discussion.)

Re:Somehow familliar

[Anonymous Coward]

> For me, this clue just shows the Firefox team is clueless

So the guy who created javascript is clueless?

If things go to plan, Adobe and Mozilla will be both be using Tamarin to host the next major ECMAScript revision (Javascript 3).

> So this is where we are ACTUALLY going.

I thought the article was clear, ECMAscript will not use threads to implement coroutines.

Re:Somehow familliar

[suv4x4]

You are, of course, aware that Adobe is open sourcing Flex, also?

Anyway, I like Flex.

Joe.

Yes and incidentally they're open sourcing it for the same reason. You see, they're feeling the heat from Microsoft .NET. They (and well the people around them, the Flash community, me including) knew about Silverlight for years now (WPF/E) and what it means for Flash as a product.

I like Flex too, but it didn't have the uptake Adobe/Macromedia hoped for, so from a 5-figure server edition for big enterprises (they sold less then 200 licenses of Flex 1), it went through cheap IDE, and will now be open sourced. If they had succeeded they wouldn't so drastically change their plans.

Re:Somehow familliar

[suv4x4]

So the guy who created javascript is clueless?

If things go to plan, Adobe and Mozilla will be both be using Tamarin to host the next major ECMAScript revision (Javascript 3).


I wonder what would've happened if Adobe didn't donate their engine.

I thought the article was clear, ECMAscript will not use threads to implement coroutines.

That's just the opinion of the Firefox tech airhead. Adobe won't exactly care what he things when it implements threading in Flash. Firefox could disable the feature if they're so scared.

Re:Somehow familliar

[Anonymous Coward]

> I wonder what would've happened if Adobe didn't donate their engine.

They would have used another, at one point they were even considering Mono as a base for Mozilla2.

> That's just the opinion of the Firefox tech airhead.

Show some respect, he's "Mr tech airhead" to you ;-o

> Adobe won't exactly care what he things when it implements threading in Flash.

Adobe currently ship 2 engines, only AS3 (ECMAScript 4) runs on tamarin. Since Tamarin is set to become the reference ECMAScript implementation (perhaps one of many justifications for contributing the code to Mozilla), I'd say Adobe do care.

Re:Somehow familliar

[suv4x4]

Adobe currently ship 2 engines, only AS3 (ECMAScript 4) runs on tamarin. Since Tamarin is set to become the reference ECMAScript implementation (perhaps one of many justifications for contributing the code to Mozilla), I'd say Adobe do care.

It already is the reference implementation, of ECMAScript portable (no eval() ). There's no better. And still, it's a superset. It'll always be a superset, since Flash isn't a browser, and needs some things working differently, and some more things ECMA doesn't have.

If you mean the AS1/2 engine - it's frozen, just forget about it. It's there just for back-compat.

Re:Somehow familliar

[mad.frog]

Tamarin is currently a superset of ECMAScript3. It is planned to become an implementation of ECMAScript 4.

ES3 (== JavaScript 1.x, basically) is a subset of
ActionScript 3, which is a subset of
ES4... which isn't yet finalized but does have a reference implementation in place.

For more info on the proposed ES4 standard and implementation, go here: http://www.ecmascript-lang.org/

Re:Somehow familliar

[mad.frog]

Not necessarily. Flash (which uses a JavaScript variant, the Tamarin engine) precompiles into bytecode, which is then converted into machine code by a JIT. So it's pretty darn fast by scripting language standards. Not as fast as straight C in the general case, but fast enough for many problems.

Re:Somehow familliar

[MrDrBob]

Yes, but Firefox, its extensions and websites don't, and that's what Pyro's using.

Re:Somehow familliar

[Anonymous Coward]

IIRC, it's intended that Fx3 can cache the internal spidermonkey bytecode.

Re:Somehow familliar

[Anonymous Coward]

I would be interested to see their implementation of th "Expose" like feature using javascript... I tried to implement it a couple years ago and it was slowwww.

Re:Somehow familliar

[zootm]

Tamarin is being integrated into Firefox for later versions (it's been open-sourced by Adobe) and so Pyro will get these advantages when that work is completed. There's no good reason to avoid doing work like this now just because optimisations in the back-end are not in place yet.

Re:Somehow familliar

[bogado]

Also it's becoming quite common for the so called "native applications" to be written in interpreted languages and run throw a VM. Lots of applications, including the latests multimedia stuff is being made using python. You just have to pass the hard stuff to a native library, like gstreamer in the multimedia case, and you will have a easily hackable UI shell and the speed advantage of compiled code.

First read

[HangingChad]

right alongside desktop apps (through compositing).

At first I thought that said through composting. Guess you'd have to call that organic computing.

On a serious note....Instead, trusted Web sites and extensions are given access to the full range of interactivity and control enjoyed by native applications today.

The "trust" issue would loom very large in that statement. Provides some interesting possibilities all the same.

Re:First read

[someone300]

As much as it makes me uneasy, potentially it's not that much less secure than downloading and running those random closed source shareware apps from download.com, and probably just as secure as providing your credit card info to a trusted site over the internet.

But yeah... something about it scares me. Not sure what...

Re:First read

[QuietLagoon]

The "trust" issue would loom very large in that statement.

Given FireFox's history of security issues, I would tend to agree.

Re:First read

[Cheesey]

Instead, trusted Web sites and extensions are given access to the full range of interactivity and control enjoyed by native applications today.

Yeah, I don't like the sound of this either. Seems like a two-level trust scheme: trusted websites have access to everything.

One of the design flaws in present day GUIs (including all the X11-based GUIs for Linux) is that one malicious application can compromise the entire GUI if it can open a window. This is true even if you take the sensible step of running untrusted applications as another user: you still have to give them access to your display, so (for example) a compromised Firefox can still act as a global keylogger even if it's running as nobody. There are ways to avoid this in X11 (using Xnest for example) but these are rarely used because they don't integrate well with other applications.

