Next Generation X11

Rene Rebe writes "The German News site Golem is running a report (babelfish translation) about the next generation X11 projects, like the OpenGL X-Server Xgl, Luminocity as well as Enlightenment 17. The report is including many screenshots and five videos."

Comments

Why isn't this already out?

[AKAImBatman]

1995: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
1998: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
2000: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
2005: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!

Nope, never heard these promises before...

Joking aside, I didn't see anything in the photos or videos that's revolutionary. Enlightenment looks like its usual "prrreeeeetttyyy" self, and X11 is shown with various transparency and warping effects that have been available on other platforms but have been largely unused.

The question of "Why have they gone unused?" seems to be pretty well answered by some of these videos. i.e. None of the applications seem to do much of anything different than current applications do. The only difference is that they have a "cool" interface. All I can say to that is, Kai's Power Tools had a "cool" interface as well. Didn't get them (or hundreds of other "me too!" programs) very far.

The truely interesting projects I've seen lately are:

1. Sun's Looking Glass Project. While it's not revolutionary in of itself, it is an excellent evolutionary step in user interface improvements. Sun really took the right path by keeping with existing Desktop designs, but improving on existing concepts like sticky notes and window shading (the ability to "fold up" a window). They've also left the door wide open for developers to leverage the new desktop for new UI concepts that fully utilize the 3D abilities of the system.

2. There was an "Ask Slashdot" a few days ago with a guy who was working on the mother of all touchpads. It was literally more of an interactive tactical plot that could have amazing uses in collaberative work.

Re:Why isn't this already out?

[Skraut]

Yeah wasn't there a Y windows in the works at one point.

While at times I've been firmly in the "There needs to be an X with less crap in it" camp, I've learned to really appreciate the network transparency. Though I do still wish it was a choice, something that could easilly be plugged in, or removed depending on the system install. The Linux Kernel is so flexable in how you can customize it for the hardware situation, its a shame you can't do the same thing for X.

Re:Why isn't this already out?

[AKAImBatman]

Yeah wasn't there a Y windows in the works at one point.

You mean this?
Fresco was the other big contender.

The Linux Kernel is so flexable in how you can customize it for the hardware situation, its a shame you can't do the same thing for X.

Modern X actually does let you plug-in, plug-out all kinds of useless^H useful crap. It has actually matured into a fairly decent system, network transparency or no.

Where this article falls flat is on trying to make us believe that any of what we're seeing is "new". We've seen it all before, just with fewer cheap effects (e.g. the "wobbly windows").

Re:Why isn't this already out?

[indifferent children]

You are probably referring to the perennial "Why Windows?" campaign.

Re:Why isn't this already out?

[StormReaver]

Perhaps the reason the effects have gone unused on other platforms is because they weren't part of the platform. If every application is individually responsible for managing the effects, then of course few applications are going to use them.

If the effects are provided by the desktop environment independently and transparently (no pun intended) to the application developer, then everyone is going to use them.

That is the whole point of these effects. They are going to be part of the underlying desktop to provide more appealing visuals. It's much more appealing when switching from window to window to see the old window quickly tear apart or fold itself away, and the new window to quickly and smoothly slide into place than for the windows to just suddenly change.

It makes a lot more sense for the underlying window manager to be responsible for those effects than to burden individual applications with the responsibility.

Re:Why isn't this already out?

[ciroknight]

I think the real problem is, everyone knows X Windows is broken, but nobody knows what to do about it. A few reactionary people set apart to create projects like Fresco and Y-Windows, but the fact is, those are just as useless without knowing what was broken in X Windows in the first place.

As a Mac OS X user, a previous Windows user, and a current Linux Desktop user, I will not be the first to tell you that X is slooow. Windows seems more responsive than most Linux desktop distros I've used, and Mac OS X puts both to shame. Java applications on the Mac (or Windows) even seem more responsive than X Windows.

I still profess that the problem lies with the widget set/window manager not being integrated into the core, but that's just my opinion, and I'm not an expert. It just seems to me that there has to be some code that's shared between the two systems, and together, both systems generate excessive overhead that can be eliminated if we weren't so obstanant in preference of either KDE or GNOME.

To be honest, I'm surprised there hasn't been a project yet to integrate GNOME into X, which I'm surprised hasn't sparked a project to integrate KDE into X (two new forks). It'd be a nice graduate project if someone had the time, and I'd love to see what a GUI on linux could actually perform like.

Lastly, the problem comes with there being absolutely no good drivers available. Honestly, even though NVidia/ATi tries, they're not up to par with what they've got on the Windows platform, and Apple developers have had the luxury of seeing the developer's specs, so their drivers are just as impecible.

I think one of the best things that could come out of open source as of now, would be out of the ReactOS department. Find a way to use ATi's and Nvidia's drivers, then wrap them in such a way they can be used to draw X. I think that'd be an ideal solution, even if 20 people reply and tell me that this is technically unfeasable, and that licences and shit keep us from doing this legally.

Re:Why isn't this already out?

[Alan]

Well, y-windows' last release is over a year ago, and Fresco's last frontpage news item was 2003-03. Looks like there has been some work on it, between then and early 2004, with minor changes (based on the nightly snapshot diffs), but not a huge amount. The snapshot tarball has been 4.1mb for the entire time. I have the feeling that both these projects are basically dead :(

Re:Why isn't this already out?

[vadim_t]

Plugged in how?

This is something I spent a *long* time explaining to some people on a forum about a related subject. Network transparency doesn't involve significant overhead, dammit!

There has to be some way for an application to talk to X. So, you remove the network protocol, how do you want to talk it to X then, magic? In order for two different programs to talk to each other there has to be some kind of protocol, no way around it.

Now, networking indeed can slow things down a little due to things like latency. But that's effectively inexistent if you're talking on the local host. And X already has shared memory communication as well. On Linux there are also the so called Unix Sockets, which is pretty much like TCP/IP, only with even less overhead since it's done locally, so it can be implemented in a simpler way.

However, as far as an application is concerned, an Unix socket and a TCP/IP one are exactly the same thing, so it makes no sense to get rid of network transparency - you wouldn't win anything with that anyway.

Re:Why isn't this already out?

[hankaholic]

This gets dragged out every... damned... time... X gets a mention anywhere.

The argument goes like this:

A> X is bloated.
B> No it isn't.
A> But what about network transparency? That's useful.
B> You actually use that crap? Fine, network transparency is neat. But there should still be a way to shut it off.

Welcome to the next point in that argument -- the realization that any time the windowing system and its processes are running in separate processes, some form of communication will have to result ni order to allow the client to do such nifty things as detect mouse clicks or draw things to the screen.

Now that you're dealing with communication between processes, there are many ways to handle the problem. They all involve some mechanism by which communication can occur.

Let's see... two programs with a need to communicate with each other... we wouldn't have a mechanism which has been tested and refined over many years to be pretty darned good at communication, would we? Where would we... hmmm... let me think, I know that at some point there must have been the need for communication at some point in the history of computing...

Oh! A network stack! It's perfect! It not only allows high-performance, low-overhead local communication via highly optimized mechanisms which are available on damned near every operating system on which anyone in their right mind would ever want a windowing system (see contest rules for details, some exceptions apply, mileage estimates provided by EPA methods), but has a built-in mechanism for communication between machines!

Hoo, boy.

The point is, sarcasm mostly aside (maybe), that there is the need for programs to communicate. This isn't a requirement which you can simply opt out of, like, say, FAT support or unneeded sound card drivers. This is a requirement that you can't get rid of, and to use something that doesn't allow network usage is basically to limit yourself to mechanisms which are much less widespread in availability than an IP stack.

An IP stack is a good tool for the task at hand, and it just so happens to be really damned hard to remove networking from it.

I'd be interested if you could provide a counterexample -- find a widely available, generally reliable IPC (interprocess communication) mechanism which is for the most part platform-agnostic and doesn't require tons of massaging in order to get the data into the right format. Bonus points abound if it does not include the ability to communicate across a network, which you requested be unavailable.

Re:Why isn't this already out?

[BlueCodeWarrior]

It doens't matter if XP has the features or not...you're forgetting that the current version of Windows is Longhorn.

The correct answer is that 'Longhorn's tight integration with hardware thanks to Microsoft's close engineering work with card makers that provides a level of performance unachievable by any other vendor.'

Because, you know, even though Longhorn is over a year away we still talk about it in the present tense.

At least if you're within Redmond's Reality Distortion Field, which is growing to be as big if not larger than Jobs'...

Re:Why isn't this already out?

[snorklewacker]

> Network transparency doesn't involve significant overhead, dammit!

Well actually, yes it does. You still have to marshal pixmaps (a wretched and primitive yet still bloated format) into a shared memory segment just so the server can pull them out of there and transfer them to the graphics driver. And X's implementation of network transparency doesn't give clients any way to tell the server to aggregate events or even tell it to "shut the hell up already" with all the mouseover events over regions where they're not listening to the events.

Network transparency is a good thing. X's implementation of it stinks.

Re:Why isn't this already out?

[AKAImBatman]

Perhaps the reason the effects have gone unused on other platforms is because they weren't part of the platform.

Nonsense. The Mac has its Genie effect, both Windows and Mac can do window transparencies, and both systems have odd-shaped window support. None of these effects have gone to much use, because they are little more than eye candy. Which is primarily what I'm seeing in these videos. The Enlightenment interface doesn't even add the cool "Dock" animation abilities into it's interface. Instead it requires the user to manually resize icons until they're happy with the results. The only problem is, I have no idea how anyone would find this useful.

Compare that to the following concepts:

1. The icon you're hovering over expands to help you see what you're clicking on and hit the target correctly. (Mac OS X Dock.)

2. Windows can be flipped over so that notes can be placed for later reference. (Sun Looking Glass)

3. Objects on a table-screen can be moved by dragging your finger, rotated by moving a second finger outside of the object being dragged, and copied by dragging two fingers around the object(s) to highlight an area. (See here)

See how much more useful those concepts are than "floppy windows"? That's not to say that the base technology isn't useful. It's damned useful, and will be the basis for future desktops. But I see no actual UI research, and the tech has been available on X11 for several years now.

Re:Why isn't this already out?

[justsomebody]

Integrating Gnome or KDE in X???

Just where do you live? Insane planet? X is underlaying layer for desktops not desktop it self.

Again, integrating Ati and Nvidia drivers in X?

People do use other cards you know. You can't make X specific for those cards, and say screw others.

What's next that you hope? Integrating X with kernel and CPU?

As for not being responsivene, either you screwed with settings, or you use some partialy supported video card.

Re:Why isn't this already out?

[WillerZ]

You need to look in either Andrew's arch repository or mine for a more up-to-date version of Y. It's still years away from being usable wherever you get it from, so don't bother unless you want to hack on it. See my other rant: http://slashdot.org/comments.pl?sid=146748&cid=122 93465.

When Y gets to the slightly-usable stage I'll submit a story to /. myself.

Phil

Re:Why isn't this already out?

[vadim_t]

Right. But you need to do this marshalling one way or another. Remove network transparency and you'd *still* have to provide data in some particular format and follow some particular protocol.

My point is that just getting rid of the network socket won't help you any. You'll still end sending exactly the same stuff but using another mechanism, incurring in pretty much the same overhead.

It's not like getting rid of the socket would suddenly free you from the need of communicating in some standard way with the server. How would it understand you otherwise?

Re:Why isn't this already out?

[Peter La Casse]

1995: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
1998: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
2000: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
2005: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!

We have really neat X11 desktops right now.

Re:Why isn't this already out?

[XMyth]

I'll give him the responsiveness comment if he's using a pre 2.6 kernel.

The rest is just gibberish.

Re:Why isn't this already out?

[amightywind]

Your cynicism is predictable and misguided. X11 has been a rich garden growing great UI ideas for 15 years. There are a large number of highly successful UI and desktop projects that make GNU/Linux/X11 the diverse and exciting platform it is. Windowmaker, Gnome, KDE, XFCE, Enlightment and a millions others. Hell I even like TWM! Disatisfied with X? Propose something better. The Xgl server work is most promising.

Re:Why isn't this already out?

[ciroknight]

Integrating Gnome or KDE in X??? Just where do you live? Insane planet? X is underlaying layer for desktops not desktop it self.

And that's a reason I think it is broken. Most current desktop operating systems have realized that the underpinning technologies for running the Graphical User Interface are just as important as the overpinnings that make it look good, and make it useful. The "middlepinnings" and the "sub-underpinnings" like OpenGL and transparency/image blurring/antialiasing/supersampling in Linux are other examples of screw ups; where everything is implemented as seperate projects, they are all going off in different directions, each leaving each other behind. If there were a central coordinating force, they'd be working together, and getting somewhere.

Again, integrating Ati and Nvidia drivers in X? People do use other cards you know. You can't make X specific for those cards, and say screw others.

I said nothing of this? I just said that X needed better drivers, and maybe a way to do this is to wrap the Windows drivers. Besides, these are the two largest graphic card companies, and other graphic card companies have already released not only specs, but open source drivers, and it was announced here. I'm too lazy to do a search and provide links, as I'm already late for class at the moment.

What's next that you hope? Integrating X with kernel and CPU?

Hell, Windows integrated their GUI into the Kernel ages ago. Mac OS X uses their graphical environment as a kernel plugin. It's a tried and tested approach, but if you are seeking my opinion, I believe it to be flawed, and that X is doing the right thing in staying out of Kernel matters.

As for not being responsivene, either you screwed with settings, or you use some partialy supported video card. I have a Radeon 9200, 7000 and an NVidia TNT2. Sadly, the TNT2 proved to be most usable under Linux, but still performed better under Windows. As for ATi, I've yet to get either video card working.

Re:Why isn't this already out?

[bman08]

Yes, I have Longhorn installed on my 5tb/sqare inch pixie dust drive, and it's damned tightly integrated with my bitboys video card. I use CherryOS on it to emulate the mac at faster than native speed!

Re:Why isn't this already out?

[russellh]

Fresco. I thought it was dead when I looked into it in 1996.

Re:Why isn't this already out?

[ciroknight]

2.4 was the only kernel I've been able to use for quite a while. 2.6 seems not to like my hardware *shrug*.

As for the rest being gibberish, why?

Take a look around at other Windowing systems on other operating systems, and tell me what we can do. I simply present the three ideas: Move your Windowing System, Rendering subsystem, and Window Manager into one software unit for managability, code reduction and speed. Design a wrapper to use ReactOS/Windows drivers under X to support better drawing. Lose the whole flamewar about which is better between GNOME and KDE (we should be working towards this anyway).

Re:Why isn't this already out?

[AKAImBatman]

MODS: Will you please give this guy back a point of Karma or two? Just because he's got fruitcake ideas doesn't mean that he deserves to be silenced.

Parent Poster: You're way off base on the "problems" with X, but it does strike me that you're grasping for what the core issue is. The remaining issue with X.org/XFree86 is not with the controls, but rather with the monolithic architecture. X11 demands the full attention of the video card plus all other hardware it controls, and does not like to give up that control. The result is that X11 tends to be inflexible for work outside of "normal" desktop usage.

What X.org/XFree86 really need to do is separate the graphical and interface device layers from the desktop interface layer. i.e. It should be possible for any system program to be able to ask for exclusive video card access, even when the Desktop is not running. Currently, you have to chose between running X11 or using the external SuperVGA library.

You'll note that this is how the X server functions on systems like Sun Sparcs. On my UltraSparc, the system is ALWAYS in graphics mode. Running X just makes the X-Server take over the graphical screen. Very modular, very flexible. Not to mention that it places the graphics card support back in the OS drivers where it belongs, and not in a server with a focused purpose. Remember, the Unix philosophy is to keep everything in simple, bite sized chunks. The X.org/XFree86 implementation is the anti-thesis of that (although they are attempting to compensate with their portable driver loader design).

Re:Why isn't this already out?

[bushidocoder]

... the realization that any time the windowing system and its processes are running in separate processes, some form of communication will have to result ni order to allow the client to do such nifty things as detect mouse clicks or draw things to the screen.

I don't know much about the guts of X and windowing in general, but what exactly is the benefit of having windowing in an external process? It would seem to me that if applications dynamically loaded each process could handle its own windowing with background threads. I realize that this is a pie-in-the-sky theory since it would be very difficult for different drawing libraries to interoperate, and that library versioning would become very difficult to maintain, but are there other major technical reasons we have an X process? Are interoperability and backwards compatibility the principal reasons for X?

Personally, I don't care about the network transparency of X - I'd prefer the remote capabilities to be built on top of the existing drawing system as remote use seems to me to be more the exception than the rule, and in the remote access scenarios I use, session management and persistence is very important to me.

Re:Why isn't this already out?

[Ed Avis]

When I first came to Linux I couldn't figure out how applications displayed things on the X server. I was used to the idea that to draw a line on screen you save the return address on the stack, then jump to some operating system address. It draws the line for you and then returns.

So... how did this work on Linux? Was the X server process mapped into memory at a fixed logical address so other programs can know what routines to call?

It was rather a surprise to find that to draw something on the screen you had to format a message and send it down a TCP/IP link or at best a Unix socket, involving at least two context switches. No wonder computers are so slow nowadays...

Re:Why isn't this already out?

[AKAImBatman]

Did I say I was dissatisified with X? No? So you ASSUMED I was dissatisfied with X. Hint: Assume makes an ASS out of U and ME.

The only point I'm being critical of is the amount of fooling around that has gone into adding OpenGL features to XFree86/X.org. The XGL extensions have been in the works for years and (as far as I can tell) they should have just been polished off and plugged into X years ago. It doesn't matter that nothing immediately uses the features. Get the features there, and allow the KDE, GNOME, XCFE, and other projects work out what they want to do with the features.

Then there's Enlightenment. *sigh* For a *decade* now it's been able to do better eye candy than everyone else. But instead of leveraging that in some useful way, they keep messing with it. Every time Enlightenment gets somewhere, they notice the "next cool idea" in UI eye candy and run off to redesign the system around it. Meanwhile, the rest of the world never manages to obtain a release.

Re:Why isn't this already out?

[vadim_t]

It's a nicer approach, though. The X server doesn't need to be a part of the kernel, which makes things safer and simpler. And it can be not installed or replaced.

Besides, context switches are pretty fast on Linux anyway, and I don't see why communication couldn't be optimized to minimize the amount of context switching needed.

One thing I do wish is that the kernel was what handled all the graphics. Then the X server wouldn't need to run as root and do strange things with hardware, which IMHO is a task best left to the kernel.

Re:Why isn't this already out?

[Anonymous Coward]

On Channel 2, the Linux community screams that network transparency doesn't add a lot of overhead to the graphics system. Meanwhile, on Channel 4, the Linux community screams that monolithic kernels beat microkernels because procedure calls are more efficient than message passing.

These aren't necessarily opposing opinions: the graphics system is different than a kernel. But in a GUI environment, both are going to be responsible for a *lot* of communication, so I find it intriguing that people are able to hold both opinions at once.

Re:Why isn't this already out?

[teknomage1]

If X wasn't a process it'd have to be a kernel service, and itegrating X with the kernel would make the kernel bloated, open the kernel up to security issues from the windowing system, and generally screw up a lot of things. But every application wouldn't be drawing it's own stuff, it'd have to message the kernel. If every application tried to draw to the screen by itself you'd have unusable chaos instead of a desktop. Look at the state of things with TWO major drawing librries, now imagine if they had to actually talk directly to the screen buffer. It'd be a nightmare and leap us back to the days of DOS 6.0.

Re:Why isn't this already out?