Is that design flaw now being extended to include web applications loaded from a possibly compromised remote server? Written in Javascript, which has proved notoriously hard to secure? Sounds nasty. Secure sandboxing should be built into every level of both the OS and GUI design so that nothing has to be "trusted".

Re:First read

[DrSkwid]

Present day is an amusing adjective to apply to X11.

Re:First read

[nnm.one]

We will soon get XSELinux at least =), I just hope that the Ubuntu developers get off their asses and start integrating it when it's released.

Re:First read

[Cheesey]

I just included that so as to make it clear that this design flaw is not confined to Windows/OS X.

Re:First read

[VGPowerlord]

Present day is an amusing adjective to apply to X11.

Given that X.Org is the X11 reference implementation, I do believe that present day applies when talking about X11.

Re:First read

[toshok]

Put it this way - it's just as secure as using a webapp you currently use. gmail? google maps? flickr? youtube? all of these include either javascript or flash which runs things client side, and they're made that much richer from it.

If you trust gmail, you go to gmail.com and use their webapp. In a pyro-world, it might be something like "you trust gmail, you download their javascript + html app and run it." In principle it's the same. The only difference is whether it downloads the js/html from the site or loads it out of an extension directory on your drive.

bad idea

[wolfgang_spangler]

Didn't we decide this was a bad idea when MS did it? They might not have done it as completely as pyro, but that just sounds like a bad idea.

I can't of course RTFA at the moment due to the flames rolling out of the webserver it is on.

Re:bad idea

[suv4x4]

Didn't we decide this was a bad idea when MS did it?

We? Mr. Gates is this you?

Re:bad idea

[itsdapead]

Didn't we decide this was a bad idea when MS did it?

No, I think we decided that it was a bad idea because MS was doing it - and was using it as an excuse for locking Windows users into IE and its capacity for "embracing and extending" web standards.

Actually, years before that we "decided" that web apps in general were a bad idea when Sun/Oracle touted their java-based thin client/Network Computer idea. I think that died, mainly, because the question was always "yes, but will it run Word for Windows and browse IE/Netscape only websites?"

Until the pyro people have put the magic smoke back into their server, its hard to even ask the right questions, but hopefully when they say "Firefox" they imply lots of standards-compliant goodness that left the applications portable across Firefox/Gecko, Webkit/KHTML or even (roll of thunder) Internet Explorer.

If anything, the Active Desktop idea was ahead of its time - back in 1995 you ran applications on your desktop, one of which was a networked hypertext reader called a "web browser" (even dynamic/data driven websites were very much "turn based") so it was far from clear why the two needed merging. In these days of AJAX and "Web 2.0" [Drops $2 into the industry buzzword swear box] the distinction between applications and web apps is blurring, and web-based technologies (Java, AJAX, Flash, DHTML) are becoming more attractive as (potentially OS-agnostic) application platforms.

One of the problems with the X Window system is that the central idea - network transparent GUI rendering - is past its sell-by date and only minimally usable with Gnome/KDE apps (I usually end up using xnest - at which point you might as well use a java VNC client in a browser...) - the cost difference between a client device smart enough to run an X server and a "thin client" capable of doing most of the GUI-related processing locally is too minimal.

These things take a few false starts to get going - I remember computer magazine front pages back in the 80s proclaiming that "The Year of Unix on the Desktop Has Arrived" - not to mention a Phillips brochure from about 1975 saying how we would have flat-screen TVs to hang on our wall like pictures Real Soon Now.

Re:bad idea

[Aladrin]

Actually, I think we decided it was a bad idea when we figured out that running an OS, then an window manager, then an application, then an (web)OS, then a (web)window manager, then the intended (web)application is horribly inefficient, and not one bit easier to program for. In fact, it's rather constricting.

I think people are seeing how easy GUI is to do in HTML/XHTML and trying to take advantage of that, but in doing so are making it more complicated. It would be a lot smarter to make an add-on for an existing GUI system (QT or GTK or whatever, it doesn't really matter) and making it easy to program for in that way. I'm a KDE person, but I'd want it on GTK for the open-source-y-ness of it, rather than built on the proprietary QT. (Yes, it's 'free' under some circumstances... But not all. I'll take my 'freedom' with fewer strings, please.)

Re:bad idea

[Anonymous Coward]

Didn't we decide this was a bad idea when MS did it?

We? Mr. Gates is this you? Mr. Gates probably still thinks it's a good idea. Anything that hurts the competition and sets in vendor lock in is Gates's and Ballmer's definition of a good idea.

I wonder how long it will take Microsoft to patent the features in Pyro...

Re:bad idea

[Anonymous Coward]

I think people are seeing how easy GUI is to do in HTML/XHTML

"GUI" is not a proper noun, retard.

Re:bad idea

[VGPowerlord]

"GUI" is not a proper noun, retard.

GUI is an acronym. Get back under your bridge, troll.

Good idea

[Anonymous Coward]

Tamarin will make XULRunner a viable platform for app development, extending this to a desktop environment is an interesting and logical idea.

The apps themselves become relatively lightweight (javascript, XML and CSS) while the runtime does all the work. It's not so different from the fabled Java desktop but will probably be easier to develop for and it appears to be the way application development is going anyway.

They should move to Fx3 pronto and make a demo ISO based on DSL or something.

Haven't we done this before?

[christurkel]

Wasn't this done with MS in Windows 98, the Active Desktop? See how well that worked? Why would anyone want this?

Re:Haven't we done this before?

[aminorex]

On this interpretation, we should never use artificial intelligence because of Clippy.

Re:Haven't we done this before?

[suv4x4]

Wasn't this done with MS in Windows 98, the Active Desktop? See how well that worked? Why would anyone want this?

That was done in 1998. It was early Web 1.0, and people didn't dig web stuff so much. But now, it's different. There are plenty of uses for a web based desktop, and to quote their site:

Internal Server Error

The server encountered an internal error or misconfiguration and was unable to complete your request.

Please contact the server administrator, webmaster@pyrodesktop.org and inform them of the time the error occurred, and anything you might have done that may have caused the error.

More information about this error may be available in the server error log.

Additionally, a 500 Internal Server Error error was encountered while trying to use an ErrorDocument to handle the request.

I think Microsoft is totally shaking in their boots at the thought of Pyro: just consider, a connected, integrated, web desktop. It's just like .NET 3.0 except it's much slower, much less secure and runs on JavaScript. Complete winner!

Re:Haven't we done this before?

[gbjbaanb]

I don't know, but (the article is /.ted) but it sounds more like MS's HTAs - their web-as-a-desktop-app system. (HTA = Hypertext Application). They used IE as a client front end to a local (or remote) web application. It was hosted in IE but without titlebar, buttons etc. It also ran in an increased security environment (as you'd expect a desktop app to interact with the filesystem, for example, that normal web apps hosted in a 'normal' browser would not get acess to).

It semed like a good idea, and enabled you to write desktop/web applications, but it never quite caught on, MS moved to Jav, sorry .NET, and lost interest in it.

Active Desktop was just a way of putting content on your dekstop instead of a static image. I think it was a little before its time due to everyone using dial-up modems instead of always-on broadband. If someone did it today, we'd have the advantage of a lot of experience in using web applications, faster networking and better security. Imagine what it could be in an Ubuntu environment, it could easily be a new desktop paradigm that makes Window's taskbar-based system look old and boring.

Re:Haven't we done this before?

[AmberBlackCat]

I don't see how Clippy counts as artificial intelligence. It wasn't learning anything. It was just responding to things. If it could learn and adapt, it probably would have been good.

Re:Haven't we done this before?

[Frumious Wombat]

That's true, because if Clippy was true AI, he'd learn from you and become more helpful. OTOH, he'd also have instructions from MS hard-coded in for features that are detrimental to you, but good for their stock price. The end result would be self-confident schizo app (c.f. HAL 9000), who would probably do Bad Things to your documents at an inopportune time.

Convergence

[Anonymous Coward]

Wow, this is a pretty cool concept. I think there are some security questions that need to be answered, but it looks like the right direction for driving the web/desktop model to new heights.

I was just reading a similar blog post about the transition of web2.0 to web3.0 and that guy was saying that a similar thing needed to happen to power up the apps.

That is an interesting read too. http://creduware.com/blog/2007/07/21/when-will-web -20-die/

Security Nightmare

[Anonymous Coward]

This sounds like a security nightmare. Think Internet Explorer bad.

From their web site: trusted Web sites and extensions are given access to the full range of interactivity and control enjoyed by native applications today.

Thanks, but not for me.

Symphony OS Anyone?

[asphaltjesus]

This guy is probably doing something similar. Symphony OS.

http://www.thecodingstudio.com/opensource/linux/sc reenshots/index.php?linux_distribution_sm=Symphony %20OS%202006-12%20Beta

Re:Symphony OS Anyone?

[Anonymous Coward]

http://www.symphonyos.com/cms/
YEP.

See also ByzantineOS (Mozilla Desktop)

[Anonymous Coward]

This idea has been around for a little while. See the ByzantineOS has had releases on Distrowatch since 2003.

It was a pretty cool idea. Basically the whole desktop is a web browser and you write apps as XUL extensions. There is possibly a future in this as corporate and institutional thin client platforms to run custom apps.

Bad naming.

[mxf8bv]

This is already the third project called pyro.
"Python Remote Objects" http://pyro.sourceforge.net/
and "Python Robotics" http://pyrorobotics.org/
and possibly there are even more.

Re:Bad naming.

[misleb]

And http://www.karppinen.fi/pyro/

Lack of network: Failure?

[theolein]

In order to head off claims that this environment would be useless if network connections were down, I should point out, that in modern business and private settings, having no network is almost as bad. In other words, being dependent on the network is probably not a liability.

Re:Lack of network: Failure?

[realdodgeman]

Still it wouldn't hurt to have Fluxbox with a few offline apps as backup anyway.

Re:Lack of network: Failure?

[MrDrBob]

People's applications should be able to run fine without network — that's what we've got NetworkManager for. Saying that people can't work without network and then using this as an excuse to depend completely on a network connection's a bad idea. I do hope this is not what you mean.

Re:Lack of network: Failure?

[ScrewMaster]

True ... I mean, when my cable connection is down I do feel as if someone removed an important organ from my body ... but I can still load OpenOffice and get some work done.

Site is already slashdotted :D

[hebertrich]

Internal Server Error

The server encountered an internal error or misconfiguration and was unable to complete your request.

Please contact the server administrator, webmaster@pyrodesktop.org and inform them of the time the error occurred, and anything you might have done that may have caused the error.

More information about this error may be available in the server error log.

ROFL

Google Cache

[Eddi3]

Here's the Google Cache:

http://72.14.253.104/search?q=cache:6EoAZGSE90IJ:p yrodesktop.org/+http://pyrodesktop.org&hl=en&ct=cl nk&cd=1&gl=us

Why this is cool

[Anonymous Coward]

It's more about that they're using popular languages that people know (HTML, CSS, Javascript) to do things that traditionally were done by unpopular languages (C, Python, etc.)

Now before you flame me about calling those languages unpopular all I mean is that not as many programmers know them, I don't mean to insult anyones language.

So Pyro (although I don't think it'll go anywhere) has some good ideas about simplifying and opening up the linux desktop to more developers, and that's something that should be applauded.

Re:Why this is cool

[Haeleth]

It's more about that they're using popular languages that people know (HTML, CSS, Javascript) to do things that traditionally were done by unpopular languages (C, Python, etc.)

Now before you flame me about calling those languages unpopular all I mean is that not as many programmers know them

Um, but surely people who know HTML and CSS but not a conventional programming language like C or Python are not programmers - they are web site developers. (Now, before you flame me, I'm not saying that they're inferior, I'm just saying they're different.)