[Em Adespoton]

Yeah wasn't there a Y windows in the works at one point.

I've been asking myself that for years now :D

Re:Why isn't this already out?

[justsomebody]

Psssssssssssssssssssssssssttt.

But here I bite again:)

Most current desktop operating systems have realized that the underpinning technologies for running the Graphical User Interface are just as important as the overpinnings that make it look good, and make it useful. The "middlepinnings" and the "sub-underpinnings" like OpenGL and transparency/image blurring/antialiasing/supersampling in Linux are other examples of screw ups; where everything is implemented as seperate projects, they are all going off in different directions, each leaving each other behind. If there were a central coordinating force, they'd be working together, and getting somewhere.

Yeah, and now imagine how large one project would that be. Companies do it this way, just because it is everything in one large product and everything is decided in that company. Not that it is better this way, but the more monolithic development hierarcy is in company, the more order it brings. Which is completely opposite of open source, where communities work on their project and communicate over specs with other project.

Now imagine such monolithic open source project. ...wrap the Windows drivers

Both Ati and Nvidia say that they use same codebase for windows and linux driver. Your wish is granted.

I have a Radeon 9200, 7000 and an NVidia TNT2. Sadly, the TNT2 proved to be most usable under Linux, but still performed better under Windows. As for ATi, I've yet to get either video card working.

Ok, I have Nvidia 4200, 5700, ati 9000 mobile, ati 9200 se and some older cards on my servers. All I can say is that both Ati cards worked like a charm under fc2 and fc3 out of the box. I even got more fps with xorg driver than with ati driver, except that I was missing one functionality (demo of spotlight for OSG didn't show any spotlight).

Oh, forgot Parphelia was a real bitch, matrox released drivers that supported 3d 1 year after I bought the card. So I don't use my Parphelia.

Best I can tell you is, insert 9200, download Ubuntu, install, test performance. That's the easiest way to test something. You can even try Ubutu live, although I can't say which driver is enabled there. Should be radeon if everything is done as it should be.

Re:Why isn't this already out?

[FrangoAssado]

While I agree with most of this, it should be pointed out that local X connections (i.e., the ones in which the server and the client run in the same machine) are usually done via UNIX sockets, not IP (witness the socket in the /tmp/.X11-unix directory while you are running an X session). The reason for it is that UNIX sockets offer less overhead than a TCP/IP stack.

Also, in modern toolkits (GTK, QT) images are usually sent via shared memory (again, only possible when the client and server are in the same machine), which is much more effecient than sending them through sockets.

Re:Why isn't this already out?

[nickos]

21. Sun's Looking Glass Project."

Take a look at the Croquet Project. It's trying to do the same sort of thing but without the Sun/Java headache.

Re:Why isn't this already out?

[0xABADC0DA]

About Project Looking Glass, Sun didn't "really take the right path" since it was done by one guy in his spare time. It really goes to show the productivity gain of using Java (or equivalent) -- it's not like other people in the Unix/X world haven't implemented similar eye-candy, but the difference is that implementing the fine details in C/C++ is a nightmare of complexity so you just get individual hacks like the 'wobbly window'. Leave it to others to implement a 'dock' for istance.

Many posts praise X windows' network transparency, but the fact is that it is too slow for realtime games and bit-movers like VNC are fast enough that X's transparency is of little real benefit. Sure current VNC's are slowish, especially when they lack a driver to detect dirty regions, but having used Radmin on windows I can tell you this model is perfectly workable (radmin is that much better than even tight VNC). Also note that the Java VNC applet is faster than native linux vnc (try playing a video using each).

It's not that I am a Java bigot so much as trying to make the point that a moderm X server is not concerned at all with low-level, for-loop performance; it's entire job these days is managing little bits of information (window dimensions, events/listeners, properties, etc). Implementing that in a complicated language like C (or C++) and having a protocol instead of an API is just too archaic these days. Manual memory management combined with lots of fine-grained data and structures is just ridiculous. It's ridiculous.

And so what if I mentioned Java? Use freakin' mono if you want, but there's no two ways about it: the X server is too old and archaic for modern interfaces.

Re:Why isn't this already out?

[Em Adespoton]

I just said that X needed better drivers, and maybe a way to do this is to wrap the Windows drivers.

That would be Windows IAx86 drivers, hmm? Although I guess embedded Linux doesn't really need those drivers, and PPC driver support can be a fork; but wouldn't some sort of plugin system akin to kernel plugins work better? --a "wrapper" plugin could be one of the options.

Re:Why isn't this already out?

[x2A]

So... you want every single process that draws to the screen to have to deal with talking to the display driver itself? Each application you want needs to have support built in for each graphics card that X would do? How would each window know whether to draw on top of other windows or behind them? How would they know whether keyboard input was meant for them?

You'd need a single process to manage all of that... and guess what?!

Or did you only want to display one application on the screen at any one time?

-2A

Re:Why isn't this already out?

[ecloud]

While at times I've been firmly in the "There needs to be an X with less crap in it" camp, I've learned to really appreciate the network transparency.

Network transparency is part of the fundamental design. You can create a windowing system from scratch that is much simpler because it isn't designed for network transparency. But ripping it out of X doesn't change much, because the APIs for programming will still be just as complex.

I finally got one of my TuxScreens running Linux last night. (Yes I know... I'm very slow getting around to some projects. This I should have done about 3 or 4 years ago.) I was impressed that it can run tinyX with only 16 megs of RAM, it's reasonably quick, and because there is no security, I can export my DISPLAY variable on my desktop machine and run any X program and have it display there. I really thought it would be harder than that, having gotten accustomed to it getting harder with newer distros, because the security people keep interfering with useability. I have yet to figure out how to do this at all with Gentoo.

However, it's much slower to display a full-screen image with xv than it is to run something simple like xclock or xterm. And modern themeable toolkits make applications much worse for network transparency. At my last job we had some old NCD Xterms and it was kindof slow to run something like Mozilla or KDE on them, while the old stuff was still as snappy as it ever was. I'm a big believer in vector rendering rather than the "pixmaps galore" approach that's currently in vogue for theming. Many programmers are not designing for network transparency anymore.

So when you say X has too much crap I'm thinking more of the "cool" modern features (and obsolete features that were cool once). At least, most of those are X extensions, and bloating of the toolkit libs. Network transparency though is indispensible IMO, and to get rid of it you might as well start over rather than use X at all. It still amazes me that X is so old, and so advanced and so up-to-date and useful at the same time.

Re:Why isn't this already out?

[vadim_t]

I did not say anything about kernels.

I'm absolutely not an expert in the area, but I find the idea of a microkernel pretty cool. But I'm not a fanatic of any of those designs, and will use whatever works for me - currently that's Linux with its monolithic design.

In my rather uninformed opinion, microkernels sound quite complex. Now, I'm not a kernel programmer so I think I'd better leave the actual kernel programmers to debate the pros and cons of the approach.

There's also that drawing a window is a pretty complex thing. I'd say that the overhead needed to send the command is probably pretty much insignificant compared to the time needed to calculate what to draw where, and execute the actual operation.

Re:Why isn't this already out?

[x2A]

It's much more appealing when switching from window to window to see the old window quickly tear apart or fold itself away, and the new window to quickly and smoothly slide into place than for the windows to just suddenly change.

That depends on whether your computer's already slowing you down or not. I want an accelerated interface, not one that throws a bunch of extra frames in between the one's I can use. An animated interface is only any good for people who don't use their computers enough to be as fast as it is.

The more people have to wait to see what they already know is there, the more they'll be turned away.

-2A

Re:Why isn't this already out?

[biglig2]

Mod the parent +0: Funny, but too subtle for the Slashdot crowd.

Re:Why isn't this already out?

[bushidocoder]

I don't want anything - I'm just asking.

That said, isn't this similar to how Win32 drawing works? Granted the display driver interface is in the NT kernel, but its my understanding that each process is responsibile for its own windowing operations, abstracted through the Win32 interface. The NT kernel also handles the input, but even if you handled input in its own seperate process and messaged each process individually, messaging the location of your mouse at each instant requires a substantially smaller communication buffer per process than a buffer capable of messaging full bitmaps back and forth.

I'm just trying to get an understanding of how it works.

Re:Why isn't this already out?

[swv3752]

What gets reported in TOP is not really the amount of memory being used. There is a fair amount of double reporting going on. Go ahead, count up the amount of ram every process is using and the total amount of Physical memory and Swap memory being used; and notice a discrepancy?

Re:Why isn't this already out?

[SnowZero]

X aggregates drawing primitives, which is really what you want. In the big scheme of a windowing system, the time processing events is lost in the noise. What do you think the average ratio of drawing primitives to input events is for most applications? I'd guess its at least 1000:1 for anything with a modern gui.

Re:Why isn't this already out?

[labratuk]

You're stuck in the early 90's.

You know. The time when, for instance, if you were writing a game you had to write support into it for each brand of soundcard you wanted to work.

Before anyone is allowed to use the word 'bloat', they should have to understand the word 'abstraction'.

Re:Why isn't this already out?

[XMyth]

2.6 has MUCH better GUI performance. On par with what I've gotten with FreeBSD (go figure) in the past. I'd keep trying new 2.6 releases if I were you, it's worth the upgrade.

As for integrating GNOME or KDE into X, I don't even know what you mean on a technical level. Perhaps you could give me an example?

You don't think that explorer on Windows compiles seperately from the MFC? It simply links to it as a library (statically or dynamically, I'm not sure, doesn't really matter). This is no different in the Unix world.

I can see how it appears to be different because you have more options here where things change radically. But really, it's not all that different.

The biggest difference is how windows are drawn and messages are passed. With these latest developments, both Windows and X are moving towards the same thing (and I think OS X has been doing this) as far as drawing windows goes.

Re:Why isn't this already out?

[bushidocoder]

I realize that, but there are places where memory is allocated inproc as nonshared - see this from GnomeLive about the total cost of buffers for X.

Re:Why isn't this already out?

[Zphbeeblbrox]

look at your average winXP installation with applications and count the processes. hrmmm... quite a few there eh? Now tell me what each of those svchost processes is actually hosting? Not easy is it. There are benefits to the way it's set up on Gnome and X. If the clock on your windows taskbar should for some hypothetical reason lock up how would you kill it? Kill Explorer? that just killed your whole window manager. Sure you can restart it with a handy little ctrl-alt-delete but you just lost all that stuff you were working on.

Me I far prefer having all those apps launching in nice easy to see processes. You don't like how much space it takes in memory? use a different one. Admittedly Gnome makes this harder than say FluxBox but the choices are there.

Explorer on my WinXP box at work takes up 20 megs in memory and has who knows how many threads. X and whatever desktop/window manager you use just make it easier to see what's doing what and change what you don't like.

Re:Why isn't this already out?

[MBGMorden]

Kill Explorer? that just killed your whole window manager. Sure you can restart it with a handy little ctrl-alt-delete but you just lost all that stuff you were working on.

Killing Explorer in Windows doesn't kill all your apps. The only thing it might kill are explorer windows you had open, and sometimes you won't see all the same tasktray icons (though the corresponding program will still be running). Other than that everything stays up. Heck you can even continue to use most programs without even starting Explorer so long as you don't minimize them.

Re:Why isn't this already out?

[ciroknight]

The largest open source project ever seems to be working fine; Linux itself.

Just because it is one body doesn't mean that it can't have sub-bodies. Compare with the Linux kernel; broken down into archetectures.

Now onto the bitch of the subject. X is huge. So are KDE and GNOME. My proposal would be to integrate X and GTK/Qt (which is what I should have said in the first place, my thinking/typing skills have been on the fritz today, and what I typed earlier was before my 1:00 class, so undoubtedly it was time compressed). This would require an entire rework of the X protocol most likely, which is something we as a community should think about.

My entire argument comes down to this; X is old. All code needs to die sometime, and X is a perfect candidate for being put out to pasture, and a good teaching example on how a windowing system can work, due to its simplicity. But we're in a world now with needs beyond what X can deliver, and we're at a point in which we're still dragging it along. Meanwhile Mac OS X has managed to implement an entire drawing engine, add nice hardware accelleration features (Quartz Extreme), make them useful (Exposé), and add to their usability once again (Dashboard). Windows "Avalon" (if it ever gets done) will add to the mix SVG interface definitions (I wouldn't doubt if Apple doesn't come out first with SVG-window rendering capabilities).

Lastly, we need to do something we really haven't done with anything up to now (well, Mozilla and GNOME/KDE excluded; they've been really good at this in the past). We need to _design_ and _engineer_ the software, not just code it. Of course, engineers expect to be paid (even I have came to this conclusion after my last project, Charity), which is another reason open source software isn't innovating as much as we should hope.

Re:Why isn't this already out?

[Ambassador Kosh]

Memory usage reported by X is not even close to accurate. The X process also memory maps your video card in most cases and even the reports the agp buffer size sometimes. So if you have a large memory card you can sometimes see X reporting that is using many hundreds of megs of ram. I have 3 vid cards in this box and X will sometimes report up to 400M of ram used. However I have 1G of ram and the system will show almost all of it free despite what X is showing.

The problem is that accurately showing up how much real memory is used is a VERY difficult process. Linux shows it one way, Windows shows it another way and neither are close to what people probably think of as right.

Re:Why isn't this already out?

[x2A]

You need [at least] a single entity to manage the whole desktop, to manage where input is sent, where things are drawn (as I said in prev post, each window would need to know which windows are in front of it. Without a single entity managing this, each app would have to communicate with each other app to ask it "do you have a window in front of me?").

This entity could be a server, such as X, or built into the kernel. Inside the kernel is faster as there are less context switches required, but you don't get the protection of running in userspace.

Perhaps being able to upload code into the kernel? Either interpreted bytecode, or a bytecode checked 'n compiled to 'safe' native code?

As for 'messaging the location of your mouse at each instant' - to each running process? How do you know if it's interested? You're better off registering to have events sent to you with the managing entity, so it can tell you when the mouse moves, moves over/out an area, clicks, etc.

Oh one other thing, threads still require a context switch, although reusing the same address space means not having to flush TLD's, you still have to save/restore CPU flags/registers/etc states.

-2A

Re:Why isn't this already out?

[ADRA]

The described concepts are base components of a more usable system.

The most important component that I see is the off-screen compositing engine. Basically, X can create and store an 'image' of every window on an active session. With that, you can add cool things like in the mac, being able to see a snapshot of all running programs on the screen shrunk.

1. I believe apple patented this, but I could be mistaken

2. I don't really see how this is a productivity gain. Maybe I'm just missing the point

3. Once again, I fail to see how this useful. Unless the UI enhancement is simple, it won't fly for normals.

I don't think leading X11 down some utopian UI for a select few people will do much good.

Floppy windows are a nice concept, and I imagine it was created as proof of concept for the underlying technologies used.

Re:Why isn't this already out?

[LionKimbro]

My point is that just getting rid of the network socket won't help you any. You'll still end sending exactly the same stuff but using another mechanism, incurring in pretty much the same overhead.

Well, what the other guy was saying makes sense to me.

If you're communicating by sockets, you have to make two context switches, right? One to call the kernel, and one from the kernel to the X server. Whereas if there was a system API, then it would be just one context switch: you call the kernel, end of story.

If we count entire paths, it would be: you-kernel-x-kernel-you, vs. you-kernel-you.

So, I would imagine that your time from calling to the time of receipt would be much faster. This could be very important in games, I would think.

Re:Why isn't this already out?

[Ian Bicking]

There has to be some way for an application to talk to X. So, you remove the network protocol, how do you want to talk it to X then, magic? In order for two different programs to talk to each other there has to be some kind of protocol, no way around it.

Sure there's a way around it: an API. X has an API (xlib, the C library). But it also has a protocol, the means by which xlib communicates with the X server. In theory the protocol could be removed now without effecting most client programs, though I assume that would effect the implementation dramatically.

In a kind of parrallel situation, the GNUStep people were trying to figure out what to do with Display Postscript -- they didn't have a good or fast implementation, and DPS was supposed to be central to the NEXTSTEP design they were copying. Then someone hacked together something that implemented the Objective C bindings that generated Postscript, and made them call the xlib libraries directly, because the whole DPS thing was taking too long. It was meant as an intermediate step... but it worked. Last time I looked (which was a long time ago) they had given up on DPS as a pointless abstraction. If they want to implement accurate printing later, they can do so by creating another, entirely separate Postscript-generating layer (but without having to worry about performance or interactivity or other non-print related things).

The same possibility exists for X. Give up on the client/server and network transparency. If it matters so much, you can easily implemented it on top of a non-transparent system (as long as you have a defined API with no back doors). If it matters so much there's probably multiple ways of implementing transparency with different tradeoffs. Various hardware would have different implementations of the xlib C API -- those implementations could share code, but only if they wanted to. All the other policy, like fonts and security and whatnot, would be implemented separately, which is already happening anyway.

We have a layered system. And that's fine (and even if it wasn't it can't be changed). But X got some of the layers wrong, I think it actually did too much. That can be fixed.

Re:Why isn't this already out?

[aichpvee]

What exactly has sun done with window shading that's so evolutionary even? I've had that for years.

Re:Why isn't this already out?

[EnronHaliburton2004]

I think there was a Who windows in the works, but the project was later assigned to third base.

Re:Why isn't this already out?

[AKAImBatman]

What exactly has sun done with window shading that's so evolutionary even? I've had that for years.

Not the Looking Glass style, you haven't. Instead of folding the window up into the title bar, the desktop allows you to shift it ro the side as if it were a book on a shelf. You can see the title on the spine, as well as a slight amount of the window's front. And since the window is on the side, it's completely out of the primary work area! :-)

Re:Why isn't this already out?

[Jherek Carnelian]

If you're communicating by sockets, you have to make two context switches, right? One to call the kernel, and one from the kernel to the X server. Whereas if there was a system API, then it would be just one context switch: you call the kernel, end of story.

That is an argument for putting the GUI in the kernel, not for getting rid of network transparency,

If we count entire paths, it would be: you-kernel-x-kernel-you, vs. you-kernel-you.

No, a "context switch" is from one process to another, a process into the kernel is not a context switch, at least not in the same way.

For example:

1) FP - process 2 process requires save & restore of floating point registers, but process to kernel does not because the kernel does not do FP

2) General registers - process 2 process requires save & restore of all general registers because each process has its own state, when you make a system call, you bring all of your general registers in with you since the "state" of the system is your currently running process. Somem will be saved off in case the kernel clobbers them, to be restored on the way back out to the user process but worst case, that's only one save & restore and if you go back to a different process (which is usually the case, even in your example) then the number of saves & restores is the same either way.

3) There's more, but I'm too sleepy to type any of the rest.

Re:Why isn't this already out?

[x2A]

Or even use the task manager to work with minimised windows!

-2A

Re:Why isn't this already out?

[Fnord]

Ok, something I always wondered why no one's done. Yes for most things, unix domain sockets are fast enough. But the X-SHM extension is MUCH faster for some operations (that mostly involve sending lots of pixmaps). But an application has to be written to use it. Well, we have an abstraction layer here that could dynamically do this, it's called X-lib. Why has no one written a modified X-lib that detects if you're connecting directly and automatically converts some operations to X-SHM?

Re:Why isn't this already out?

[Fnord]

That was SOMEWHAT of the case with Win9x. What a lot of people don't realize is that NT and later actually have a display server. Its just very well hidden.

Re:Why isn't this already out?

[Brandybuck]

You are the classic case of someone not seeing the forest for the trees. "But how can this single oak tree sustain and entire woodland ecology? It can only support a half dozen saprophytes and one squirrel family!"