The people who write the code that drives web apps are programmers, but they all know languages like Ruby, PHP, Java, and so forth. That's what they use to write web apps in the first place. You can't use AJAX if there's no server for you to send your AJAX requests to, and the code running on that server is not written in HTML. It could theoretically be written in Javascript, of course, but that's pretty rare, if it happens at all. Why? Probably because more programmers know languages like PHP and Python than know Javascript...

Basically, I do not believe there is any significant number of people who are capable of using Javascript to do more than copy and paste trivial scripts, but do not know any other programming language. Yes, Javascript can be used to write complex applications; but my point is that the people who are capable of doing that are also capable of using conventional tools for the same purpose, so while this project might indeed enable some interesting new applications, it is unlikely that there will be many people capable of developing with it who would not be equally capable of developing without it.

Re:Why this is cool

[jlarocco]

It's more about that they're using popular languages that people know (HTML, CSS, Javascript) to do things that traditionally were done by unpopular languages (C, Python, etc.)

Now before you flame me about calling those languages unpopular all I mean is that not as many programmers know them, I don't mean to insult anyones language.

The fact that you called C and "unpopular language", no matter what you meant by it, completely invalidates any opinion you may have on any programming related topic.

Re:SUE!!!!!

[Anonymous Coward]

Microsoft will likely patent this, innovate it into Vista+1 (7), and then use those patents to threaten the Pyro project for coming up with the ideas in the first place.

Coral got it

[Anonymous Coward]

Yay for coral!

Internet Explorer

[springbox]

Huh? What's this? For Linux users who wish they were running Windows?

Informative parent

[Cheesey]

I am glad to hear that someone is working on a real solution to this problem, because many people do not even seem to be aware of it. XSELinux sounds like exactly what is needed. More information here.

This site has ceased to be!

[Qbertino]

It's not pinin'! It's passed on! This website is no more! It has ceased to be! It's expired and gone to meet 'is maker! It's a stiff! Bereft of life, it rests in peace! If you hadn't posten it on slashdot, it wouldn't be pushing up the daisies! Its metabolic processes are now 'istory! Its off the twig! Its kicked the bucket,shuffled off 'is mortal coil, run down the curtain and joined the bleedin' choir invisibile!! THIS IS AN EX-WEBSITE!!

Very Old Idea:

[Anonymous Coward]

Does anyone beyond me remember Workspot?

mmm

[MrShaggy]

Didn't Microsoft try this with XP and IE??

Look what happened to them..

I will believe it when I see it.

Re:mmm

[ddrichardson]

Still the dominant mainstream operating system and web browser despite its failings and many better products - that should cheer up Pyros developers.

Heh.

[stonecypher]

So, Active Desktop is making a comeback, is it?

(c) Monty Python, the dead parrot sketch

[cheros]

And the website isn't down, it's merely resting. :-)

the desktop future

[m1h41]

oh my god it's full of ads!

Re:the desktop future

[bogado]

Guess you never heard of the adblock extension, have you?

ARG! POPUPS!

[Ungrounded Lightning]

Agree with your comments.

The whole thing sounds to me like running web applications as popups that are exactly like the locally run application windows, both in appearance and in all, or most, of the infrastructure behind them.

I can't think of a better way to assist malicious web apps in fooling the user by removing any clue that they're not something local and innocuous.

A phisher's dream.

Per your request.

[Ungrounded Lightning]

If you hadn't posten it on slashdot ...

[My english is better than most other people's german, so please point out mistakes politely. Thank you.]

"posted"

(Unless there's some variant I'm unaware of, perhaps in UK English or a reference to a comedy routine, where "posten" would be proper or humorous in that sentence.)

Your use of English idiom is excelent.

"mashup?"

[Anonymous Coward]

What the fuck is a "mashup" and why is it good?

not as many programmers know them?

[r00t]

To really "know" your favorite languages (HTML, CSS, Javascript) is quite rare. Lots of us can get some minor thing half-way working on one browser. Writing solid code is not at all a trivial task; very few people can manage. Not even Google always gets it right.

While I've seen some very buggy and unportable C code, fixing one platform doesn't tend to break all the others as is the case with the web stuff.

Re:not as many programmers know them?

[Ansoni-San]

To really "know" your favorite languages (HTML, CSS, Javascript) is quite rare. Lots of us can get some minor thing half-way working on one browser. Writing solid code is not at all a trivial task; very few people can manage. Not even Google always gets it right. Knowing HTML, CSS, and Javascript has nothing to do with knowing exactly the way they're implemented on every browser under the sun. Writing solid code IS trivial but it's not all going to work in any browser. The typical process is writing solid code missing about 20% of the useful features the language provides and then wrestling with it to get it working in all browsers, kind of like stretching skin over a drum. And then breaking it and putting it back together with gum and spit, and giving this version to IE. The resulting code which works in all browsers is not "solid code"....it's superfluous backwards code that happens to please all the browsers you have tested in.

I just wanted to point out that knowing the languages has nothing to do with having it work as intended in even a single browser. I don't think you can say code isn't solid just because a browser doesn't understand the language the code is written in.

Re:not as many programmers know them?

[toshok]

If anything, this project might represent a breath of fresh air when it comes to the problem of platform differences between javascript/layout engines - you only have to worry about one. Granted, there might be bugs between different versions of said engine, but that's a problem that already exists in the native code space.

Memory usage

[Anonymous Coward]

Firefox is notorious for its memory usage. It has to be restarted a lot so the memory usage can be controlled. Do we really want this for the whole desktop? Login-logout cycles every hour so your box won't swap because your bling-bling desktop system is taking up all of the physical memory?

Desktop environment built on bugs?

[DogDude]

Until Firefox can solve some of the many, many bugs still present in it's product, I think that this entire discussion is silly.

Re:Desktop environment built on bugs?