If you run X in synchronous mode, so that every draw command gets processed one at a time by the server, then yes it will be dog slow and nasty. But that's not the default. Normally your commands get queued up and sent over in batches at certain defined times. In this type of scenario the X11 way is every bit as fast as the win32 way (for example).

Today I was using netscape from Solaris over a slow network to a 100Mhz P1 running FreeBSD/XFree86. No problem. It wouldn't be suitable for running tomorrow's hot new first person shooter, but neither is win32, which is why Windows has DirectX. If you want to argue that X11 needs a DirectX analog, I will fully agree with you. But that's a completely different issue.

Re:Why isn't this already out?

[TheRealMindChild]

I think the "bitching" stems from the fact that X requires a seemingly unrelated subsystem to be functional.

I could be talking out of my ass here, but I would expect that what one would REALLY want is abstration at the DISPATCHING level... use some mechanism to dispatch the communication between the client and server, then an OS with no need for networking could use X by replacing the implementation with

Re:Why isn't this already out?

[Brandybuck]

This could be very important in games, I would think.

The graphics engine needed for a game is very different than the graphics engine needed for most other things. The former almost requires a kernel bypass to write directly to the video hardware. But that's overkill for drawing buttons and edit fields and window frames.

Re:Why isn't this already out?

[vadim_t]

Um, and how exactly would you communicate? So, you remove the protocol. And then?

An API is an interface that can be provided by instance for a library. Or it can be provided by an application and used by a library (plugins). But your application isn't *linked* to the X server, so it can't just call functions in it!

The only way to make what you say possible would require each application to be in the same address space as the X server, or for the X server to be in the kernel. Both approaches are quite horrible, IMO.

If the application is a module in the X server then a failure in the application can bring the whole server down. And there's no possible to have applications running with different permissions. Just wonderful, we can go back to MS-DOS levels of security and stability!

X in the kernel is a more viable approach, but forget about it. The kernel developers don't want it there for a good reason, and this isn't exactly removing the network protocol anyway - it's an entirely different system that wouldn't even look remotely like X.

I get the idea that you don't know much about software design. I'll say it again: You can't get rid of network transparency that easily! It has to be there because there has to be some way to talk to the server. I'll try to explain this way. Here are the most common forms of Unix IPC:

Network sockets
Unix sockets
Pipes
Shared memory
Message queues

Guess what? Except shared memory all of those are pretty much the same as network sockets. RPC (remote procedure call), which is probably what you mean by "API" works through network sockets. And shared memory is inconvenient to use for some things.

So, again: Unless the X server moves into the kernel, or the application moves into the X server there's simply nothing better than networking to make the application and server talk.

Re:Why isn't this already out?

[FooBarWidget]

Repeat after me: all modern graphics systems use IPC! If you want to get rid of the context switch overhead then you have to write directly to the hardware. There isn't a single operating system which allows you to directly write to the video hardware (with some exceptions, like DGA). Not MS Windows, not MacOS X, not BeOS, not whatever. X as it is right now is not much different from MS Windows or MacOS X as far as communication is concerned: on all platforms, the application has to send a message to an external entity which then draws stuff to screen. You have to do this, otherwise you'll get into synchronization problems with other applications.

So why don't people complain about the IPC overhead in Windows? Because the anti-network transparency hype is overrated.

Re:Why isn't this already out?

[Jherek Carnelian]

These aren't necessarily opposing opinions: the graphics system is different than a kernel. But in a GUI environment, both are going to be responsible for a *lot* of communication

The problem is that within a kernel everybody runs at the same privilege level, although micro-kernels do have the concept of different subsystems running with different privileges, it usually ends up not being a valuable feature.

But the X server does require privileged access to hardware, often the privilege is enough to do bad things to the system if misused. Thus you can not (safely) just have the user making procedure calls into the gui because they would, by definition, need to run at a high privilege level which means the rest of the user process could also do bad things to the system too.

In theory, with either special hardware and/or a fancy compositing manager, regular user processes could each draw in their own shared memory buffer (either in system memory or on the card) and just about all the gui calls could be implemented via (safe) procedure calls that just scribbled on the private buffer and relied on the compositing manager to (safely) composite them all together on the actual display. But, AFAIK, today's hardwware does not support that kind of use and so the software isn't there either.

Re:Why isn't this already out?

[nickos]

Stop using KDE, GNOME and apps that use qt or GTK and switch to a lightweight window manager and apps that use the lighter toolkits and your system will fly along. This should demonstrate that the problem is not with X but higher up the GUI stack.

Re:Why isn't this already out?

[vadim_t]

And that's what you people don't get - it's not unrelated!

Networking is simply the best way to communicate client and server. You can choose from: Network sockets, unix sockects, pipes, queues and shared memory.

TCP/IP sockets and unix sockets are functionally equivalent (except that the later are only local and have less overhead). Pipes are unidirectional. Queues aren't used much by anything AFAIK, and shared memory is an inconvenient mechanism for exchanging messages.

Re:Why isn't this already out?

[Brandybuck]

Why has no one written a modified X-lib that detects if you're connecting directly and automatically converts some operations to X-SHM?

Just a wild guess here, but "because no one really needs it?" I have a tiny bit of trust that the guys working on Xorg have already thought of this rather obvious idea, and rejected it for valid reasons that we on the outside can't immediately discern. My initial guess is that the complexity isn't worth the miniscule performance gains, especially since the major GUI libraries already use shared memory where appropriate.

Re:Why isn't this already out?

[4of12]

I've learned to really appreciate the network transparency

I do, too.

Unfortunately, in the world at large when you mention X Windows you get a confused look as the person wonders if what you're talking about is either somehow related to Mac OS X or is a "Brand X" pale imitation of the wonderful brand name product that is genuine "Windows".

I suggest X.org get with the times.

You've heard of iSCSI for accessing SCSI disks over IP? It's cool. It's recent. It's what all the hipster and groovy people are into.

The next generation of X Windows should not be called something like X11R7, X12, or Y.

It should be called something like iVideo with its own RFC and an i so all the PHBs and IT journalists will know it's cool because it flows over the network, dammit.

[OpenGL hardware level performance is one checkbox item; the other is some infinite resolution support so SVG, PDF, etc. can be rendered without as much intermediate interpretation.]

Re:Why isn't this already out?

[poofyhairguy82]

First let me say: good post. It brings up a good discussion. I hope to not offend you with my response.

The "middlepinnings" and the "sub-underpinnings" like OpenGL and transparency/image blurring/antialiasing/supersampling in Linux are other examples of screw ups; where everything is implemented as seperate projects, they are all going off in different directions, each leaving each other behind. If there were a central coordinating force, they'd be working together, and getting somewhere.

You mean a central force like Apple or Microsoft. The Linux Desktop is what it is - a pile of OSS projects. Despite tons of calls to "unify" it will never happen. Heck, I'm happy enough that such projects work together AT ALL, let alone act together well. If you can't stand the patchwork nature of the X, your only choice is to pull out a credit card and buy a Mac. Apple seems to do everything you want, but Linux doesn't want to be a crappy version of OSX (linux wants to be linux, it just happens to be a crappy version of OSX).

As for not being responsivene, either you screwed with settings, or you use some partialy supported video card. I have a Radeon 9200, 7000 and an NVidia TNT2. Sadly, the TNT2 proved to be most usable under Linux, but still performed better under Windows. As for ATi, I've yet to get either video card working.

Now I have figured out where your complaints come from- your only modern hardware (the TNT2 is not modern in a graphic card sense) is ATI hardware. ATI LINUX DRIVERS ARE CRAP!!!! this isn't Linux's fault, and nothing drastic needs to be done on the Linux side to force ATI play along. The company with either correct itself, or lose marketshare to Nvidia. I recently traded a 128mb 9600 pro in my Ubuntu box with a 64mb 5200FX. In Windowsland, the 9600 pro is twice as good, but in Ubuntu the 5200FX does everything AT LOT faster. You complaints about X sluggishness is foreign to me, and I use xcompmgr and my desktop is accerated to a point that I never got Windows to achieve (equal to OSX the few times I've used it). I know what you'll say- "Its not practical to tell people that they must buy new hardware in order to not have a sluggish desktop." To that I respond "my $30 Nvidia card was A LOT cheaper than a new (or even used) Mac." One day, ATI might actually have some decent Linux drivers (for 2D) and your sluggishness might go away...

Re:Why isn't this already out?

[rookworm]

Would you care to explain it for the slower-witted among us?

Re:Why isn't this already out?

[Ian Bicking]

By standardizing on only an API, the protocol becomes ad hoc. That is, it could be a kernel call, or IPC, or send data over the network, or do the work in process, or whatever. There are real and valid reasons for all of those options, even if each might not be reasonable in all cases. For instance, using the same API for printing is reasonable, and should probably be done in process. There's no reason to have a standard protocol, because there aren't multiple viable implementations of that protocol (at least on the client side, and I believe the protocol provides a difficult abstraction barrier for efficient servers).

If it's best to put the display code in the kernel, then this opens up that possibility, but does not require it. It opens up a wide variety of possibilities, where a protocol is limited to a client/server architecture. X's client/server architecture is the oddball in the world of display architectures, while a standard API is the norm.

I believe this is the way the GGI project is architected, so it's not a crazy idea at all.

Re:Why isn't this already out?

[poofyhairguy82]

Meanwhile Mac OS X has managed to implement an entire drawing engine, add nice hardware accelleration features (Quartz Extreme), make them useful (Exposé), and add to their usability once again (Dashboard). Windows "Avalon" (if it ever gets done) will add to the mix SVG interface definitions (I wouldn't doubt if Apple doesn't come out first with SVG-window rendering capabilities).

Meanwhile, Apple only puts OSX on very particular hardware (that they sell and know everything about), and therefore it costs them a lot less to do things. Thats why OSX is ahead. Its not developed better, but its closed nature makes it easier to program for. The only fair comparison would be Windows, as it has to suffer from the same "needs to run on a billion different configurations" problem that x86 Linux does.

Re:Why isn't this already out?

[civilizedINTENSITY]

I have to agree that window shading is window shading. Shade it up, or shade or over, it effectively replaces the square with a strip. Is there a benefit to a vertical strip rather than a horizontal strip?

Re:Why isn't this already out?

[vadim_t]

There is a standard API, the X library. It's not like its interface changes every week or something. I don't know of any applications that talk the X network protocol directly.

So, we have that already. You can change the network protocol, and it IS being changed and extended (Xrandr, Render, Shape, etc). But still, you won't get rid of it.

Please take an Unix book and open it. The classic Unix server either forks or multiplexes to handle multiple clients. If it forks, how does the parent server talk to the children? Usually using shared memory, unix sockets, or pipes. How do the children talk to the clients? Shared memory, network sockets, unix sockets, or pipes.

How does the X server talk to clients? Network sockets, unix sockets or shared memory. There you have it. It works *exactly* the same way virtually every other Unix server works. Nothing at all oddball about it.

Re:Why isn't this already out?

[AKAImBatman]

Is there a benefit to a vertical strip rather than a horizontal strip?

I would think that the benefits would be obvious:

1. Moves the shaded window out of the primary work area automatically.

2. Allows you to partially see any activity that may be going on in the shaded window.

Re:Why isn't this already out?

[Crazy Eight]

I didn't see anything in the photos or videos that's revolutionary.

I suppose it's not revolutionary, but I think that screenshot showing a translucent mplayer window is pretty impressive. The story a while back showcasing Luminosity had a video demonstrating Xv stuff showing up in a pager. That was on a laptop with Intel integrated graphics. As it stands right now I can see the Xv colorkey on the inner edges of a xine window if I move it around too fast. That's on a Geforce.

Re:Why isn't this already out?

[darkwhite]

plug-in, plug-out all kinds of useless^H useful crap.

Screw you. Lack of proper X support for switching between my laptop's LCD and external DVI output is my biggest problem with Linux right now. WinXP does it with zero user intervention.

Device hotplugging and autoconfiguration is the number one hardware support priority for Linux on the desktop. (The Linux kernel is not all that hot on some types of hotplugging either, and userland support for what it can do is braindead in almost all distros.)

Re:Why isn't this already out?

[fikx]

actually X is treated as the video driver for the most part. There,s good and bad with this idea, but it seems to be the mentality steering a lot of choices. Me, I like having the most problematic and changing driver put one layer above the kernal. That way, a crash in it doesn't barf everything else. Plus the ability to share the device across a network for free!
If only X protocol itslelf was allowed to be a general purpose driver instead of being held back by backwards compat, design confusion, desktop-specific kludges etc.

Re:Why isn't this already out?

[darkwhite]

er, hardware support problem, not priority. It should be a priority, though. I want to see a distro with actual no-shit autoconfiguration of all relevant devices.

Re:Why isn't this already out?

[Ian Bicking]

There is a standard API, the X library. It's not like its interface changes every week or something. I don't know of any applications that talk the X network protocol directly.

Exactly, which is why this can be resolved in a backward-compatible way (for clients, not servers of course). Which is exactly what I said in my original reply. Keep the xlib library, turn the X protocol into an incidental implementation detail which can be replaced. If you have an X server, then some protocol will be present. I don't believe the presence of an X server process is absolutely required.

Nothing at all oddball about it.

I said, very specifically, that a client/server display is an oddball architecture. Not that the way it implements client/server is odd.

I know how this stuff works, but I'm talking about architecture, not the implementation details you are focusing on. The architecture I'm talking about exists right now, it is mainstream (outside of X), and can be backward compatible with most X applications. GGI (using an API-based architecture) was a great idea, and I don't know why no one ever paid it much attention. I'm not arguing for Y, or Berlin, or any of those other things. I am arguing that network transparency should be implemented on top of this architecture, not as part of it, and X inappropriately mixes the two (and mixes all sorts of other things in there too, most of which it does poorly).

Maybe the X architecture isn't really to blame, but the traditional monolithic X development process. Maybe this architecture is only appealing because it makes it easier to route around that process. But there's a benefit in that too.

Re:Why isn't this already out?

[AKAImBatman]

Screw you. Lack of proper X support for switching between my laptop's LCD and external DVI output is my biggest problem with Linux right now.

Oooookkkaaayyy... And you are swearing at me for this because... ??? It's not like I wrote X-Windows. I just commented that it's gotten a lot better than it used to be. Hotplugging video monitors IS a big deal, but otherwise X tends to do reasonably well at cold-plugging devices.

Of course, the *correct* solution would be to move all the hardware stuff to the kernel and allow the X-Server to just do the Windowing System. This is the way that commercial Unixes are implemented, and it works quite well. Unfortunately, as you said, the Linux kernel hotplug support is dicey as well. Works quite well on FreeBSD, though. :-D

Re:Why isn't this already out?

[AKAImBatman]

actually X is treated as the video driver for the most part.

Which is a problem. Video Card support is a completely different thing from Window Management. They are related, but only in that the latter builds upon the former.

Me, I like having the most problematic and changing driver put one layer above the kernal. That way, a crash in it doesn't barf everything else.

What do you *mean* by this? Barfing in the hardware can crash a system just as surly as a kernel mode exception.

If you mean that the Windowing System should be in the user-mode area above the driver layer, then I agree with you.

Re:Why isn't this already out?

[hankaholic]

I don't know much about the guts of X and windowing in general, but what exactly is the benefit of having windowing in an external process? It would seem to me that if applications dynamically loaded each process could handle its own windowing with background threads.

If this were to work, you'd have each window responsible for drawing to the screen itself, deciding which mouse input was its own, accepting or ignoring key presses, etc., etc., etc.

Assume that you have two overlapping windows on the screen. If both simply drew to the screen without any form of communication, they'd be overwriting each other.

The background threads which handle this sharing of hardware (screen space, input devices) would still have to communicate with each other. By allowing the X server to handle things like deciding whose window gets drawn on top, you get things like overlapping windows without each and every application having to explicitly worry about deciding which parts of their windows are obscured, or how to communicate with every other running X process. It's much easier to have a single process (the X process) handle hardware management. Window drawing and management policies also must be maintained, but recognizing that this is a problem with many solutions the designers of X made this a modular feature, allowing the user to choose their window manager (or none at all) as they chose.

Re:Why isn't this already out?

[darkwhite]

Because you called output device hotplugging in X useless.

The hardware stuff (the video drivers) are already partially in the kernel (I'm not sure how the functions are subdivided between the X driver and the kernel driver). Completely splitting X into a hardware support/renderer infrastructure part and a windowing system part sounds like a great idea to me, too. And then, making the renderer accept vector data and use the GPU to process it would be awesome. OS X awesome.

Re:Why isn't this already out?

[GoCoGi]

You need a process that manages your windows. So you can have X in the kernel, or in userspace. Since it would bloat the kernel we go for userspace. Now some form of Inter-Process-Communication is needed:
If any sane person had decided to develop a non-network transparent windowing system, he or she would eventually have created something very similar to X11 (without the possibility of using TCP sockets) and then realized that it was trivial to add the network transparency to the system anyway.

Re:Why isn't this already out?

[hankaholic]

True, but a networking mechanism is still required, and there's not a large difference between using UNIX domain sockets and using IP past opening the socket initially. You also seem to have missed the point about being platform-agnostic -- UNIX domain sockets are nice, but merely being nice doesn't make them available on non-UNIX platforms.

Extensions (such as SHM) allow for more optimized actions to occur, but I'd imagine that supporting shared memory added more to the actual code base than allowing two whole types of network sockets. Xserver/Xext/shm.c is 1257 lines long. Grep around for TCPCONN in the X source -- there's really not much to supporting TCP over supporting UNIX domain sockets.

Essentially my point is that communication is necessary, and it's wrong to assume that allowing network transparency is a costly decision in terms of implementation costs, or that removing network transparency would even be a meaningful operation.

Re:Why isn't this already out?

[ScytheBlade1]

Ctrl+shift+escape, ctrl+tab to the processes tab, end the explorer.exe process(es). Ctrl+tab back to the first tab, tab to the new program button, hit space, type in "explorer", and hit ok.

Congragulations, you just restarted explorer without:
-Touching the mouse
-Having to worry about explorer responding or not
-Rebooting
-Closing any open programs
-Losing any data

Additionally, switch from the initial craptastic XP theme to the "standard" theme (2k style), and the RAM usage will just plain drop. Disable of of the other "fun" features (like customized menus based on usage), and explorer will only use less memory.

The "svchost" processes are rarely actual "servers" "hosting things", rather they commonly provide access to random bits of explorer and other shell-related functions.

Re:Why isn't this already out?

[MrHanky]

Ah, the new Karma Whore scheme: Type a lot of mindless drivel that makes no fucking sense whatsoever, while saying OS X is better and X is slow, blah blah blah. Make some nonsensical suggestions for how to improve the situation.

Voila! Instant +4, interesting.

Slashdot is now officially News For Dorks: Brainless Apple Fans Pretend What They Say Matters.

No disrespect, ciroknight, but from a technical POV, you're talking crazy-talk. That's fine, but not worthy of a `5, interesting'.

Re:Why isn't this already out?

[The_Wilschon]

DON'T ask on the y-devel mailing list, "Is Y-Windows still alive?" The developers are very cranky, especially about such questions. Basically, the answer is "We're busy, and it doesn't seem as important to us anymore. So, if we get bored, we'll do alittle work." The actual reply however, is much more caustic. Here is the archive of a full discussion of the state of Y.

Also, there seems to be some awfully weird licensing kind of stuff going on there lately... The developers have been claiming that someone "stole" the source code to Y, and warning people against "pirated" versions... That sent all kinds of red flags up in my head. And here is the archive of the discussion of the "stolen" source code (see Y-windows update)

Re:Why isn't this already out?

[AKAImBatman]

Because you called output device hotplugging in X useless.

Ah. I was trying to be funny. As in "what is all this useless crap hooked up to my computer!" I didn't intend for my comment to be taken so seriously. :-)

The hardware stuff (the video drivers) are already partially in the kernel

Ehh... sort of. A "standard" X11 driver sits outside the kernel, but controls the device. The trick that NVidia has been pulling (which I think is true of ATI as well) is that they've been taking their Windows Driver code, compiling it to a Linux Driver, then writing a glue driver for X11 to load.

In short, all the groundwork has been laid, but the XFree86/X.org servers continue to think they should be controlling the hardware.

Another intersting point is that FreeBSD controls most of its input hardware (e.g. the mouse), and X11 merely talks to the BSD drivers. That's the way it *should* work, but that hasn't quite percolated through to the Linux guys yet.

Completely splitting X into a hardware support/renderer infrastructure part and a windowing system part sounds like a great idea to me, too. And then, making the renderer accept vector data and use the GPU to process it would be awesome. OS X awesome.

Glad you agree. :-)

Re:Why isn't this already out?

[vadim_t]

Whoops. Ignore AC post.
But what you're talking about has nothing to do with X.

If there's no X process, then what handles windows? GGI, at least from what I can see on its website is simply an API that provides access to low level functions, pretty much like SDL.

However, I don't see anything in GGI that indicates it has any ability to draw windows, widgets or anything remotely GUI like. If you want a thin layer that lets you draw stuff on the screen you can have it already: The framebuffer.

Enable support, boot with a VESA framebuffer console. Knoppix and many other distributions boot with a VESA console if you don't want to mess with the kernel config. Now write junk to /dev/fb0 to test it. It's currently there and it's usable. But it has nothing to do with windowing systems.

And again, network transparency can't be implemented on top of it because networking is the best IPC mechanism Unix has. It's that simple. If you want to avoid networking the only two options are: X in the kernel, which won't happen. Or application as a module in X, which won't happen either.

Something has to do the GUI stuff. You know, manage the different applications that have access to the same display. Send messages to other applications when appropiate. All of that requires communication. If all this stuff doesn't happen inside one unique process, then that means IPC. And good IPC means networking.

Re:Why isn't this already out?

[evilviper]

As a Mac OS X user, a previous Windows user, and a current Linux Desktop user, I will not be the first to tell you that X is slooow.

You are sure as hell NOT the first person to tell everyone this. Newbies yell it from their soapboxes on /. all the time...

Problem is, X is still, and always has been VERY fast. The things that are slow are GNOME and KDE. Use another window manager and you'll see just how fast it really is. Even a 100MHz machine makes a very responsive system with X11 and Blackbox/Afterstep/XFce/Enlightenment/IceWM/etc.

Of course, lack of knowledge of the subject won't stop you from telling everyone the solution to all their problems... You were doing the same thing yesterday, B.S. about dual-core Opterons and G5s you know nothing about...

Re:Why isn't this already out?

[man_of_mr_e]

Indeed. Not only do you have this you->kernel->x->kernel->you transition, but you also have the various attempts to optimize that (such as batching) which, while cutting down on the context switches drastically, also increases latency. It also makes it more dificult to do certain actions efficiently (like getting or setting the value of a single pixel).

X works a *LOT* better on a dual processor system, since X can be running on one processor and the current app on another, but I doubt various processor balancing algorithms keep this in mind.

I'm not convinced that this is really the problem though, since architectures like BeOS used a similar technique and were very fast. I just think the mechanisms X uses to marshal data suck donkey dick.

Not everything works well as a serial data stream, and that's what network transparency seems to enforce on X.

Re:Why isn't this already out?

[SpinJaunt]

yeah, you got it in one, why windows(r)?

Re:Why isn't this already out?

[evilviper]

Heck you can even continue to use most programs without even starting Explorer so long as you don't minimize them.

You can still minimize and restore tasks. Instead of going to the Explorer taskbar, they become basically shaded at the bottom of the screen. Explorer isn't really the window manager in the Unix sense of the term.

Re:Why isn't this already out?

[AKAImBatman]

Also, there seems to be some awfully weird licensing kind of stuff going on there lately... The developers have been claiming that someone "stole" the source code to Y, and warning people against "pirated" versions... That sent all kinds of red flags up in my head. And here is the archive of the discussion of the "stolen" source code (see Y-windows update)

Dude, check the date.

Re:Why isn't this already out?

[man_of_mr_e]

Maybe.. just maybe.. someone could start a "foundation" or something to collect donations to PAY nVidia and ATI to port quality drivers to Linux.

Then again, maybe that would start a bad precedent, and they'd expect a payment every time a new card was released. But then the alternative is putting up with crap drivers.

Re:Why isn't this already out?

[darkwhite]

didn't intend for my comment to be taken so seriously

You'd take it seriously too if you spent as much time as I did trying to get my laptop to dock properly :)

And it's still not working :(

Re:Why isn't this already out?

[man_of_mr_e]

One might argue that if X were properly designed then the choice of Window manager shouldn't slow down the performance of the windowing system.

Of course i'm sure you will disagree. It's like arguing that an application shouldn't be able to crash the OS.

Re:Why isn't this already out?

[NotoriousQ]

What X.org/XFree86 really need to do is separate the graphical and interface device layers from the desktop interface layer. i.e. It should be possible for any system program to be able to ask for exclusive video card access, even when the Desktop is not running. Currently, you have to chose between running X11 or using the external SuperVGA library.

What do you mean by this?

Do you mean separate the device drivers from the X11 libraries? Separate it from the window/keyboard events? Make it a render only environment? How do I not make the choice between X11 or SuperVGA? How can I run them both?

I am confused by what you are trying to say. The graphical system is a shared resource. X11 becomes the manager of that resource, and naturally does not want any other people trying to simultaneously manage it (SuperVGA).

You'll note that this is how the X server functions on systems like Sun Sparcs. On my UltraSparc, the system is ALWAYS in graphics mode.
So it is on linux, if you include the frame buffer drier in the kernel. In fact this is required for non-x86 machines.

Running X just makes the X-Server take over the graphical screen. Very modular, very flexible.
The same on linux, it simply switches the vt, hence the updates to framebuffer from a different vt are ignored. Thus X takes over the frame buffer, and the video card. Why is this modular? What modules are you even talking about? Frambuffers?

Not to mention that it places the graphics card support back in the OS drivers where it belongs, and not in a server with a focused purpose.
Sure, if you want to run the system with all the acceleration that a frame buffer can provide (hint:none). The correct way of doing this is closer to what the DRI and nvidia people are doing. Stuff the hardware specific things into the kernel driver, and pass a more machine independent calls via a device file into the user mode. Then in the user mode, more advanced processing can be done (driver specific gl implementation, driver specific video implemenation, etc). This should be , and is in the user mode of all major systems.

Remember, the Unix philosophy is to keep everything in simple, bite sized chunks. The X.org/XFree86 implementation is the anti-thesis of that (although they are attempting to compensate with their portable driver loader design).

This is the most sane statement of the above. The X.org is not very modular, but that is mostly due to their build process. What I would like to see is separate switchable plugins providing driver specific implementations of each of X11's extensions. That would make X11 a whole lot more modular.

Also the X11 base needs to be cleaned up a bit. There is X11 API that became very deprecated that all of it has been replaced by an extension. (The old X11 font handling comes to mind)

However, now that I have ranted a bit, I will say that I still do not get what you meant in your post.

Re:Why isn't this already out?

[teknomage1]

There is a kernel framebuffer module that can provide access to screen pixels directly but you would still need some sort of manager to prioritize which application could draw. I believe you can use fluxbox with the kernel framebuffer to achieve this.

Re:Why isn't this already out?

[evilviper]

One might argue that if X were properly designed then the choice of Window manager shouldn't slow down the performance of the windowing system.

I don't see how you can even make that argument. Having a slow window manager means you will have slow response from the windowing system, because it's up to the window manager to tell X11 what to do.

It's not a parallel track like a single program crashing the OS... It's a serial relationship, where it's up to one to pass info through to the other. It's like saying your cursor shouldn't jump around the screen, just because your mouse is dirty...

Re:Why isn't this already out?

[mpcooke3]

I'm not convinced it's X rendering that makes linux distros seems slower than windows. Applications and windows usually take longer to appear. For example gedit takes nearly as long as word to load on my PC and is about 10times longer than notepad.

After massive improvements and work to nautilus is about launches at a reasonable speed if you aren't using the file explorer version (which of course most people do once they've worked out how to re-enable it!). Maybe it's got something to do with library caching/toolkits being preloaded on windows, but whatever it is linux needs it.

Re:Why isn't this already out?

[fikx]

I guess my prefernce is diff from most people in that I'd like X (at least on Linux) to do more not less. I think it or a superset of it should be THE graphics API for everything, not just to create windows. we've got the window manager concept to actually "manage" the windows anyway
As far as "barfing" I'd rather the kernel provide the lowest level possible access to the vid card and let the X server actually manage the card. when the card locks or the driver code does, the machine keeps going underneath. From my experience it works that way now. X runs with above average privaledge and manages the vid card but crash leaves the machine breathing.

Re:Why isn't this already out?

[biglig2]

Say "Y windows" out loud. Good question, isn't it?

Re:Why isn't this already out?

[AKAImBatman]

Do you mean separate the device drivers from the X11 libraries? Separate it from the window/keyboard events?

No, not quite. Yes, the driver needs to be in ther kernel. But the API layer above that should provide exclusive access to the video card services. The X11 system can then build on top of that, but something else (e.g. a video game) can request that X11 temporarily relinquish control.

I am confused by what you are trying to say. The graphical system is a shared resource. X11 becomes the manager of that resource, and naturally does not want any other people trying to simultaneously manage it (SuperVGA).

You've just stated the problem. X11 is controlling the graphics system. That's a problem, because there's no way to do anything graphical *without* X11. This particularly sucks as X11 support for things as simple as changing the screen mode or reporting hardware capabilities, well, sucks.

The same on linux, it simply switches the vt, hence the updates to framebuffer from a different vt are ignored.

I'm going to ignore the lameness of this statement.

The correct way of doing this is closer to what the DRI and nvidia people are doing. Stuff the hardware specific things into the kernel driver, and pass a more machine independent calls via a device file into the user mode. Then in the user mode, more advanced processing can be done (driver specific gl implementation, driver specific video implemenation, etc).

Now you've got it! The trick is that we don't need X mucking things up at this level. X should be one step above. If X is directly managing hardware, then it's doing something wrong.

Re:Why isn't this already out?

[man_of_mr_e]

Well, the cursor jumping around the screen exposed a design flaw in the way analog mice worked. This was fixed in the "optical" variety with an entirely new design.

My point is, maybe the way window managers work is broken. Just because it gives you what you want doesn't mean it's the best way to do it.

Re:Why isn't this already out?

[misleb]

What X.org/XFree86 really need to do is separate the graphical and interface device layers from the desktop interface layer. i.e. It should be possible for any system program to be able to ask for exclusive video card access, even when the Desktop is not running. Currently, you have to chose between running X11 or using the external SuperVGA library.

Are you kidding me? Who even uses the SuperVGA libraries? I haven't used svgalib in YEARS. I don't see why I would want to. There is no hardware acceleration. Seems to me that X, more than any other graphical environment, splits the desktop from the graphics layer pretty good. And that seems to be the major complaint people have about it! That GNOME, for example, is not integrated with the graphics server. Who cares if you can do graphics outside of X?

You'll note that this is how the X server functions on systems like Sun Sparcs. On my UltraSparc, the system is ALWAYS in graphics mode. Running X just makes the X-Server take over the graphical screen. Very modular, very flexible. Not to mention that it places the graphics card support back in the OS drivers where it belongs, and not in a server with a focused purpose. Remember, the Unix philosophy is to keep everything in simple, bite sized chunks.

What use is the graphics mode on a Sun outside of X? I'm not a regular Sun user, so maybe I am missing some amazing abilities, but last I checked all it does is make scrolling text on the console really fucking slow. What's more annoying is when you hit an invalid key and the console "blinks." I swear, you can see it paint the screen.

Really, if you just want a bare graphics console with no desktop, just run X without starting any window manager or use twm or something. Sheesh.

-matthew

Re:Why isn't this already out?

[justsomebody]

The largest open source project ever seems to be working fine; Linux itself.

Linux is just kernel. 5mio+ lines of source. Both X and openoffice are bigger. But if you think Linux as distribution this is not one OSS project, but collection of many.

My entire argument comes down to this; X is old. All code needs to die sometime, and X is a perfect candidate for being put out to pasture, and a good teaching example on how a windowing system can work, due to its simplicity. But we're in a world now with needs beyond what X can deliver, and we're at a point in which we're still dragging it along. Meanwhile Mac OS X has managed to implement an entire drawing engine, add nice hardware accelleration features (Quartz Extreme), make them useful (Exposé), and add to their usability once again (Dashboard). Windows "Avalon" (if it ever gets done) will add to the mix SVG interface definitions (I wouldn't doubt if Apple doesn't come out first with SVG-window rendering capabilities).

Code being old is not reason to die. Reason to die would be if someone introduced better solution that would make more benefit than lost time would mean for the project. ...But we're in a world now with needs beyond what X can deliver, and we're at a point in which we're still dragging it along.

What would that be?
Expose? nope. As you see Xgl extension is more than capable for this.
Quartz? nope, cairo, glitz and poppler are more than capable replacement.
Dashboard? gdesklets and konfabulator are here now for a long time but I still haven't found practical use besides eye-candy.
Avalon? nope, xforms being w3c standard is the way to go.
Spotlight? nope, you have beagle
Apple first in SVG? No, it wont. waimea already did that. cairo and glitz as system parts are following its lead

Just because it is one body doesn't mean that it can't have sub-bodies. Compare with the Linux kernel; broken down into archetectures.

Linux distro being about 40 to 50 mio lines or more. Who would maintain it? You? You would need one central figure or organization that is coordinating everything (and remember that this is not one way project, you've got office, x, desktop, kernel, networktools, cli utils...). And that would be a huge task for OSS.

Lastly, we need to do something we really haven't done with anything up to now (well, Mozilla and GNOME/KDE excluded; they've been really good at this in the past). We need to _design_ and _engineer_ the software, not just code it. Of course, engineers expect to be paid (even I have came to this conclusion after my last project, Charity), which is another reason open source software isn't innovating as much as we should hope.

Well, secret is simple. Write free for free and commercial for commercial. I always write at least two projects, from which one is OSS. From the other I either count to be paid in time (writing software for maintaing my servers, and this means that this one is OSS too) or in money (if there is a customer who ordered it, this is not OSS). I always try to help patching projects (that one is for free, because in the end I benefit if software works better from default)

I never thought OSS as charity, I benefit more than enough money from it, in fact last year my job was more or less based on OSS.

Re:Why isn't this already out?

[Veshtaj]

It's not like I wrote X-Windows. I just commented that it's gotten a lot better than it used to be.

That reminds of a joke (or observation?) floating around back in the 1990s:

X-Windows: Faster than it used to be. This is not hard.

Re:Why isn't this already out?

[steeviant]

1. The icon you're hovering over expands to help you see what you're clicking on and hit the target correctly. (Mac OS X Dock.)

I'm a Mac OS X user, and have been for the past 3 and half years, and though I have always liked the dock, I despise the icon zooming feature not because it's flashy or eats CPU/GPU, but becaue it's awful from a usability standpoint. Things that you are trying to click on should not move away from, then toward your mouse as you get closer.

The idea of the entire dock expanding and contracting, and having different sized icons is hideous. Drop/click targets should not move as you're trying to click on them. Yuck.

2. Windows can be flipped over so that notes can be placed for later reference. (Sun Looking Glass)

This could be achieved using other methods. Programs have been popping up windows to allow you to edit preferences and other data for quite some time now. There's no reason that a sheet couldn't drop down from the window title bar or pop up from pressing a button. That feature of looking glass stemmed from the developer desperately trying to find a use for a new technology rather than it being a needed feature looking for a good implimentation. Not to mention the fact that the flipping effect could easily have been simulated in a 2d environment.

3. Objects on a table-screen can be moved by dragging your finger, rotated by moving a second finger outside of the object being dragged, and copied by dragging two fingers around the object(s) to highlight an area. (See here)

And while we're on the subject of hardware that no-one has yet, why don't we just use a neuro-interface to transmit the computer's interface directly into the users brain using our 50Ghz 20 CPU superduperwoopercomputer. :D

Re:Why isn't this already out?

[Zphbeeblbrox]

Didn't say they were servers. But they do host things. They load DLL's and applications hide behind them. I know cause I've seen it. Certain applications load through the svchost. You don't see the application though you just see svchost.exe. Its really annoying when your hunting that elusive piece of adware or virus.

Re:Why isn't this already out?

[Anonymous Coward]

Erm, the obvious difference is that the GUI subsystem in Windows runs in kernel mode, so applications don't use IPC to make graphics-related calls, they use a system call interface. Obviously this entails the overhead of transitioning from the client process to kernel mode when a batch of commands is sent from the client, but the kernel-mode GUI subsystem is then able to directly interact with the hardware. With a client/server design, on the other hand, doing the same thing requires a transition from the client process to kernel mode, then from kernel mode to the server process and then from the server process back to kernel mode. It's patently obvious that the latter model is far less efficient.

Re:Why isn't this already out?

[cahiha]

I don't know much about the guts of X and windowing in general, but what exactly is the benefit of having windowing in an external process?

It makes things work better. Microsoft thought they could get away without it and regretted it. Apple thought they could get away without it and regretted it.

Are interoperability and backwards compatibility the principal reasons for X?

No, the principal reasons are that the design works really well and has stood the test of time. The designs you propose were the designs people started out with 30 years ago.

Personally, I don't care about the network transparency of X - I'd prefer the remote capabilities to be built on top of the existing drawing system as remote use seems to me

You don't pay for it, so don't worry about it.

Re:Why isn't this already out?

[rookworm]

Yeah, but that doesn't make sense ("why windows in the works"-- "what windows in the what??!!?"), and if it did, it wouldn't be funny.

Re:Why isn't this already out?

[njh]

Have you looked at the Cairo-graphics project? (cairographics.org) Everyone is porting their apps and DEs to it. It will give Quartz a run for its money (probably not core image for a while, but it will happen if core image turns out to be actually useful).

Gtk+ already runs on cairo, btw. I believe that gnome 2.12 or whatever is planned to run on cairo.

It does happen, it's just that sometimes it takes an external example to motivate free software people to do it (such as MacOSX and Avalon). This isn't necessarily a bad thing.

Re:Why isn't this already out?

[AKAImBatman]

I despise the icon zooming feature not because it's flashy or eats CPU/GPU, but becaue it's awful from a usability standpoint.

I'll never understand why people don't like the dock. I think the feature is excellent as it allows an otherwise cluttered area to become useful. Just think of Windows. How many times do you have no idea where the heck your program is because all of the taskbar buttons look like "[tiny icon]..."? And I use a double-tall taskbar!

In any case, you can turn the effect off. Just adjust the dock zoom setting to zero. You can also set the initial size of the dock itself.

This could be achieved using other methods. Programs have been popping up windows to allow you to edit preferences and other data for quite some time now. There's no reason that a sheet couldn't drop down from the window title bar or pop up from pressing a button.