[Saurian_Overlord]

Why? Microsoft has been building desktop environments on buggy web browsers for years, and they seem to be doing pretty well for themselves.

Re:Desktop environment built on bugs?

[misleb]

The bugs in Firefox are not nearly as significant as the shortcomings of XUL. XUL is definitly better for programming "rich" applications than HTML, but once you start using it you realize that the Mozilla team only developed XUL just enough to get the browser up and running. Last time I used XUL there were big parts that just didn't get any attention. And even if all the widgets did function as they were meant to, they'd still be a more limiting than common desktop toolkits. And then there's Javascript, which also shows its limitations when an app starts to get rather complex. Javascript, despite being object oriented on the surface, doesn't even have a proper class inheritance system.

-matthew

Re:Desktop environment built on bugs?

[Anonymous Coward]

of course it doesn't have class inheritance. it has prototype inheritance.

/. effect

[Mazin07]

Obviously their webserver is running as a web app from within Firefox, or else it would be able to serve more than 5 simultaneous clients at once.

Re:/. effect

[timbury]

"it would be able to serve more than 5 simultaneous clients at once." As opposed to 5 simultaneous clients at different times? ;-)

Re:/. effect

[Mazin07]

Maybe it'll serve the clients sequentially?

*grumbles about slashdot language nazis...

Re:sig

[Alan Doherty]

"There are 11 types of people in the world, those who know binaries and those who don't."

i realy hope ths was typo as the quote/joke is wrong in so many ways

the correct one is know/uderstand are interchangeable but understand is the common one
There are 10 types of people in the world, those who know/uderstand binary and those who don't.

as 11 == 3 thus makes the joke nonsense
and binaries equally makes nonsense of the joke

Slow

[Anonymous Coward]

Usability and practicality aside... wouldn't scripting your OS in Java just be too slow?

Re:Slow

[Anonymous Coward]

I would think so. Good thing it's not Java.

Re:slashdotted after the first comment -- Back now

[orphennui]

It's back up now. Sigh...

My personal project hosting provider, DreamHost, actually *moved* the main directory on pyrodesktop.org due to too much server load. I'm using the totally bundled MediaWiki version that DreamHost provides, mind you. I was on a train to London, and just realized something was up now.

So I've just moved the wiki directory back. Let's see how long it stays up...

GNOME - got a toolkit make a destop from it.

[krischik]

Ohh - please no. We had this allready. No, not Active Desktop. I mean GNOME. Got a tookkit for a graphic processor (GIMP) and then think: Why not make hole desktop from it. Stupid idea - just because it works for a graphic processor or bowser it does not mean it will scale up to desktop environemnt. Even after years GNOME is not as stable as KDE. Don't believe me? How about:

Unable to start the settings manager 'gnome-settings-daemon'.
Without the GNOME settings manager running, some preferences may not take effect. This could indicate a problem with Bonobo, or a non-GNOME (e.g. KDE) settings manager may already be active and conflicting with the GNOME settings manager.

There is a damm lot more to a desktop environment then pretty looks. It look easy at the beginning and sounds like a great idea but becomes quite complex over time. And in the end you have a mess like GNOME. And it is a mess: gtk, cairo, pango, bonobo and no common line.

Martin

Re:GNOME - got a toolkit make a destop from it.

[MrDrBob]

You sir, have no idea of what you're talking about. Quoting an error message with no context means nothing. Do you even know what GTK, Pango, Cairo and Bonobo are? What's a "common line"?

Re:GNOME - got a toolkit make a destop from it.

[krischik]

You sir, have no idea of what you're talking about. Quoting an error message with no context means nothing. It was not meant to be a bug report - it was meant as an example for a bug which appeart on my system and won't go away. Actulay it also appeared once on my solaris system. Solution: delete all setup directories.

Bottom Line: The setup system of GNOME is not stable. Smallest mistake in your files and BANG!

Do you even know what GTK, Pango, Cairo and Bonobo are? What's a "common line"? A single download side with a single directory per version would be a good start. And yes I know them: From time to time I need compile Gtk+ from scratch and it is hell you get a set of fitting version together from all the different sites. And this is just Gtk - I don't even want to know what trouble one is in when one wants to create a hole GNOME.

On the other hand: If KDE anounces a new release I go to kde.org and everthing is there. Sourcecode or binaries and a unified subversion archive. KDE is just one project with a common line and not a humble jumble of dozend half independent projects.

Martin

Re:GNOME - got a toolkit make a destop from it.

[saned]

If you want to build GNOME from source, try http://www.gnome.org/projects/garnome/ HTH

Re:GNOME - got a toolkit make a destop from it.

[MrDrBob]

If you make a mistake in any configuration file on Linux, things aren't going to work properly. That's the way of things, and there's no simple fix. It's just the same for KDE. You might be looking for the GNOME FTP site (ftp://ftp.gnome.org/Public/gnome/sources/). It doesn't include Cairo, since that's a separate project. When compiling, the configuration script should check that you've got the minimum required versions of all dependencies, which should stop compilation from failing. GTK+ (etc.) will compile fine with any version later of a dependency than its minimum required version, so you don't really need to get a "set of fitting version together". Regardless, you should be using your distribution's package management system, since it'll likely have distribution-specific changes. You just go to kde.org; I just go to http://www.gnome.org/start/2.18/ (linked to from gnome.org).

Re:GNOME - got a toolkit make a destop from it.

[krischik]

If you make a mistake in any configuration file on Linux, things aren't going to work properly. That's the way of things, and there's no simple fix. But just using the standart gui (gnome-control-center) I should not be able to mess up my configuration - If I am then gnome-control-center has not sufficently protected me. Messing up should only be possible with an text editor or gconf which I did not use.

It doesn't include Cairo, since that's a separate project. Which is what I critisise: To many separate projects.

Martin

Re:GNOME - got a toolkit make a destop from it.