I was wondering if someone was going to mention that. Thinking like a Mac user, eh? :-)

The reason why the notes on the back of the window make more sense, is that it is a far more natural position for them to be. e.g. When you look at a movie or game in the store, what the first thing you do after you pick it up? That's right, look on the back!

The alternative methphor would be akin to having a special pocket on every product that people are somehow trained to check. Not very natural as things stand today.

And while we're on the subject of hardware that no-one has yet, why don't we just use a neuro-interface to transmit the computer's interface directly into the users brain using our 50Ghz 20 CPU superduperwoopercomputer. :D

Ok, this is why we're on the subject of UI Research. The tabletop interface may not be commercially available yet (otherwise it wouldn't be research), but it is something that exists today and can be viewed in the video.

Re:Why isn't this already out?

[NotoriousQ]

The X11 system can then build on top of that, but something else (e.g. a video game) can request that X11 temporarily relinquish control.

Ok. So what you are basically looking for is an api for X to stop sending stuff to the video processing code, and let the programs do it directly. Hmm..there is some point to that statement, but then these video processing drivers will need to have a way to translate things like xvideo->video card specific calls. In other words, all the drivers will need to present the same interface.

To some extent this is already done in X. The only thing that X offers on top of that are windowing primitives, and a level of indirection for the desktop (great for xinerama, virtual desktop, etc).

You've just stated the problem. X11 is controlling the graphics system. That's a problem, because there's no way to do anything graphical *without* X11. This particularly sucks as X11 support for things as simple as changing the screen mode or reporting hardware capabilities, well, sucks.

I will to some extent disagree here. X's fixed resolution/desktop size issues do suck (although XRANDR pretty much fixed most resolution based annoyances), but reliquishing all control does not help at all. The only thing that separating the hardware specific parts from X will get you is an ibility to use those drivers outside of X. I do not see how that would suddenly help the situation.

Now you've got it! The trick is that we don't need X mucking things up at this level. X should be one step above. If X is directly managing hardware, then it's doing something wrong.

The thing is, X is not mucking around that level. The driver part of X is mucking around the hardware level. The rest of X is completely hardware independent.

So basically what you are suggesting is to move the X video drivers out of X, and make them separate .so with a more or less standardized interface based on X extensions. In other words if your video card supports video capabilities, then the driver will provide XVIDEO. If it has GL support, then it will provide GLX, etc. All of these interfaces should follow some spec.

Do note however, that this will probably not be kernel code. A lot of capabilities will probably require some emulation, as no card will support everything in GPU. And having that much code in the kernel is scary.

To some extent I like the idea. That way I can use the video card API with whatever app. There are a few tricky things. First, there is a similar issue as that encountered by the terminal: who owns the screen. This is the main purpose of X. X owns the screen, partitioning it to other applications. If you do not have a manager, issues will become apparent quickly. Although I will find it humorous to receive a TSTOP in response to a GL call.

The hacker in me absolutely loves the idea. I would be able to write GL programs on the console, and will get my screen by simply using termio. However, the realist in me suggests that 99% of the time you want a manager anyway. Even if I am playing a full size game, there are programs that need to be able to grab the screen and display critical messages. And in that case the drivers that are in X will only be directly used by X, their functionality multiplexed among applications, and hence there is no real need to have the driver elsewhere. (Of course this assumes you will use X as the only manager, but if not you can simply move the driver to the other manager).

Basically, I just want to say that what you are complaining about will not bring about wonderful new functionality, but will simplify the architecture, and make it less tied to X (assuming video card capabilities can be standardized to one API).

Re:Why isn't this already out?

[Ian Bicking]

If there's no X process, then what handles windows? GGI, at least from what I can see on its website is simply an API that provides access to low level functions, pretty much like SDL.

GGI isn't a replacement for X, but is an example of that architecture applied to a subset of what X does. GGI has multiple backends. The GGI API is translated to the backend differently depending on the nature of the backend. For instance, some hardware implements some high-level rendering in the hardware. For other hardware some of that rendering may have to be done in software. A network backend, for instance, would turn those API calls into some form of communication and send it over the net.

Under GGI that would be, at best, a VNC style of network transparency, and you'd get a rectangle on the other end of the connection. But GGI isn't a replacement for X; if you were to implement X this way then you'd have multiple xlib implementations which forms a higher level API than GGI, with concepts of windows and whatnot.

If you want a thin layer that lets you draw stuff on the screen you can have it already: The framebuffer.

Funny you should mention it. The problem with the framebuffer is that the API is bound to the implementation. It's a framebuffer; but that implementation doesn't fully describe the hardware available. GGI has a higher level API, and so the GGI API can be translated into commands to the graphics card that take better advantage of the card's ability. Again, this isn't an X replacement, but an example of an architectural style (which I'm asserting is superior to X's client/server style).

Though you were right when you said that X is moving in the direction of an API. Something like the Render extension is the style I'm talking about (at least from my understanding of how such extensions work). It's primarily an API, not a protocol. It can be implemented in terms of the stable X protocol, but to be efficient that is usually shortcutted, using some protocol I know nothing about and probably isn't standard. (Yes, it does it with shared memory or somesuch -- but shared memory isn't a "protocol"; you need to define what you put in that shared memory, not just where it is, and AFAIK these extensions to not try to define that what in any formal way.)

Re:Why isn't this already out?

[Fulg]

The graphics engine needed for a game [...] almost requires a kernel bypass to write directly to the video hardware.

These days are long gone. Windows has been proving it with blazingly fast GDI, DirectX and OpenGL apps that are completely hardware-independent, yet 100% hardware accelerated. Hell, even current (and future) game consoles don't allow you to poke directly at the graphics hardware anymore (PlayStation2 excluded).

Ugh, can't say I miss programming VGA registers by hand and dealing with all the semi-compliant hardware... *shudder*

Re:Why isn't this already out?

[cahiha]

One might argue that if X were properly designed then the choice of Window manager shouldn't slow down the performance of the windowing system.

And it doesn't. What's your point? Do you even know what a window manager is?

Re:Why isn't this already out?

[bankman]

While at times I've been firmly in the "There needs to be an X with less crap in it" camp

What crap? And why would anyone want to remove network transparency?

Re:Why isn't this already out?

[cahiha]

Well actually, yes it does. You still have to marshal pixmaps (a wretched and primitive yet still bloated format) into a shared memory segment just so the server can pull them out of there and transfer them to the graphics driver.

How the hell do you think this works on, say, Macintosh or Windows? Magic elves make the bits appear on your screen? X11 gives you the same choice as any other system: you can either let the server do a copy-conversion, or you can use direct rendering. Most sane people will choose letting the server do the copy for most applications.

And X's implementation of network transparency doesn't give clients any way to tell the server to aggregate events or even tell it to "shut the hell up already" with all the mouseover events over regions where they're not listening to the events.

Huh? You have detailed control over what regions you get what events for. But what's the point anyway these days? Mouse events are not a performance bottleneck on X11.

Network transparency is a good thing. X's implementation of it stinks.

First of all, X11 isn't an "implementation", it's a protocol

Secondly, implementations of X11, like XFree86, do pretty much the same thing Macintosh or Windows do: they have a graphics server process that's separate from the application, and they use IPC to talk to it.

The only thing that "stinks" is that people like you who evidently have no idea what's going on opine about X11's design.

Re:Why isn't this already out?

[cahiha]

These aren't necessarily opposing opinions: the graphics system is different than a kernel. But in a GUI environment, both are going to be responsible for a *lot* of communication, so I find it intriguing that people are able to hold both opinions at once.

What can I say? The design of the Linux kernel sucks. Tanenbaum was right. But the Linux kernel works, so why worry about it? The fact that it isn't a microkernel is largely Linus's headache, and as long as he can live with it, nobody is going to worry about it.

And what is happening with X11 and Linux is the same that's happening on Windows and Macintosh: all of those have a graphics server process and a monolithic kernel. Microsoft and Apple came to those designs after trying a lot of other possibilities, so there is probably a reason for it.

Re:Why isn't this already out?

[vadim_t]

Aha, I think I'm starting to understand the way you think at last. But I still don't see the point.

So fine, GGI has the API concept you like so much, and provides an uniform interface to anything. It's modular, cool and shiny. Now, how do you suggest to make a GUI environment following the same system?

The thing you seem to be missing is that X, or its replacement can't be done this way. So okay, we'll try to do it the Windows way. We'll come up with a nice API so that we call CreateWindow, and a window is made. Now, how does that actually happen?

Well, something needs to make that window. It can't be the application. In the case of GGI/SDL it's simply a problem of translating that high level command into pixels on the screen. But that won't work, you need to be able to run several applications on the same desktop. Just letting several processes call functions and draw stuff where they please is only going to create a huge mess or something useful.

So, how do we solve this? We need an arbitrator, which will assign and manage resources. It will ensure whatever is drawn is drawn in the right order. It will keep track of who owns what, and what is where. It will send messages when something of interested happens, like a paint event.

How, this can't go inside this magical library. How would you decide who is in control? That won't work. The arbitrator must be its own entity and dictate the rules to all who use the services it provides. One example of a system like this is the kernel. A program can't just take a 64K chunk of memory starting at 0x80abcdef, hell, no. The program instead calls the kernel and asks: "I want 64K of memory", and the kernel makes sure there's enough of it, that there's some location big enough, and that the program is allowed to have that much. Then, it adds it at the end of the current process, marks it as used, etc.

Same here. Now, who would arbitrate this stuff? Well, one possibility is the kernel. However, as I mentioned already, this won't happen. The kernel developers don't want it and with good reason. They say that if it can go in user space then it will have to go in user space. And since you can perfectly do all this stuff in user space that's where it will be.

Now that we decided it will be in user space, we know it will run as a process. This process manages resources used by other programs, which also are processes. All of this means both the manager and the users need to talk. And how do we do that?

Well, we can pick: Network sockets, unix sockets, shared memory, pipes, queues. I'll leave out queues because I'm not that familiar with them. Pipes are unidirectional. You can get the same effect, just better with unix sockets. Shared memory is great for sharing data but not for implementing protocols. In a complex event based system you need to be able to poke things and to be poked, and networking does this a lot better.

So, we arrive at using networking yet again! Amazing, isn't it?

Re:Why isn't this already out?

[cahiha]

I could be talking out of my ass here, but I would expect that what one would REALLY want is abstration at the DISPATCHING level... use some mechanism to dispatch the communication between the client and server, then an OS with no need for networking could use X by replacing the implementation with

X11 doesn't require networking; it runs fine over non-network IPC mechanisms and is commonly run that way.

Re:Why isn't this already out?

[NotoriousQ]

You got me more interested in the issue, and I am currently reading a great article on this topic written by keith packard So unless you are Keith Packard, or have read it already, I highly recommend it.

Re:Why isn't this already out?

[cahiha]

but there's no two ways about it: the X server is too old and archaic for modern interfaces.

How do you think Windows or Macintosh work? All major GUIs use an X11-like architecture these days, with separate server and client processes talking via IPC.

Re:Why isn't this already out?

[vadim_t]

Uh... and why would you want something as messy as that?

I'll say it again - Unix sockets are one of the basic and most used IPC methods in Unix systems. They do NOT have any noticeable performance impact!

Doing some strange crap with memory wouldn't give you any better performance. Besides it sounds more complicated and a lot less secure.

Re:Why isn't this already out?

[AKAImBatman]

Ok. So what you are basically looking for is an api for X to stop sending stuff to the video processing code, and let the programs do it directly. Hmm..there is some point to that statement, but then these video processing drivers will need to have a way to translate things like xvideo->video card specific calls. In other words, all the drivers will need to present the same interface.

This is mostly true. The exception is that extensions like XVideo would be potentially working with both the X-Windows APIs (I'm about to use this area of the screen here) and separate video decompression drivers (ok graphics card, X11 says I can stick the video here). If support is missing in some area, then you simply couldn't install the extension.

To some extent this is already done in X. The only thing that X offers on top of that are windowing primitives, and a level of indirection for the desktop (great for xinerama, virtual desktop, etc).

Well, that's the problem. X combines the two concepts into one. Look at how Windows handles it:

Video Drivers -> HAL -> GDI -> Win32 -> MFC

The clear layers of separation help keep everything in line and ensure that a given area doesn't overstep its bounds. In the X world, everything from part of Win32 on down is mixed up in one big glob.

The only thing that separating the hardware specific parts from X will get you is an ibility to use those drivers outside of X. I do not see how that would suddenly help the situation.

It would help from a focus perspective. If the drivers were moved to the OS level, then the focus would be on defining *good* interfaces for things like changing the screen mode and giving device caps. Right now, no one can fix those issues because they're all tied up in other areas of the server. Once you move hardware out of the domain of the X server, the code suddenly becomes leaner, and FAR more portable.

The thing is, X is not mucking around that level. The driver part of X is mucking around the hardware level. The rest of X is completely hardware independent.

Which is again my point. Why is X managing hardware? That's the job of the OS.

So basically what you are suggesting is to move the X video drivers out of X, and make them separate .so with a more or less standardized interface based on X extensions. In other words if your video card supports video capabilities, then the driver will provide XVIDEO. If it has GL support, then it will provide GLX, etc. All of these interfaces should follow some spec.

Bingo. :-)

Do note however, that this will probably not be kernel code. A lot of capabilities will probably require some emulation, as no card will support everything in GPU. And having that much code in the kernel is scary.

Of course not. NVidia's drivers aren't part of the kernel on Windows or Linux. They're kernel mode drivers. The higher level APIs would of course run in usermode, but the option could be later added for kernel mode for the performance junkies. (e.g. Solaris provides this option for high performance workstations.)

There are a few tricky things. First, there is a similar issue as that encountered by the terminal: who owns the screen.

That's easy. Who owns the screen now? Whoever last requested it, or whoever the user has forced. If I run X, does X not grab the screen? But what happens if I hit CTRL+F1? Now who owns the screen? This is the same thing, with the exception that more than one graphical TTY can exist.

However, the realist in me suggests that 99% of the time you want a manager anyway.

Even if it were 100% of the time, I think it still makes sense to split these two components. The graphics driver and the windowing system are two separate layers that shouldn't be forced together. All one has to do to convince themselves is run an X server such as WeirdX *inside* an X server such as X.org.

Re:Why isn't this already out?

[jericho4.0]

You bring up a good point, which brings up another good point.

The design of X and Linux might not be the best of all possible designs, but we use them because they work, and nothing else does. For all the bitching about X's network overhead, I, for one, could not live without it.

Re:Why isn't this already out?

[man_of_mr_e]

I was responding to someone that claimed that the Window manager is what slowed X down. What I should have said was "If that's true..." I don't know if it's true or not, but you seem to be certain. Many people in this thread have claimed that running a light weight window manager made X fast.

Re:Why isn't this already out?

[Tore S B]

I want to see a distro with actual no-shit autoconfiguration of all relevant devices.

Ubuntu gets close.

Re:Why isn't this already out?

[AKAImBatman]

Nice link. Packard basically blames all of X's woes on the architectural problems I was describing, and then backs up his statement with common examples such as display hotplugging. (A salient point that I was taking undeserved flak for from another poster in this topic.)

So, does everyone agree with me now? ;-) (Me? Big head? Never.)

Re:Why isn't this already out?

[vsync64]

Now tell me what each of those svchost processes is actually hosting? Not easy is it.

I've not found tasklist /svc to be a hardship, myself...

Re:Why isn't this already out?

[NotoriousQ]

People were argueing with you about X11's display hotplugging. I thought everyone knew it sucked. XRANDR is great for a single display, but restarting X (or starting a new one) is annoying, and should be properly handles by xinerama (or something similar) IMO.

However the bigger problems are that the driver support still sucks.

But here is some more troll-candy :) for you. atitvout.

The reason I said troll-candy is due to that it both helps your and my comments at the same time. It is an application that goes over X to switch a mode in a video card, without X realizing the problem. Unfortunately X sometimes reacts poorly to the change.

Your (would-be) point: if the driver provided the right functionality then any program should be able to switch displays.

My point: the ability to switch displays should only be allowed by the driver which currently owns the terminal. That means if X is running, X is the only thing that is allowed to make the call.

So back to our discussion: So suppose we split the x-server into the hardware controlling part and a desktop part. How many assumptions must the desktop part make about the hardware part.

Think the current terminal nightmare: application crashes leaving terminal in raw mode, application now in control of the terminal does not see the problem causing the terminal to become useless.

This same issue would be the case in your new design. How do you think one should deal with that?

PS> Agree with you? I did not know we were arguing? (In fact, once I understood your point, I wholehearedly agreed, with the exception of this whole bypassing the screen manager business).

Re:Why isn't this already out?

[vsync64]

Additionally, switch from the initial craptastic XP theme to the "standard" theme (2k style), and the RAM usage will just plain drop.

Empirically, I and coworkers have found that skipping that step and simply setting the themes service to "disabled" provides the most noticable improvement.

Easiest way to get Explorer not to suck CPU is set your performance options (I think in "My Computer" somewhere) to "Best Performance". Then go to Explorer and turn status bars back on.

Re:Why isn't this already out?

[aichpvee]

I'm somehow doubting you've ever used window shading yourself and are just fanboying off the looking glass demo. Nothing that I've seen from them is anything more than eye candy which has been done by by other people (translucency, video icon previews, etc) and less than useless "features" such as flipping a window around to write a note on the back (both useless "feature" and useless eye candy) and their bastardization of window shade.

Putting the window to the side doesn't take it out of the "primary work area" and negates the purpose of window shade in the first place, which is to quickly get a window out of the way.

The current window shade used with drop to back and multiple desktops is already so much more useful than this crap that sun is trying to sell that it's pathetic.

Re:Why isn't this already out?

[AKAImBatman]

People were argueing with you about X11's display hotplugging. I thought everyone knew it sucked.

Oh, they know it sucks. I was getting *blamed* for it because of a comment I made about X's improved ability to handle pluggable hardware. Yeash, people can be testy about this stuff.

However the bigger problems are that the driver support still sucks.

But... it's something that could more easily be focused on if the driver was separated from X.

It is an application that goes over X to switch a mode in a video card, without X realizing the problem. Unfortunately X sometimes reacts poorly to the change.

Well, there's a few points on this front:

1. A lower level driver would have to define event for when screen mode and geometry changes happen. Any X server built on top should listen for those changes and pass the information down to the root window to do something about. (Presumably, the root window is the Window Manager, which would then take care of moving any windows that need to be moved.)

2. Applications shouldn't be changing the screen mode of X. Remember my point about changing virtual terminals? Any program that wants to have exclusive access to the screen should be in a separate TTY. That way the screen mode issues don't matter, because the mode will be restored when the X server is active. This gets into your next point of:

My point: the ability to switch displays should only be allowed by the driver which currently owns the terminal. That means if X is running, X is the only thing that is allowed to make the call.

Which is fine. X should control its tty with an iron fist. But Quake III should also be able to control its TTY with an iron fist. Since both can't occupy the screen at the same time (which is what actually happens today), there is no issue.

Think the current terminal nightmare: application crashes leaving terminal in raw mode, application now in control of the terminal does not see the problem causing the terminal to become useless.

That was one of Packard's points in his article. If the OS controlled the graphics hardware, then it would be able to detect this situation. Once it detects that the program has crashed, it closes the TTY and pops the next one off the stack. Life continues on.

PS> Agree with you? I did not know we were arguing?

Sorry, just me being flippant again. I thought the whole big head comment would tip you off. :-)

Re:Why isn't this already out?

[NotoriousQ]

Sorry, just me being flippant again. I thought the whole big head comment would tip you off. :-)

It did. I just could not express my sarcasm in a sentence form. Verbally that could be done using intonation, but no such luxury in slashdot posts.

Just to stuff in one last comment, the whole graphical tty business is cool, but I still think that a single manager is preferable due to the critical notification. No application should ever look for which TTY is active, and then try to find the owning application to display a warning to you. I will still believe that one single manager is the right way.

And I am waiting till they properly split the codebase. This is the reason why I am watching kdrive very closely, as it has the potential to become the properly architectured X11.

Re:Why isn't this already out?

[AKAImBatman]

Just to stuff in one last comment, the whole graphical tty business is cool, but I still think that a single manager is preferable due to the critical notification.

I'm still not clear on what sort of critical notification should interrupt a game. Unless the system is about to blow up (the infamous "the processor is on fire" interrupt ;-)), there's really nothing that won't wait for you to exit your game (or hit the key combo to switch the screen back to check). The only thing that *might* be important enough (and I stress *might*) would be an OS/X11 level event. In that case you don't need to check, just tell the OS "I want to be in charge." A bit like setting the active window in Windows.

This is the reason why I am watching kdrive very closely, as it has the potential to become the properly architectured X11.

True 'dat. :-)

Re:Why isn't this already out?

[misleb]

From my experience it works that way now. X runs with above average privaledge and manages the vid card but crash leaves the machine breathing.

Apparently you don't own an NVIDIA card. Man, the binary driver is like 2.5MB in RAM. And trust me, when X freezes, it takes down the machine. But this has only happened with particular driver versions and settings. In general it is solid.

-matthew

Re:Why isn't this already out?

[NotoriousQ]

I'm still not clear on what sort of critical notification should interrupt a game.

An IM from a girl I asked out. ;-)

Re:Why isn't this already out?

[Brandybuck]

When you're using DirectX, you *ARE* dealing direct with the hardware.

Re:Why isn't this already out?

[runderwo]

A process calling into the kernel *is* a context switch, because the "context" is defined by the combination of the processor state (registers) and activation record (the stack). When a system call is made, the GP registers are saved, and the stack is switched away from the process stack to a kernel stack. In fact, the only bit of context that isn't switched on a system call (at least on IA32) is the FP regs, and this is a Linux convention.

At least one other architecture (SPARC I think) has this nifty construction called register windows, where to save registers for a context switch, only one instruction needs to be executed (to change the register window pointer) rather than one instruction per register.

Re:Why isn't this already out?

[Brandybuck]

it's just like networking, file systems, et al.

Apples and oranges. You put stuff in the kernel because you need direct hardware access, not because you want speed. The only place where you're worried about speed in the kernel is when you're handling interrupts. And GUI interrupts are so seldom and so non-urgent that it makes no sense to handle them in the kernel.

If you don't need direct hardware access, the only speed hit you're going to have in userland is context switching, but even that is so insignificant it's not worth worrying about unless you're running a 33MHz 30386.

You'll get more performance improvements if Nvidia and ATI released complete specs than if you dropped network transparency shoved the X11 into the kernel.

Re:Why isn't this already out?

[AKAImBatman]

An IM from a girl I asked out. ;-)

LOL! :-D

Well yes, that should interrupt a game, but I'm afraid that it won't under pretty much any system. You have to listen for the IM sound to get played through the mixer, then pause and minimize your game. Nothing different from what I'm proposing.

Re:Why isn't this already out?

[runderwo]

they can do so by creating another, entirely separate Postscript-generating layer (but without having to worry about performance or interactivity or other non-print Linux related things).

This already exists.

Re:Why isn't this already out?

[NotoriousQ]

It does now. I have gaim set up to popup windows only when certain people wend them, but not for others. It pops up right over ut2004.

Re:Why isn't this already out?

[AKAImBatman]

Ok, technically it's not supposed to do that. The game should have complete control over the screen and the GDI should be detached. Sounds like you're exploiting a bug. :-)

(And yes, it is a bug. Specifically an undesirable side effect of sharing the same screen space between the GDI and DirectX subsystems. Microsoft's literature tends to caution against accidental GDI calls when in DirectX modes.)

Re:Why isn't this already out?

[evilviper]

Well, the cursor jumping around the screen exposed a design flaw in the way analog mice worked. This was fixed in the "optical" variety with an entirely new design.

Analog mice work just fine. Optical mice have just as many flaws as their predicessors.

maybe the way window managers work is broken. Just because it gives you what you want doesn't mean it's the best way to do it.

You're being so vague about it, I can hardly argue...

There really is nothing wrong with the way window managers work, it's a matter of junk-in, junk-out. If you window manager, which you depend upon, is slow, bloated, buggy, etc., you have a problem. Just as it's bad if you are depending on some other poorly designed piece of software. Somehow, X11 gets all the blame.

Re:Why isn't this already out?

[NotoriousQ]

It is not a bug. The game does not have a complete control over the screen. The manager has that. If I configure the manager to put up a window over the screen, I surely want that done.

BTW, this is not in windows. This is on X11/linux, and X11 allows popping windows over the top fullscreen window.

And there are no side effects, save for blitting delay.

Re:Why isn't this already out?

[AKAImBatman]

BTW, this is not in windows. This is on X11/linux, and X11 allows popping windows over the top fullscreen window.

Ah, ok. In Windows that's a bug. In X11, well, technically it isn't designed for playing games in the first place. :-) However, I thought that games were supposed to set the "Always on top" flag. Granted, the window manager could ignore it, but the feature does exist and is supposed to be honored.

Re:Why isn't this already out?

[kidlinux]

"Enlightenment looks like its usual "prrreeeeetttyyy" self"

You've completely overlooked the major accomplishments the Enlightenment Project has made. Yeah, it's gorgeous (is that a bad thing?), but the achievment is the Enlightenment Foundation Libraries that make it all possible.

The design that went into the libraries makes it easy to design an app that's functional, visually pleasing, and well integrated with the rest of the desktop and applications.

You've really got to visit enlightenment.org to understand how well designed the system is.

Yeah, they've created a pretty interface, but they've got the technology to back it up.

Re:Why isn't this already out?

[NotoriousQ]

I believe that UT2004, just pops itself up and sets a fullscreen flag.

But even if it did say "always on top", all that a popup has to do is to have "always on top" as well. I do not actually know how the stay on top contention is restored, and unless defined in ICCCM, diferent windowmanagers will handle it differently.

Re:Why isn't this already out?

[NotoriousQ]

What's more annoying is when you hit an invalid key and the console "blinks." I swear, you can see it paint the screen.

This effect is probably solely responsible for the origin of the infamous "slowaris".

Re:Why isn't this already out?

[lemody]

> ... doesn't give clients any way to tell the
server to aggregate events or even tell it to
"shut the hell up already" with all the
mouseover events over regions where they're
not listening to the events.

a window receives only those events that are in it's 'event_mask'. this mask can be changed on the fly by calling XChangeWindowAttributes().

Re:Why isn't this already out?

[Mancat]

Take a fast car.

Skinny guy drives fast car. Car goes fast.

Massively fat guy on the brink of ten thousand simultaneous heart attacks drives car. Car goes slower.

Enjoy the example.

Re:Why isn't this already out?

[Mancat]

*sigh*

X is not broken. It has been ahead of its time for many years, and continues to be. Name another display system that is modular, cross-platform interoperable, and network transparent. You can't.

Please stop this useless, redundant troll.

Re:Why isn't this already out?

[Fulg]

When you're using DirectX, you *ARE* dealing direct with the hardware.

No, you are not. You have been (successfully) misled by the marketing drones at Microsoft :)

Waaaay back in DirectX 1.0 (when the only useful component was DirectDraw) there was a fast path leading directly to the hardware framebuffer, but this is long gone.

I write games for a living (and prior to that I wrote DX and GL drivers for Windows) so I can assure you, at the application level, DX is not letting you in on some hardware action, everything is abstracted to a common layer (the DirectX runtime). Only the driver, supplied by the hardware vendor, is able to talk to the hardware.

Re:Why isn't this already out?

[cahiha]

For version 4.0, released in 1996, the graphics components were moved from a user-mode module to a kernel-mode module

True, but the point is: it is running in a different context from the application.

Moving the graphics components from a server process to a kernel module was a good trade-off on Windows

No, it wasn't a "good tradeoff", it was a way of partially mitigating the consequences of trying to continue to hold on to an obsolete and poorly designed graphics API in Windows. Microsoft needed X like client/server separation, but they either didn't even realize what they were doing, or they weren't willing to do the engineering necessary to design a decent protocol and create a new API to do it. And Microsoft has paid a heavy price for it in increased kernel complexity and reduced kernel stability.

Moving X into a kernel-mode module on Linux would be a much less favourable trade-off because,

Moving X into a kernel module wouldn't help you because X was designed for client/server operations from the ground up. You don't gain any significant amount of efficiency by moving the X server into the kernel. In addition, the Linux kernel appears to handle communications between user processes more efficiently than the NT kernel as well, further reducing the need to put anything into the kernel relative to NT. That's why the question has never come up.

Re:Why isn't this already out?

[billcopc]

Yep. This is why WinXP is almost easy to use. You just plug in whatever and 10 seconds later it's installed. I just did that with an old gamepad I had lying around, didn't need to download a driver, didn't need to say "Next->Next->Finish", just plug in and go.

Some Linux distros have some infant form of autodetection, but it's mostly used at boot time and nowhere else. It's not like I can hot-plug NICs anyway, but it's still a bitch having to find the right driver for it sometimes. You'd think the PCI id would be enough to pick the right driver without a single user prompt.

Re:Why isn't this already out?

[cahiha]

If you really don't understand the difference between transitioning to kernel mode and transitioning/sharing data with a user mode server, I don't think you have sufficient understanding of the subject to discuss it in a meaningful way with those of us who do. Sorry.

Of course, there are "differences". Of course, putting stuff into the kernel makes it easier for the NT graphics server to access user mode data structures. But if you don't understand why that isn't relevant in this context (e.g., you still need conversions between hardware independent bitmaps into hardware dependent ones), then you really don't have a sufficient understanding to discuss the matter in a meaningful way with people who do.

The rest of your post is generally wrong and/or irrelevant to the issues at hand.

Well, you don't know what you're talking about. But don't feel bad about it--you share that with a lot of people in this industry. After all, all the flaky, bloated, shitty software that ships must come from somewhere, and it comes from people with misconceptions like yours.

Just keep in mind is that Microsoft is now on at least the fourth design of graphics subsystem (95/98, NT-user-mode, NT-kernel-mode, now Avalon). Apple had to throw out their first window system and buy themselves a new one. Sun tried twice (SunView, NeWS) with a new window system and failed both times. BeOS didn't succeed either. Neither did the AT&T Blit, or DPS, or any of the others.

The X window system design, on the other hand, has stood the test of time, with huge longevity, numerous implementations, excellent scalability, and excellent performance. Evidently, the people who designed X got it right more than 20 years ago. The fact that you don't understand what they achieved or how they did it is only your loss.

Re:Why isn't this already out?

[Jherek Carnelian]

SPARC register windows suck in practice, they have all kinds of ugly side-effects like making longjmp hard, making super-scalar hard, etc, etc. Sun has managed to maintain compability, but it was expensive and register windows are primarily just legacy cruft now.

A process calling into the kernel *is* a context switch, because the "context" is defined by the combination of the processor state (registers) and activation record (the stack)

Well then, a function call is a context switch because the stack pointer changes and so do some of the registers.

What matters in the context of this thread is the performance (I did say, "not a context switch, at least not in the same way."), and a switch into the kernel is a lot less expensive than a switch to another process. Not saving FP (which is a unix convention, not a linux convention) as well as the segment and other misc registers makes kernel calls cheaper than scheduling a new process.

Re:Why isn't this already out?

[FooBarWidget]

And yet OpenGL games on Linux run just as fast as the same OpenGL games do Windows (of course, assuming you have an NVidia card).

Re:Why isn't this already out?

[CoughDropAddict]

As a Mac OS X user, a previous Windows user, and a current Linux Desktop user, I will not be the first to tell you that X is slooow.

As a previous BeOS user, I'm here to tell you that OS X, Windows, and X are ALL slooooow.

Resize a Tracker window in BeOS and compare with resizing the Finder on OS X -- you will be blown away at the difference (sadly, you'll probably never get a chance to do this). I don't care how fast your Mac is, there's no comparison. (BTW, I'm a Mac fanboy in most respects).

No windowing system has in responsiveness what BeOS had 8 years ago, even with 8 years of Moore's law.

What's the difference? I'm pretty sure it's multithreading.

Re:Why isn't this already out?

[Zphbeeblbrox]

interesting! I'll have to remember that one. Course the fact that it's not common knowledge still makes it hard for the average guy to find out that information. So my point still stands. Thanks for the tip though.

Re:Why isn't this already out?

[Zphbeeblbrox]

Except in inetd they aren't hidden processes which is what my point was. And system services is a broad term if your latest adware app happens to think it is one.

Re:Why isn't this already out?

[cahiha]

Thanks for the kinder, more detailed response.

People who understand operating systems architectures really do disagree about how much of the GUI belongs in kernel mode. [...] For NT, I think putting the GUI in kernel mode is pretty obviously the right choice. For Linux, I think leaving it in user mode is the right choice. It's all a matter of the right trade-offs for the system design, the hardware and the usage patterns.

You are optimizing a narrow set of criteria, mostly related to efficiency and responsiveness. But what "the right thing" to do isn't a purely technical question: you also need to take into account the effort that goes into developing and maintaining it, to what degree the GUI supports innovation and new interaction styles, how well it is specified, and many other aspects.

The people who designed X11 didn't worry about squeezing the last drop of efficiency out of the GUI by putting things into the kernel, they made a choice that maximized maintainability, utility, and portability, and that is why the design has survived for 20 years, while Microsoft and Apple have had to do one rewrite after another.

but you seem a bit emotional about X, and I don't usually find it productive to discuss technical issues

I find it frustrating that a large percentage of engineers seem to just not care about design and engineering principles, beyond a small simplistic criteria that they can easily quantify. It makes no sense to optimize performance of a GUI at this point: even if Win32 graphics were faster (I don't think it is), X11 is more than good enough. I think it's actually not because they don't know that they should be doing something else, it's that they simply don't have any good ideas other than how to do performance tuning.

That's why Microsoft Windows, after twenty years of development, the supposedly best minds in the industry working on it, and billions of dollars poured into it, still doesn't work any better in any practical sense than half a dozen open source kernels with a twenty year old window system on top of them.

A few microkernel hold-outs [...] Others think user-mode servers are generally a poor design, and prefer everything in either the client process or kernel mode.

Proponents of both of those positions are making the same mistake: they think it matters and worry a lot about efficiency. There is a third approach: UNIX isn't a microkernel, yet it keeps a lot of stuff out of the kernel. That's why it doesn't have ACLs, metadata, or GUI kernel services. Like X11, and unlike any kernel design from Microsoft or Apple, the UNIX design has stood the test of time.

The world really is ready for something better than UNIX/Linux and X11, but Microsoft, Apple, Sun, and any of the other big names with money just aren't stepping up to the plate.

Re:Why isn't this already out?

[cortana]

tasklist /svc

Three ENGLISH articles instead of Babelfish

[Flywheels of Fire]

I got tired of reading the article on Babelfish. A bit of googling found me these three relevant articles:

1. Seth Nickell has posted a few videos showing the Luminocity window manager doing some super Open GL hardware acceleration tricks.

2. Interview: Rasterman Speaks of Enlightenment .17

3. XGL file format specs

Re:Three ENGLISH articles instead of Babelfish

[Nagus]

The parent linked to the wrong XGL. This is the Xgl that was meant in the article.

Re:Three ENGLISH articles instead of Babelfish

[McLoud]

Oh, great! Now we should see an editor posting these 3 links in an article and get /. crowd blaming on dupes.

bablefish

[jbeaupre]

I enjoyed reading the machine translation from german. Makes you think about about how language works and it's down right funny. My favorite line (from a comment): "With open SOURCE is too much abgekupfert." Don't know what it means, but I find my self agreeing...

Re:bablefish

[tveidt]

> "With open SOURCE is too much abgekupfert."

Ah, looks like Troll v.24,7, German Edition

Re:bablefish

[One Childish N00b]

'abgekupfert' means 'plagiarised' - the comment translates as 'with open source, [there] is too much plagiarism'. Take from that what you will...

(IANANGS - I Am Not a Native German Speaker)

Re:bablefish

[toxis]

abgekupfert is the perfect form of the verb abkupfern (Kupfer = copper) which comes from the old profession of engraving famous paintings in copper and other metals.

Though it takes a lot of talent those engravers (Kupferstecher) were not creative by themselves and if today a German says something is abgekupfert he/she means it is still just a copy and ignores the hard work behind it.

Y Windows

[BlacBaron]

I recall seeing this a while ago Y Windows

Re:Y Windows

[BlacBaron]

Release wise yes, but from the mailing lists there appear to be a few people still working on it.

Re:Y Windows

[WillerZ]

Don't go there expecting anything you'll be able to use in the near future. I fully expect HURD 1.0 to be released before we're done.

Please don't join the mailing list and ask "is anyone still working on this?" or "when will feature x be included?", because I'm tired of telling people to fuck off. We're working on it, we'll work on the features _we_ want in the order we feel like doing them. If you want something done you can do it yourself or pay someone else to do it for you.

Apologies for the rant: the usual followup to that link being posted on /. is a stream of fools bitching at Mark/Andrew/me for not working hard enough on Y. I work on it in _my_ time, and people telling me what I ought to be doing usually causes me to go do something else entirely.

Phil

( phil -at- y -hyphen- windows -dot- org )

Re:Y Windows

[AKAImBatman]

Just a friendly suggestion, but you might be able to keep people off your back if you occasionally add a news item. It doesn't have to be any big news (like a release), just something to let everyone know "Hey, we're still here! Don't bug us about it!" It can be as simple as a status report or "I had this cool idea and I'm working on coding it." It could even be the line, "Yes, we're still here." (Although that won't keep anyone's interest for very long. ;-))

Just my two cents.

Re:Y Windows

[youknowmewell]

Phil, get back to work and make me some cool graphics pronto!

Re:Y Windows

[WillerZ]

I considered using this: http://en.wikipedia.org/wiki/Mark_V_Shaney and cron to do something fun with the mailing list archives, but I figured people might notice. Perhaps they wouldn't...

Phil

Re:Y Windows

[WillerZ]

I guess I asked for that.

Re:Y Windows

[Creedo]

Where are the private arch repositories(yours and Marks)? I have never seen a link that I can recall.

Re:Y Windows

[WillerZ]

I don't care if no-one else gives a shit about it. Doesn't stop me having fun. Market share, PR, and all that crap is for companies and egotists.

Phil

Re:Y Windows

[Sleepy]

Heh... nicely put! :-)

Seriously, do yourself a favor and just say something on the website like "not much status to report; check back in about 3 months for the next News item".

You want to at least differentiate yourself from projects that really are abandoned, if only to reduce noise emails and list postings.

Re:Y Windows

[Queer Boy]

people telling me what I ought to be doing usually causes me to go do something else entirely.

Yeah, really you ought to be doing something else entirely. I've seen the beast.

Re:Y Windows

[WillerZ]

Mark's is linked off the website but RSI means he doesn't develop much these days.

Mine is linked from here http://www.y-windows.org/pipermail/y-devel/2005-Fe bruary/001903.html

And Andrew doesn't seem to have linked his publicly.

Phil

Re:copied?