[MrDrBob]

But just using the standart gui (gnome-control-center) I should not be able to mess up my configuration - If I am then gnome-control-center has not sufficently protected me. Messing up should only be possible with an text editor or gconf which I did not use. There's not much that can go wrong with control centre. If you tell me what you did, I'll see about fixing it.

Which is what I critisise: To many separate projects. So everyone should implement their own drawing library, instead of linking to and using a standard one which is stable and efficient, and requires no duplication of effort?

Re:GNOME - got a toolkit make a destop from it.

[toshok]

I'm going to go out on a limb and suggest that Alex and I have more experience developing desktop systems than you do. So I thank you kindly for your suggestion that it's not easy, but I'm pretty sure we already know that.

Also, your thoughts on gnome vs. kde are irrelevant to this discussion. Pyro exists because we believe *neither* of them represents the future.

Re:GNOME - got a toolkit make a destop from it.

[SanityInAnarchy]

But just using the standart gui (gnome-control-center) I should not be able to mess up my configuration - If I am then gnome-control-center has not sufficently protected me.

First, spelling check. It's "standard", with a d.

Second, are you absolutely sure there's no way it could have messed up unless it was gnome-control-center? Really? For example: bad RAM, bad disk, etc could all be at fault here. I'm not saying it couldn't possibly be gnome-control-center, but understand that it can't possibly protect you from anything you could possibly do -- and, in fact, a powerful tool should make it possible to hose your system if you were really trying to. (For exmaple, gconf.)

Third, you should have had backups, anyway. It's not difficult -- back up /home and /etc and you can rebuild just about your entire desktop from scratch, or carry it from distro to distro.

Which is what I critisise: To many separate projects.

There's no getting around this. Or rather, none of the ways around it are worth it.

You could force everything to be completely integrated, so that installing a single app requires pulling in tons of dependencies, many of them completely unrelated. Maybe it's just me, but this seems to happen a LOT more with KDE apps than with GTK ones -- maybe because they're typically KDE apps, not QT apps, whereas the other side is usually GTK apps, not GNOME apps.

That's probably what you meant -- KDE does have a pretty bad case of not-invented-here syndrome, and it doesn't seem to be using Cairo, for example. But GNOME and KDE both use dbus and X11, for example. Would you rather that these not be separate projects?

Or, you could force every app to rewrite all the functionality by itself, because you don't like having a separate project for a shared library. Obviously, this is a retarded idea, and you're a retarded person if that's what you were suggesting.

Re:GNOME - got a toolkit make a destop from it.

[SanityInAnarchy]

I have to wonder -- why, exactly, do you think that your gnome-settings-daemon is at all related to GTK+?

That's right, it's not.

I hate GNOME as much as the next guy, but let's be fair here -- it's not that GNOME sucks because of GTK+, and really makes no sense to suggest that a graphical toolkit has anything to do with the rest of the desktop -- it doesn't. All GTK+ does is the graphics.

Kind of like -- oh, I don't know -- QT. Which is also used for KDE's graphical tools (Krita).

I mean, really, what you're saying is kind of like saying Linux is better than Windows because Linux doesn't have a start button.

Re:GNOME - got a toolkit make a destop from it.

[SanityInAnarchy]

This is the third post of yours that I'm replying to. You must be either 13 years old or intentionally playing dumb -- the spelling errors are making it hard to read. It's "whole" when you mean complete, not "hole". In fact, "hole", as in security hole, is the opposite of "whole", as in complete.

A single download side with a single directory per version would be a good start.

Great, so when there's some tiny security patch in some library, you have to go re-download the entire fucking library. But hey, at least you know you've now got KDE version 3.5.6 build 135812903510912385709123, and not build 135812903510912385709122 or 135812903510912385709124.

Fortunately, the KDE developers aren't as stupid as you are. Trying to install the "kde" package suggests an additional 158 packages on Ubuntu -- and I already have kubuntu-desktop installed.

On the other hand: If KDE anounces a new release I go to kde.org and everthing is there. Sourcecode or binaries and a unified subversion archive.

Watch this:

sudo apt-get install gnome-desktop-environment

Need source?

apt-get source gnome-desktop-environment

Was that so hard? Opening it up in Aptitude, I find the package is called 2.14.3.3ubuntu1. Looks like a pretty specific version to me -- and yes, a "common line".

As an added bonus, since I'm letting package management handle it, instead of trying to download all the sources by myself, I can actually get little, incremental updates without having to download every single KDE app again.

KDE is just one project with a common line and not a humble jumble of dozend half independent projects.

Oh, bullshit. How do you even define that in an open source world? You can't.

I challenge you to find one consistent definition that proves GNOME is a mess of independent projects without also proving the same of KDE.

There are advantages to KDE. They do involve integration. But I seriously doubt it has anything to do with how the projects are organized or distributed. If anything, it's because there are parts of KDE that were designed from the ground up to make it very easy to mix and match components to build your app -- for example, Kontact integrates KMail, KOrganizer, Akregator, KAddressBook, KNotes, KArm and MultiSynk into one app -- or you mix and match whichever ones you want. Konqueror will load and run more than one kind of document viewer -- including odd ones like KPDF -- right there in the browser.

But, I'd almost argue that the only "integration" was creating a system like KParts and providing a forum for developers of these different projects to meet. It's precisely because the applications involved have so little to do with each other that they're so light and fast. For the counterexample, just compare Kontact to Evolution or Outlook. Evolution and Outlook are both hugely bloated, slow, unreliable, unmaintainable apps. Probably insecure, too. Kontact does basically the same stuff, but it's nice and modular -- and really, running Kmail in Kontact, for example, isn't any slower than running, say, Thunderbird or Kmail by itself.

(And I'm assuming you mean "KDE is just one project with a common line and not a jumble of a half dozen independent projects." I'm amazed what you actually typed is as readable as it is.)

Discussion on someone's blog

[Anonymous Coward]

There's been some vocal discussion about Pyro on someone's blog (http://squeedlyspooch.com/blog/archives/002095.ht ml), which is probably relevant to this article.