[Doctor Crumb]

so? OSX has a very nice graphics architecture with lots of potential. X11 is old and crufty, and these sorts of effects require large portions of code to be rewritten. If it can suddently render a genie effect efficiently, imagine how quick it'll be to render more mundane windows!

Re:incongruencies

[afd8856]

Download a skin and stop bitching. Nuola is working quite nice for me.

Why is this in the Linux section?

[Rick the Red]

X11 is not just for Linux, you know!

Re:Why is this in the Linux section?

[jellomizer]

But we dont talk about them

Re:Why is this in the Linux section?

[dzeuthen]

> X11 is not just for Linux, you know!

Oh, sorry, you're right. It needs to be in the Amiga section.

Re:Why is this in the Linux section?

[shaitand]

Right, cuz you know IRIX has a serious chunk of the desktop market.

Re:Why is this in the Linux section?

[Sabalon]

I posted that once in a discussion about Fresco or whatever the "from-the-ground-up" replacement for X on windows was.

Got a long flame about how HPUX, AIX and Solaris are all going to go away and Linux was the only thing that matters so other OS compatability didn't matter.

With advocates like that, who needs Microsoft!

How about doing something actually useful ?

[Anonymous Coward]

How about implementing dynamic X server reconfiguration to allow connecting and disconnecting external monitors to laptops on the fly? How about using different resolutions on these monitors?

Right now Linux/X11 is horribly behind both Windows and Mac OS X, being unable to detect an external monitor being connected and change resolution accordingly.

Re:How about doing something actually useful ?

[ssj_195]

I'm sure they are working on this also. Unfortunately, XFree was stagnant for such a long time that there is a lot of catching up to do. I'm quite happy to see this kind of "blue sky" research as long as it is happening alongside stuff that will be immediately useful. Very happy, actually, as it all looks rather sexy :)

Re:How about doing something actually useful ?

[Anonymous Coward]

"Right now Linux/X11 is horribly behind both Windows and Mac OS X, being unable to detect an external monitor being connected and change resolution accordingly."

So, is Windows being so advanced the reason why it regularly resets my dual display to a single? Or is it so advanced that it requires that display #1 be on the left and display #2 on the right -- that is if I want the mouse cursor to appear in the right place when moving it from display to display?

With Linux, my dual display does not have the same problems -- and yep, it's the same video hardware. Both systems use 2 monitors with different resolutions on both.

Re:How about doing something actually useful ?

[gsasha]

That maybe works on a workstation (even then, question is how much manual tuning you had to do for that). But on laptops things are much worse, at least on my Thinkpad R40 I *never* seen Linux discover the two monitors, much less allow me to do advanced switching (like showing the same/alternate desktop on the monitor connector, disabling it or enabling the video out of the laptop...). On windows it worked like a charm.

Re:How about doing something actually useful ?

[gsasha]

Nope, not moving no goal posts, siree. BTW., very strange to hear that Windows wouldn't let him rearrange the physical location of the monitors - IIRC, that's easy and works well. I say that Windows *worked*, implying that I use Linux on my laptop exclusively. If there is a possibility for such advanced monitor configuration for R40, then I don't know about it (but I would gladly hear. Honestly, I didn't check in the last 1/2 year or so).

Re:How about doing something actually useful ?

[Spy Hunter]

Maybe Windows works great on your laptop, but I've had nothing but trouble this week moving desktops around and connecting them to different monitors. Windows rarely does the right thing, and sometimes even has to be rebooted just to notice the change. Also, sometimes after I disconnect my TV Windows likes to switch around my preferences so that any video that starts goes fullscreen on my main monitor instead of the TV, covering the player, desktop, and mouse pointer until I can figure out how to quit the video player with the keyboard (Alt-F4, unless I've focused another application with the invisible mouse, in which case I'm screwed).

Windows could use a lot of improvement in that area. Maybe some of it is the fault of NVIDIA's drivers, but maybe some of Linux's problems are their fault too (when using their driver).

Re:How about doing something actually useful ?

[Kihaji]

So, is Windows being so advanced the reason why it regularly resets my dual display to a single? Or is it so advanced that it requires that display #1 be on the left and display #2 on the right -- that is if I want the mouse cursor to appear in the right place when moving it from display to display? No, it's so advanced that the drivers for your card have a different designation for what is primary and what is secondary depending on OS/Manufacturer.

Re:How about doing something actually useful ?

[Mr Teddy Bear]

Oh please, you can move the monitor positions around in Windows. You could do that since Win98 when multiple monitors were first implemented. Windows has much better multi-monitor support than Linux does. Granted, Linux has come a long way with DUAL monitor support... but in my case.. i have 3 monitors and i have never gotten it to work quite the way i want it to. (Though still usable.. just needs some more work is all.)

Re:How about doing something actually useful ?

[otis wildflower]

How about implementing dynamic X server reconfiguration to allow connecting and disconnecting external monitors to laptops on the fly? How about using different resolutions on these monitors?

How about GL accelerated xinerama?

Re:How about doing something actually useful ?

[MikeFM]

You mean like the nice Mac 'feature' of changing my screen resolution when I boot up before turning on the monitor? Yeah, great idea. Annoys the shit out of me.

Not that improvments couldn't be made. Just don't follow the example set by Windows or MacOS.

Re:How about doing something actually useful ?

[shaitand]

"Windows rarely does the right thing, and sometimes even has to be rebooted just to notice the change."

Yeah but that IS working great compared to X. Unless you get really lucky, rebooting does not fix your monitor in X either. The X server has to be reconfigured by a sys admin for every monitor plugged into it.

Re:How about doing something actually useful ?

[Brandybuck]

Conference room R7. Tensions are high. The CEO and VP of Sales are waiting for the presentation that will make or break the careers of two dozen directors, managers and project leads. The presenter plugs the cable from the Epson projector into his Dell/WinXP (with SP2!) laptop. Nothing happens. The presenter presses several different key combinations. Nothing. The secretary squirms in her chair. An engineer leans over and presses a button on the projector. Nothing. The CEO clears his throat. On the wall of the conference room are two glowing red words: "No Signal." Five minutes pass as very people offer helps suggestions, the cables get unplugged and plugged several times, and various buttons are pushed. The Sales VP is starting to get up to leave, when finally an image appears on the screen. Then PowerPoint crashes...

Re:How about doing something actually useful ?

[Politburo]

Also, sometimes after I disconnect my TV Windows likes to switch around my preferences so that any video that starts goes fullscreen on my main monitor instead of the TV, covering the player, desktop, and mouse pointer until I can figure out how to quit the video player with the keyboard

Windows isn't arbitrarily "switching around" your preferences. It's coping with the situation. What you've done is tell your Media Player that its default screen position is Monitor 2: 0,0. When you disconnect the TV (monitor 2), Windows can't put the Media Player there, or else you'd be here bitching that Windows is putting the Media Player on a non-existant monitor. Windows does the next best thing: it puts the Media Player at Monitor 1: 0,0.

Also, I don't know any Media Player that obscures the mouse pointer while the mouse is moving. Sounds like an inferior program or driver. From my experience, most programs are not written to work properly with multiple monitors. Furthermore, have you tried Alt+Enter to exit fullscreen mode? I don't know what app you're using, but alt+enter is the default "toggle fullscreen" keystroke. F11 is also used sometimes.

Re:How about doing something actually useful ?

[DennisZeMenace]

AMEN. That's a classic open-source problem. Everybody wants to work on the cool features, but nobody wants to do the grunt work to turn X into a freaking usable system (which is why X still can't do things MacOS could do 15 years ago, remember those monitors you could just rotate 90 degrees and the desktop and windows would resize by magic ?).

DZM

Re:How about doing something actually useful ?

[cahiha]

How about implementing dynamic X server reconfiguration to allow connecting and disconnecting external monitors to laptops on the fly? How about using different resolutions on these monitors?

That's been there for, oh, at least a decade in XFree86. Of course, other X11 servers have had it longer.

Right now Linux/X11 is horribly behind both Windows and Mac OS X, being unable to detect an external monitor being connected and change resolution accordingly.

No, it isn't.

Re:How about doing something actually useful ?

[Spy Hunter]

You think I don't know about Alt-Enter? This fullscreen mode is different from normal media player fullscreen mode. I guess it is actually a function of the driver, which makes it NVIDIA's problem. You can set it so that whenever you start an application that plays video, the video "takes over" a display of your choosing, obscuring everything on the display (in addition to playing in its normal window, which can be made fullscreen or not independently). This is great when hooked up to a TV or projector because you can stick the player on the first monitor, and only the video ever gets shown on the TV. PowerPoint has a similar capability to take over the second display, but that's a feature specific to PowerPoint.

Anyway, whenever Windows realizes that the TV is off (which is sometimes hours after it actually has been turned off) the video starts covering the first display, and the next time I start a media player, it takes over the display until I manage to quit the media player, which is sometimes hard when you're basically blind. Like I said, it's probably an NVIDIA problem. Some of Linux's problems may be due to NVIDIA drivers as well.

Re:copied?

[cabraverde]

To me a lot of these effects are just copied from OS X

Are you implying that that's a bad thing? OS X has many nice GUI features. I'd like to see some of them on my Linux desktop

screenshot mirror

[Anonymous Coward]

enlightenment
luminocity
xgl

UI stuff is tough to do open source.

[Viltvodlian Deoderan]

So in another 4-5 or more years X will have the same stuff that OS X has had for a while? This highlights the problem with Opensourcesoftwaredevelopmenten. Things go swimingly until some really un-fun interface code needs to be written. At that point, you really want to pay someone to do the grundge work. Auf Weiderscrheiben, Mike .

Re:UI stuff is tough to do open source.

[mattyrobinson69]

its because xfree86 was so stagnant for so long, now xorg is the popular x11 we are seeing fast development again. hopefully this will continue and we will get a great x11 again.

Re:UI stuff is tough to do open source.

[prockcore]

Things go swimingly until some really un-fun interface code needs to be written.

No, it highlights the problems that *every* OS faces.. revamping older libraries is difficult.

Look at how many OSX apps still use Carbon (including many Apple apps). Look at how much legacy crap is in Windows.

Updating framework libraries is difficult to do cleanly and unobtrusively. This isn't unique to opensource in the slightest.

Re:UI stuff is tough to do open source.

[mz001b]

does OS X have network transparency?

Re:UI stuff is tough to do open source.

[Lussarn]

You do understand most of OS X is open source? Does the BSD parts of OS X sucks? I better tell jobs so he can promote going back to cooperative multitasking and no memory protection.

Re:UI stuff is tough to do open source.

[poot_rootbeer]

Things go swimingly until some really un-fun interface code needs to be written.

It's not the interface code that's the bottleneck; I'm sure there are enough coders out there who can write a windowing API and write it flawlessly.

It's that the Open Source Community is well-stocked with developers, who understand the hardware/software interaction intimiately, but lacking in interface designers who understand the software/human interaction. It's almost as if "usability studies" and "focus groups" are anathema to the OSS model, shunned as a useless relic of bloated corporate suit-wearers.

Apple understands the importance of the human interface and it shows. What can be done to get open source projects to give more consideration to interface design?

Re:UI stuff is tough to do open source.

[mveloso]

It used to, back in the NeXT days. It's unclear what happened to it. I never heard that it was removed, but nobody uses it.

Theoretically you could shim the ObjC runtime to send UI stuff over the wire. That would be a great hack, if anyone had a lot of free time.

Re:UI stuff is tough to do open source.

[t_pet422]

... but lacking in interface designers who understand the software/human interaction.

You're absolutely right. I'm reminded of an interview with Bill Hill, the man who designed ClearType for Windows. He says that the most important operating system is not Windows or Linux or the Mac; it's homo-sapien 1.0, and there's no upgrade in sight.

Re:UI stuff is tough to do open source.

[otis wildflower]

Apple understands the importance of the human interface and it shows. What can be done to get open source projects to give more consideration to interface design?

Pick a good team of UI folks and let them block releases for usability issues?

Re:UI stuff is tough to do open source.

[cahiha]

Sorry, no go. You can make graphics appear remotely, but that isn't network transparent windowing. The reason Jobs nixed the remote parts of NeXT in OS X is probably because he realized that.

Re:Meanwhile Today On Earth...

[tveidt]

> come by any Apple Store and pick up a mini

This is illegal where I live. Here, we have to give money to a store in order to get something from them. Sigh.

X free of CPU and RAM usage

[Anonymous Coward]

Many complain that CPU speed does not increase much from the user perspective but what if the new X11 tech brings us GPU based jpeg decompression?

Surf your photos and they go straight to the GPU instead of storing a CPU decompressed bitmap in RAM, the speedup would be incredible. Low CPU usage in laptops as GPU does the work.

Remote X11 display without recompression of the network stream? It would become as fast as surfing. Requested jpegs being send straight to the receivers GPU, simply upgrade the GPU in school computers to get very fast thin client Linux boxes.

Look at Apple's Core Image in Tiger: possibilities will be amazing.

Re:X free of CPU and RAM usage

[Nagus]

Dude, jpeg decompression is so efficient that it's basically free. I mean, loading the actual data from disk (or network) takes a hell of a lot more time than decompressing it.

Much more interesting is the ability to render SVG images with hardware acceleration. The xsvg renderer will give us that ability (when used with glitz as cairo backend).

Resolution-independent graphics, rendered at high speed. That is what will make for really amazing possibilities.

Re:X free of CPU and RAM usage

[Spy Hunter]

Um, JPEG is already very fast. Even if you could accelerate it on the GPU, there's really no reason to. The speedup would not be "incredible", more like "unnoticable". The RAM usage would not go down, you would just use video RAM instead of system RAM, which is *not* a good trade (video RAM is much more precious).

The bottleneck in remote X11 display is *not* decompressing and recompressing JPEGs, it is the network. Modern remote-display systems (NX, VNC) already use JPEG compression. And they already work very fast over a LAN; in many cases fast enough that you don't even notice them. You don't have to imagine "very fast thin client Linux boxes", just go set some up!

Re:X free of CPU and RAM usage

[mingrassia]

Much more interesting is the ability to render SVG images with hardware acceleration.


The real area for this to make a difference is with mobile devices. This is where an open standard like OpenVG will make things really interesting!

Re:X free of CPU and RAM usage

[darkwhite]

render SVG images with hardware acceleration...Resolution-independent graphics, rendered at high speed

Yep, that's one feature that I can get really excited about. The luminous transparent eye-candy is completely irrelevant compared to this. A fast, fully SVG toolkit and desktop environment would be a breakthrough, and modern GPUs probably have plenty enough power to render SVG on-chip.

bootstrapping problems

[CowbertPrime]

We were discussing the X11 OpenGL server at the LWE X BoF session. IIRC, the current problem with full native implementation of the OGL server is that starting the ogl server requires the dri layer, which requires an X server to be running.

HCI

[paithuk]

It has actually be shown a number of times that fancy features (such as integrating a physics engine into the desktop as so) actually leads to a more complex and harder to use system. I have to congratulate these guys on what they've achieved, but at the same time I have to wonder if this is the right direction to take, especially since Linux's only major flaw is in fact its lack of usability.

Re:HCI

[coolGuyZak]

The XGL desktop extensions are currently being designed with usability in mind. The developers are attempting to do 2 things:
1) Make the desktop seem more real to a user
2) Decrease or entirely remove graphical lag from the system. This is already apparent within x.org's work. The windows slide around seamlessly (with the Composite & XDamage extensions loaded).

Adding the "wavy windows" code into it makes windows seem more like real objects, and thus immerse the user more.

I agree with your opionion that the interfaces are currently pretty bad, but they are constantly improving (insert obligatory KDE/Gnome/Freedesktop.org reference here). And of course, with great 3D power comes even greater responsibility to use it correctly. And there will be people who abuse it.

And now for something completely different.

While I see 3D desktops (3DD) to be interesting theoretically, they have yet to find practical uses. Many times this is because they break the 4th wall, and thus seem less real to users. We havn't found the proper way to use the technology provided to us.

However, it is also my opinion that if we keep trying different things, we will eventually stumble onto the right track. How long it will take is debatable.

Re:HCI

[KillerBob]

Here's as good a place as any to point this out...

I keep looking for a 3d-accelerated desktop. I could get what I want from Mac, but quite frankly, I can't afford to buy one. Otherwise, I probably would buy one for my next computer.

See, the major advantage, to me, from having a 3d-accelerated desktop is eye candy. I would find floating around on a 2d screen extremely annoying. But I would very much like to have decent transparency effects, and nicely rendered windows. Properly anti-aliased fonts would be nice, as would having a "real" lighting source to throw shadow on rendered buttons/user interface (maybe one that moves and changes colour to indicate the time of day). All of it is eye candy, and doesn't actually affect the usability of my computer at all.

*but*, it's something that I think would look cool, and would make using the computer a little less mundane. Extensions that do some of this are available for Windows and Linux, of course, but none of them actually *use* my graphics card's rendering capability. The last time I used transparency on my Windows machine, it was a nightmare. You wouldn't believe the graphics lag once I had more than one window open. I've got a 256MB ATI Radeon 9600Ultra video card in that system, but the windows weren't being rendered by it at all. It was all CPU-based graphics rendering, and would have looked the same on a 1MB video card using a VESA interface.

The real benefeit of using a 3d-accelerated user interface isn't so that you can create a 3d user environment, it's so that you can use the graphics card's capabilities to offload processing of eye candy and free up system ticks.

Re:HCI

[DoWeHaftTo]

Various thoughts:

o While fancy X11 visual effects remind me of fancy Movie effects - usually shallow but entertaining eye candy - perhaps they might inspire and enable new UI approaches.

o One of these days we are going to get beyond pixel-addressed GUIs and get back to the future with display postscript or similar resolution independent display technology. Just waiting for the display devices to get to decent resolution I know, (a significant fraction of paper would be nice), but with all the effort going towards displays of all kinds, particularly for computers, cell phones, HDTV (woefully lo-res that it is - your cell phone may soon have higher res), and better yet, digital cinema, the research into display technologies is roaring as people really want it. Now if the software (legacy windowing systems for example) will be able to keep up...

o If IBM or Sun or ? ever really wanted to take market share from MS the answer is pretty obvious - just make Linux into a viable desktop platform. It all starts from there. Sure MS Office replacements are an essential part of any such strategy (and apparently coming along), but the main thing is to get the apps onto Linux. And what does this have to do with X11? Think OSX. Put a cool, professionally designed face on Linux and get vendors to write their apps to it. Same as OSX. But with Linux and its current apps as well as the apps IBM and/or Sun could bring to the party there is more than enough to prime the pump for the commercial vendors (I'm suggesting IBM/Sun/etc donate a lot of stuff up front like they already have). So my feeling is that X11 visual effects are one thing, but a professional OSX-level platform is the larger and much more significant vision.

TB

PS. Agree with the parent about HCI - why is Apple about the only major computer system vendor that has a clue in this area? Maybe some companies haven't figured out most EEs and CSs don't know F about UI design (it's not the code stupid).

Re:HCI

[otis wildflower]

I keep looking for a 3d-accelerated desktop. I could get what I want from Mac, but quite frankly, I can't afford to buy one. Otherwise, I probably would buy one for my next computer.

$574 is unaffordable? The mini has a 32MB graphics adapter that is Quartz Extreme capable.

Swap out the slow-ass HDD and you've got a solid little 'puter. G4 =~ Pentium M...

Re:HCI

[coolGuyZak]

Yeh. That too.

My point is still valid though. Eye candy can be used to improve the usability of a system. A few more examples that your post made me remember:

Transparency & "Z Buffering" of windows- Makes them seem more like they are in "the background", which brings better highlighting to your main window.