Download and try .. still up and running.

[Anonymous Coward]

http://pyrodesktop.org/releases/compzilla-0.1.xpi

the wiki part is /. but the install / download still works.

Hmm...

[JustNiz]

This just sounds like its follwing Microsoft's trend of adding even more layers of middleware to suck down all our CPU time and resources just to do the simplest of tasks.

Version stabilty

[krischik]

Got another one: I a desktop evironment you can's just declare plugin "incompatible" from one version to the next - the way firefox does not. A desktop evironment just jsut have to keep compatible with your past.

This is the most stupid idea I heart this year!

Martin

Re:sig

[Anonymous Coward]

Whoosh

User interactivity?

[macraig]

Given that Firefox already has issues with ignoring user input at various times, I guess Pyro will also bring that ability to ignore the user to the Linux desktop, as it has existed in Windows since the beginning?

Re:sig

[Anonymous Coward]

Are you dense?!

Thats the whole point of the joke

he obviously _doesnt_ understand binary.

Its a variation. Thats why its funny.
I swear you cant tell a joke to a geek.

Re:sig

[TheRaven64]

It's the amalgamation of two (old) jokes:

There are only three sorts of mathematicians [sometimes people], those who can count, and those who can't. And:

There are 10 sorts of people, those who understand binary, and those who don't. And, yes, combining them does make you sound like someone who's heard the second and incorrectly remembered it.

Great, I'll go buy another 2G RAM

[Lalo Martins]

I thought about something like this recently, and decided against it, because Firefox is very memory-inefficient, and seems to (I may be wrong) leak more than Harry Potter spoilers. So while it sounds interesting, I wonder if I could really run it, in practice, without having to log out every day or so to release memory.

Re:Great, I'll go buy another 2G RAM

[toshok]

(I may be wrong)

maybe you should make sure you're right before posting? And I doubt you thought of doing it the way we're doing it :) If you're already running firefox for a browser, and choose not to run anything but native apps (which is likely the case now, since we haven't really explored how to even do the js+html apps), you shouldn't see much difference in memory usage. Our code is pretty small, and we've been careful about client-side memory usage. Much of the firefox memory usage complaints center around objects that live on past page loads. There aren't any page loads in pyro (other than loading start.xul, anyway).

A desktop manager written in Javascript?

[Anonymous Coward]

It must run fast!

Re:sig

[wish]

There are 11 types of people in the world, those who can count in binary, and those who can't.

Linux projects bundling or supporting update?

[sysadmintech]

Are any projects going to bundle or supporting download of pyro desktop? I wait until I here one does that way it will run fine with the project. Anyone who has been using python had to know this was coming from other similar type current projects.
MS can only keep program developers out of work for so long.

Cool

[Anonymous Coward]

I already imagine debates about Firefox vs. Opera, where Opera fans will say Firefox is a good desktop environment, all it needs is a good web browser...

Re:Cool

[VGPowerlord]

No, I think that a good desktop environment needs to be multi-threaded. Heck, single-threaded apps are bad enough.

(I also replied to this because of my signature)

He must be wrong, not me, I know everything.

[DrSkwid]

You don't know binaries as well as you think you do.
All the other responders are also wrong.

Re:He must be wrong, not me, I know everything.

[Ignominious Cow Herd]

I know binaries. I've worked with binaries. You sir, are no binaries.

See also "Try and Try again" (Mozilla Desktop)

[Anonymous Coward]

"This idea has been around for a little while."

Several ,several whiles Even Flash wants in on the act.

Eich != airhead

[mad.frog]

Firefox tech airhead

Brendan Eich is not only the original inventor of JavaScript, he's one of the smartest guys I've ever met. Calling him an "airhead" is really pretty sad.

More to the point, he's correct: threads suck, in that mere mortal programmers have pretty much no chance of getting them right in the general case, and even superstars have problems with them. We need a better programming paradigm for taking advantage of multicore, but threads are NOT the right approach for a scripting-level language like ECMAScript.

Re:Eich != airhead

[VGPowerlord]

We need a better programming paradigm for taking advantage of multicore.

Such as?

I mean, threads are already independantly executed elements that have a shared memory space. What do propose instead, forked processes that created a shared memory location? Wouldn't forking a GUI process not only waste a ton of memory (as it creates a copy of the entire process's memory space), but also make it duplicate its windows?

A single-threaded GUI application is the digital equivalent of having a call center with only one person working in it. They can only answer one phone at a time, putting everyone else on hold (or just ignoring them) while they do so.

I can't think of a really good analogy for the single-GUI thread, multiple worker threads scenario (which is how KDE and Microsoft Windows appear to work).

The point of this is that javascript can still operate in a single thread, it'd just have to communicate with the GUI thread to have the GUI updated. As someone else mentioned previously, all the other major web browsers do Javascript in a separate thread.

Re:Eich != airhead

[suv4x4]

Brendan Eich is not only the original inventor of JavaScript, he's one of the smartest guys I've ever met. Calling him an "airhead" is really pretty sad.

Being smart != being right.

While JS is a nice scripting language, the choice of basing an entire platform on it is horrible and we see the disadvantages in every single XUL app out there.

Threads aren't nice, soft and warm from a programming point of view: yes, they are hard and error prone, if used poorly. No one sais decent applications should be easy to do. XUL is just "making it nice" from programmer's point of view while abandoning the point of view of the user.

While I'm all for transactional memory or whatever your threads-workaround favorite is, the matter of fact is this is still being researched and worked on, and threads are used in practice to deliver.

I'm not saying put threads in the userspace (i.e. web pages than run in 10 threads), but put it in the engine for XUL, so Firefox can perform decently. It's not that much to ask for a browser that claims to be the best in the world.

Re:Eich != airhead

[mad.frog]

I'm not saying put threads in the userspace (i.e. web pages than run in 10 threads), but put it in the engine for XUL

That's basically what he's saying: threads may have a place in the underlying runtime, but not in the toplevel userspace scripting language.

great

[botkiller]

Wonderful, now my desktop can be a memory eating hog that hasn't worked right sine v. 0.8 AND crash all the time. Awesome.

Re:great

[martin_henry]

stability of Firefox
mod parent Funny

a new wave

[kbox]

In Pyro, Web apps run in windows on the desktop, right alongside desktop apps

revolutionary! Watch out MS! :/

I don't know about anyone else, But the main reason i use FF is because i don't want my browser to be an integral part of the OS.

They always say, A feature that isn't a solution to a problem, is a problem in it's self.

finally

[elgatozorbas]

Pyros for porn

Use the hammer!

[Brandybuck]

Use the hammer! When you need to insert a screw use a hammer to bash it in! When you need to saw some wood, use a hammer to break it apart! When you need a pair of tweezers to carefully manipulate a tiny electronic part, use a hammer to bash it into smithereens!

When you're so tunnel blind that all you can see is the web, then everything starts looking like a web page.

well that's rich

[10am-bedtime]

lots of "rich" stuff there; sounds fattening. i see "incredible" "leveraging", too. the clincher, however, is "trusted". now we can rest easy.

bingo!

Ridiculous!

[soupforare]

We already have emacs, why would anyone switch from that?

Re:Ridiculous!

[DaveV1.0]

EMACS?!? You heathen! Everyone knows VI is superior to EMACS in every way except bloat!

Quickly, we must marshal the troops for a return the holy text editor wars!

Not so fast!

[Gazzonyx]

I'd agree with you 95% - the only thing that Office has, that I can't readily find without setting up a server or datamining app, is pivot tables. I don't like Microsoft Office in the least bit, but I'll give credit where credit is due. Pivot tables make my job so much easier since I'm stuck with a legacy Access database solution and can't get my boss to let me migrate to SQL (our front end uses DAO, we'd have to wrap calls in ODBC, JDBC, or ADO.NET - I'm pushing for anything but the latter). There is just no other way to visualize data right from the queries, in real time, so easily and with so much versatility.

It also has a bit to do with the structure of our tables and queries. Pulling related data and referencing from inventory and invoices and PO's would be a hassle unless we set up a real client/server paradigm (sorry, I think I was forced to read a white paper recently, I never say 'paradigm') and used a data mining app (I would use Oracle's whole stack just for their datamining and datawarehousing that JBoss just doesn't have) to translate all the data.

Other than that, yeah, you can't use Office for reporting of any useful kind. I just happen to be in a crappy, legacy position where Excel just happens to be useful, for lack of options.

Expose in Javascript? Damn!

[aschoeff]

Am I the only one to be mightily impressed at the idea of implementing Expose in javascript??

I haven't looked yet at how well they accomplished this, but damn, I love the idea of having a common Expose-like function available to me on all the modern OS's I am forced to use daily.

If it's GPL'ed I will check it out.

great

[senatorpjt]

A combination of the speed and stability of Firefox with the app and hardware support of Linux.

It's all coming true...

[Wikipedia]

Microsoft's Attempt to Dissuade Netscape from Developing Navigator as a Platform: http://www.albion.com/microsoft/findings-21.html

Re:Expose in Javascript? Damn!

[toshok]

the pyro expose code is adapted somewhat from compiz's (it uses compiz's code to compute the destination x/y/w/h of the windows.)

the compiz code is MIT licensed, and the pyro adaptation is MPL'ed.

Don't use Firefox for it...

[SanityInAnarchy]

No one complains about Konqueror's kind of similar integration...

But anyway, I think this kind of concept could be done well. Just two obstacles:

No web browser that's advanced enough to run AJAX apps is simple enough to be as secure and stable as we need. It's not that we couldn't write such a browser, it just hasn't been done.

HTML/CSS/JavaScript will always suck, performance-wise. We really should create a better standard GUI toolkit for the web, if we're intending to run applications on it.

uh oh

[PermanentMarker]

we've seen this before in windows 98, and many people where against it.
Microsoft had to remove this GUI or change so that win98 could also run with other web brouwsers...
Just wondering how it will end up in linux

Correction

[SanityInAnarchy]

Great, so when there's some tiny security patch in some library, you have to go re-download the entire fucking KDE libraries.

Fixed it here. Sorry about that.

Great idea. Basically, like google-gears?

[walterbyrd]

http://gears.google.com/

BTW: I don't know why so many posters are saying that msft tried to do this in 1998, or whatever. Msft has never had anything like this. Pyro would not make FF and integrated part of the desktop. And msft is always 100% proprietary. With msft it's all about vendor-lock-in, Pyro would do just the opposite.

I can see a lot of advantages to browser-based apps. The apps can run on any platform. The apps can be server or client based. The apps can run locally, or a local intranet server, or a remote internet server.

Explorer?

[Anonymous Coward]

So in other words, Active Desktop and explorer?

Re:sig

[JuanCarlosII]

There are 10 types of people in the world, those who understand binary, those who don't, and those who understand that this is a joke about ternary.

Re:sig

[Daychilde]

...those who think it's about ternary, and those who know it's a joke about quaternary? ;-)

Re:sig

[duguk]

There are two types of people in the world. Those who put people into one of two groups, and those who don't. I'm one of the one's who don't.

Tried It

[Russianspi]

Wow, lots of talk, but who's tried it? I just ran it on a test machine - an older laptop that was humming along running Ubuntu Feisty. The machine was a tad slow before running it, but it has crawled nearly to a halt upon trying out pyro. My older (read: slow) hardware may exaggerate the slowdown, but it is ridiculous. Also, my desktop is showing at about 100x its normal size, which means that I am lucky to see one of my launchers. Thought (regardless of the security problems) that I would give this a shot, but my short test has come to an end. Next time I see some FUD about this, I'll look the other way.

Re:Tried It

[Russianspi]

After a restart: "Ahhh. There's no place like gnome!"