Scaling the desktop down to fit in the pager, with live updates.

Properly antialiased fonts (taken directly) - increases readability & decreases eye strain.

"Redrawless windows" - When moving windows around, you don't have to worry about flicker. At all.

Then, on top of all of that is bragging rights.
Me: "Oh, yeh... I thought it'd look neat to have my buttons sparckle when I mouse over them"
Them: "How did you get it to look so crisp? Oh! Dude, when you move the window the computer doesn't slow down either!"
Me: "Oh, that's the pixel shaders."
Them: "Wait, you have a pixel shaded desktop?"
Me: "Heh. And you said Linux sucked."

Then, after finding out it's linux, they'd call it a horrible misuse of resources... and then I get to say "And Windows isn't?" Hardy har har.

That was rather rambly.

Well, back to work.

Re:HCI

[KillerBob]

When I add in the extra costs for accessories I need in a computer, convert it to Canadian dollars, and then remember that I'm a full time student who quit his job in December, yes, that's unaffordable. Especially since I already spent my tax refund on a car repair. ... Quite aside from the fact that that particular computer is a downgrade from *both* of the computers I currently use. One's an Athlon XP 2400+ w/ 1024MB of RAM, a 360GB hard drive, and a 256MB Radeon 9600Ultra, and the other is a 1.4GHz Athlon M laptop w/ 512MB of RAM, 40GB HDD, and a 32MB Radeon mobility card.

Besides, both run Slackware perfectly. I don't need a new computer at the moment. When I'm better positioned to buy a new computer, I'll likely buy a mac. Right now, I can't afford to buy one that's actually an upgrade. Hence, I'll wait until somebody writes a 3d-accelerated UI and releases it under the GPL or BSD license.

Well sure, but...

[FreeLinux]

surely you can see the immediate need and usefulness of transparent windows and wobbly windows. Not to mention that the present versions of X11 are only using from 50 to 100 megabytes of memory when modern systems have 512 to 1 gig available. I think once we get the bugs worked out of these new features, then we can look into more advanced stuff like "hot-plug monitors" and dynamic resolutions.

beware

[qortra]

That E17 article is insanely old (almost 4 years old). Normally, it wouldn't matter, except that I think E17 has gone through redesign[s] since then. Just check out the screenshots; nothing like the last taste of E17 we got late last year. I wouldn't count on any of the information being useful or up-to-date

Re:incongruencies

[The Real Nem]

Or create a new toolbar and drag the address bar onto it. My address bar sits alone on its own toolbar taking up the entire screen width.

Movie representations of computer UI

[Winterblink]

I remember watching movies like Hackers, which is a fairly decent movie overall, and totally laughing at their representation of the user interfaces on the computers. From the seriously hacked up and personalized desktops on everyone's PCs to the "flying through the mainframe" hacking at the end of the film, I was convinced it was there as a joke.

But it seems nowadays desktop environments are getting to be SO customizable and graphically "enhanced", I start to wonder whether those old movies weren't jokes but rather premonitory.

Re:Movie representations of computer UI

[mattyrobinson69]

there was another film i saw (had the white bloke from pulp fiction, cant remember his name). some dude was writing a virus (or worm or something) and it said on the screen with a very graphical interface "creating virus" or something.

it made me laugh (isn't gcc good enough for these people that they need a framebuffer and a custom compiler/frontend that says "creating virus" but gives no output)

Re:Movie representations of computer UI

[coolGuyZak]

Does that meant that in the future, the script kiddies will be able to fly through my system like superman?

Re:Movie representations of computer UI

[radish]

Swordfish, John Travolta.

You're welcome.

Re:Movie representations of computer UI

[xtremee]

The name of that movie is Swordfish.

Re:Movie representations of computer UI

[Chyeld]

They were jokes.

Even if the common user's interface is some sort of stylied 3-D world (and really, while it would be cool for games, who would actually want to try to work using something like that), a real 'hacker'/power-user/informed user is going to stick to the minimalist text UI as it gives you more direct access to the underpinnings of the system.

Re:Movie representations of computer UI

[Drooling Iguana]

"It's a UNIX system. I know this."

Re:Movie representations of computer UI

[lumpenprole]

Ah, it's the other way around. Right now some designer is thinking "Wow! With all these new killer extensions I could put together a desktop just like the one at the end of Hackers!"

You can start to freak out when the Lawnmower Man fans get their hands on it.

Re:Movie representations of computer UI

[Winterblink]

Oh god, Swordfish... that whole "making the virus" scene literally had me cringing at how horrible it was. He did everything but shove the keyboard down the front of his pants and fuck it six ways from sunday. I like coding, but not THAT much.

Re:Movie representations of computer UI

[mattyrobinson69]

the only bit i can remember from that film are that "creating virus". i must have blocked the rest out.

Re:Movie representations of computer UI

[Winterblink]

Unfortunately I watched the rest between the fingers of my hands covering my face. I think I even quoted C3P0. "Oh, how horrid!"

Re:Movie representations of computer UI

[bicho]

ah... Jurassic Park...
I don't quite remember that scene though.

Re:Movie representations of computer UI

[Anubis350]

you mean like all those guys who develope on osx and turn off aqua to use wmaker, oh wait......

Re:Movie representations of computer UI

[mrchaotica]

No, more like all those guys (like me) who develop on OS X and have a bunch of Terminal windows open. It may have pretty transparency, but it's still a text interface...

Re:Movie representations of computer UI

[mrchaotica]

That was a real interface, you know -- it was made by SGI.

Re:Movie representations of computer UI

[m50d]

Not for hacking. Hackers largely still use command lines. Movies don't like having plain text going (although there was real hexediting of that half-copied disk, which I was glad to see), so they'll use the 3d toys (jurassic park anyone?) but the fact is that's often where an experienced user is most productive. Hackers won't care much about looking good, they'll just use the text mode.

Re:Movie representations of computer UI

[Queer Boy]

I start to wonder whether those old movies weren't jokes but rather premonitory.

premonitory | adjective
a strong feeling that something is about to happen, esp. something unpleasant

Re:Movie representations of computer UI

[Winterblink]

Is it a good thing then that our computers are capable of doing these things? Sure it's neat to have crazily distorted windows and such, or windows that are translucent so you can see what's behind -- but how USEFUL is that in most circumstances? Not saying it's a BAD idea per se. :)

Re:Movie representations of computer UI

[m50d]

Yep. And it was distributed in the "toys" category.

Why not X12 instead

[Anonymous Coward]

How long will they continue on this "11" series. Isn't it about time to upgrade to X12

Re:Why not X12 instead

[node 3]

How long will they continue on this "11" series. Isn't it about time to upgrade to X12

It only goes to 11.

Re:Why not X12 instead

[FuzzyBad-Mofo]

Yeah, but 12 would be.. one more!

Re:Why not X12 instead

[rookworm]

It only goes to 11.

Yeah, but when you need that little extra kick, you can bring it up to 12!

Re:Why not X12 instead

[bankman]

It only goes to 11.

Excellent, finally there is binary versioning. That's what they should use for the kernel, it would make Usenet discussions with newbies far more interesting.

agree with the article...

[uodeltasig]

I would have to agree with the article on the following point:

The problem here is to implement in all drivers sufficient 2D-XRender-Hardwarebeschleunigung - those actually simply only one special case of the existing 3D-Beschleunigung represents.

I've had a hard enough time trying to figure out how to say "Hardwarebeschleunigung" let along trying to implement all the drivers for it.Despite this it is good to know however that...

Without the parallel running videoaufzeichnung the animations ran absolutely liquid.

Karama Reedemer: Below is the babelfish translation to the mirror. Mirror dot translation

Re:agree with the article...

[argent]

Now for the bad news: Arbitration and translation schemes have had unfortunate clenirations with the ridgeway armiphlage.

Re:agree with the article...

[One Childish N00b]

hardwarebeschleunigung = Hardware accelleration.
videoaufzeichnung = Video drawing/rendering.

There seriously wasn't an English site on this? and Babelfish is a useless translation engine anyway, I had to go back and read the original German article to avoid getting a headache - my German isn't perfect, but the English-with-German-technical-terms was addling my brain...

Re:agree with the article...

[jd]

If the animations are running absolutely liquid, I suggest the reviewer cuts back on the acid.

Re:agree with the article...

[CrosseyedPainless]

Don't you just hate machine translation in the Slow Zone?

Re:agree with the article...

[LarsWestergren]

Ahh... I love you guys. Quotes from my favourite book. :-)

I have seen a couple of Slashdot users with inspired names - Twirlip of the Mists for instance.

You've read Deepness in the Sky too, I hope?

Re:agree with the article...

[forkazoo]

IMHO, babelfish is handy for somebody like me, who can mostly read the German, but doesn't have the vocabulary for all of it. I primarily read the german, and use babelfish as a secondary reference for translations on some of the individual words.

But, yeah, for a non-speaker, babelfish would be pretty much useless. When I use it from a language I don't understand, I am never sure if I understand the translation.

Re:copied? (-1, Irrelevant)

[orasio]

Some are.
Some aren't.
Some were even available in demos ten years ago.
Some are obvious, and the fact that OSX implements them, means nothing.
For example, that "expose" feature that is so praised, is an obvious improvement on the window idea, and there were already papers written, and many people already implemented it in some way (heck! I even keep all my windows shaded, so when I shade the one I'm using, I can see all of them at the same time!). I didn't think OS X was copying me, or "stealing my IP", put in a more fashionable way.

also beware

[DrWhizBang]

That is not the XGL that you are looking for. This is.

Not very good googling.

Re:Too Late - I've already bought a Mac

[bloggins02]

Sorry, Unix is God, but I've bought a Mac.

Considering that the newest version of Mac OS is Unix, how does that sentence make sense in any way?

What I want

[RealProgrammer]

I want an interface that lets me think in 3D.

• I want to be able to grab an edge of a window and push that edge out of the way, with the window aspect changing according to how far I push it. Grab the left border and move it right, and the contents of the window compresses (e.g., making the text look skinny).
• I want to be have a large virtual desktop which I can zoom out away from to show groups of screen objects (windows, icons, local backdrops, etc.), and zoom in on to show the objects close up. The objects should not all be in the same plane, so when I zoom in on one set of object I can still see ("far off") other tiny sets of objects. One effect of that would be to allow hiearchical groups of objects.
• I want to take a group of objects and wrap them in a box, which I can label arbitrarily. The box should have variable opacity, perhaps password security, and should respond to signals (it should be a process).
• I don't want to have to use a pointing device. If necessary, I'd rather use a subvocal microphone/sensor, keyboard mouse driver, eyeglasses, or a chin strap than a mouse, touch pad, trackball, or nipple.
• I want a video driver / X server that outputs stereovision to two displays (or two halves of a single display).

And I want it to be Free.

To answer the obvious retort: every time I get started learning X programming, my feeble little brain starts to hurt. Kudos to you wizards out there who grok X.

Re:What I want

[Anonymous Coward]

Zooming and keyboard bindings I agree with, the rest of your proposals do not seem compelling.

Re:What I want

[duerra]

http://www.hamar.sk/sphere/

Re:What I want

[RealProgrammer]

That's for a system I don't use.

Re:What I want

[duerra]

Well then next time you should specify Free as in Speech, else you may get Beer.

Picky picky picky ;)

Re:What I want

[glwtta]

Like all of these 3D desktop ideas, this doesn't sound all that compelling.

Face it, monitors and input devices are two dimensional (well mice at any rate, keyboards are one dimensional), simulating a third dimension adds complexity - both in use and in implementation - and doesn't add anything to useablity or productivity. Sure, you get about 5 minutes of "Oooh, shiny!", but that's about it.

Navigating 3D space with any of the current input devices is a huge pain in the ass, trying to do useful work with a large amount of data on such a thing will get frustrating very quickly. They make it look cool in movies, but that's becuase it's scripted.

Re:What I want

[veg_all]

hee hee.

"nipple."

hee.

Re:What I want

[TuringTest]

# I want to be have a large virtual desktop which I can zoom out away from to show groups of screen objects (windows, icons, local backdrops, etc.), and zoom in on to show the objects close up. The objects should not all be in the same plane, so when I zoom in on one set of object I can still see ("far off") other tiny sets of objects. One effect of that would be to allow hiearchical groups of objects.
# I want to take a group of objects and wrap them in a box, which I can label arbitrarily. The box should have variable opacity, perhaps password security, and should respond to signals (it should be a process).
# I don't want to have to use a pointing device. If necessary, I'd rather use a subvocal microphone/sensor, keyboard mouse driver, eyeglasses, or a chin strap than a mouse, touch pad, trackball, or nipple.

Then, you should support the Archy project so that it achieves it's full potential.

Re:What I want

[m50d]

All I want is 3d for arranging my windows. The windows should stay flat and look the way they do, but I want to be able to rotate myself around them. Go around the back so the window on the right is now on the left. Tilt up a bit to see my media player. Desktop switching with my mousewheel is something, but I want to be able to continuously scroll.

Re:What I want

[evilviper]

Face it, monitors and input devices are two dimensional (well mice at any rate

Mice are 3-D, it's just that the third dimention doesn't do anything right now. Pivot your mouse clockwise/counter-clockwise for the 3rd dimentional movement. Trackballs are even easier to add a 3rd dimention to.

simulating a third dimension adds complexity - both in use and in implementation - and doesn't add anything to useablity or productivity. Sure, you get about 5 minutes of "Oooh, shiny!", but that's about it.

You could have said the EXACT SAME THING before GUIs came out... Graphics add complexity, and don't add anything to usablitity or productivity.

Re:What I want

[Hatta]

You could have said the EXACT SAME THING before GUIs came out... Graphics add complexity, and don't add anything to usablitity or productivity.

And you'd have been right.

Re:What I want

[bmcent1]

I'll add one to the wish list:

I love focus-follows-mouse in my window managers. But what I REALLY want is...

Focus-follows-focus!

I dunno how to implement it, maybe with a web cam. It would be nice to have text being typed automagically going into the terminal or application I'm looking at. Guess it might not work well for people who hunt-n-peck.

Re:What I want

[sp0rk173]

Glad to see someone on slashdot celebrates 4/20.

Babelfish translation?

[c0ldfusi0n]

You mean that babelfish translates?!

Man, all this time i was thinking it was only generating random words in given language. All of it were lies. LIES!

It's a Translation!

[duerra]

Now Slashdotters have an excuse for not reading the articles!

Cairo is not a GTK fork!

[nickname_unique]

Please don't post here such nonsense! Read at least the first sentence on http://cairographics.org/introduction: Cairo is a vector graphics library designed to provide high-quality display and print output.

That's what it is. A 2D vector graphic library with multiple backends, which means you can draw something and choose if you use as drawing backend X11, a PNG file, a PDF file, glitz (OpenGL) or something else.

Gtk3(?) will _use_ Cairo and it's X11 or glitz backend to draw it's widgets!

PARENT IS TROLL

[rjw57]

There are so many errors there that I must assume you sir are a troll :).

Re:PARENT IS TROLL

[DrWhizBang]

oh, well. I bit. ;-)

Re:luminocity = longhorn in linux?

[DrWhizBang]

Oh brother.

There is a language being developed codenamed cairo..

No. Cairo is a 2D vector graphics library, not a language. ...to be used in Linux...

or Windows. or Mac. It is a cross platform library.

its a GTK fork...

No. It is not. The CVS head version of GTK uses cairo for drawing. ...and its special feature is ability to use opengl rendered screens in place of bitmaps for window drawing...

Among its features are multiple drawing back-ends. One is OpenGL, another is Render. Because it is a vector library, it may or may not render to bitmaps - depending on the backend.

A product is already being developed using this called luminocity.

Luminocity is a fork of the metacity window manager that has a built in composite manager that renders to OpenGL.

Now that that's been cleared up...

Un-fun code

[truthsearch]

I couldn't disagree more (you knew somebody had to, right? ;) Plenty of high quality un-fun code is written in the open source community. Think every line of the Linux kernel or GCC was fun to write? It's not as much the fun as how badly someone wants it. People have been toying around with this sort of thing for a long time. But there doesn't seem to be enough real community demand to get a big enough team to hammer it out.

I know nothing of graphics programming. But if I was very interested in having accelerated window animations I'd learn OpenGL and help out. There will always be someone who wants that itch scratched.

Re:Un-fun code

[MikeBabcock]

Part of this is that some people don't understand why people write OSS.

Some of us just code because we love it.

Some of us also happen to be perfectionists and want code to be better than it is now, for no better reason than because it could be.

Those people end up creating things like Enlightenment.

Re:copied?

[justsomebody]

You mean the ONE and ONLY 3D effect in OSX, when you minimize to dock (or maximize from)??? You talk about lot of effects.

Translucency is planed to be window specific as I know. Additional menu in window system menu where you define translucency.

wait..

[SpongeBobLinuxPants]

I thought X11 was dying/dead/bad and Xorg was good? Did I miss TFM?

Re:wait..

[coolGuyZak]

X11 is the standard that X clients and servers use to communicate. What you are thinking of is XFree86, which for all intents and purposes is dead/dying/bad.

Re:wait..

[mattyrobinson69]

Xorg is a fork of XFree86, they are both implementations of X11 (along with a few others that afaik died a death)

Re:wait..

[SirTalon42]

Xorg and Xfree are an implementation of the X11 standard. (your thinking of XFree86, which stagnated for years and is now dead, confirmed by netcraft)

Re:wait..

[Nasarius]

XFree86, which stagnated for years and is now dead, confirmed by netcraft

Yeah, but I don't think they know it. The funny thing is, they released 4.5.0 (4.4.0 was the one that marked the controversial license change) just a month ago, and I never even heard about it. All the Linuxes and FreeBSD (not sure about NetBSD and OpenBSD) have ditched it in favor of X.org; I don't see why they bother.

Re:wait..

[chill]

SCO still uses XFree86. :-)

Re:wait..

[Byzantine]

OpenBSD 3.7 (to be released next month) will have Xorg (or more properly has Xorg, since the OPENBSD_3_7 is live). 3.6 still has XFree86.

Re:wait..

[WilliamSChips]

Wait... SCO sells products? I thought they just leeched money off of Microsoft and, in return, sued Linux...

Re:wait..

[Gandalf_007]

It should be pointed out that the "XFree86" in OpenBSD 3.6 was 4.3.99.something forked before the license change, much like Debian has done. This was actually something the OpenBSD devs made a big deal of, and I think it's referenced in the 3.6 song.

NetBSD 2.0 shipped with XFree86 4.4 (they were one of the few distros that didn't have issues with the new license), but I believe even they plan to eventually transition to Xorg. I know it's possible to build NetBSD with Xorg now, and while I haven't been tracking -current, it's quite possible Xorg is now the default. If not yet, it will be.

I was disappointed to see that DragonFly is still using XFree in the 1.2 release they just made. At least, the package repositories/ports tree it references are.

Re:Meanwhile Today On Earth... Today???

[justsomebody]

Today????

Your earth has wrong date set. Call your planet administrator.

Meanwhile, on my earth April 29 will pass by just normally. Just as nothing would happen.

Re:copied?

[mattyrobinson69]

as far as i can tell, its much more smooth dragging a window whilst using kompmgr (not sure about xcompmgr) than it is without.

If only the composite stuff would suddenly become stable enough for everyday use.

AGP on your 486?

[green pizza]

Unless you're still using a 486, you shouldn't have to worry about JPEG decompression using up your CPU cycles. It doesn't require that much power.

That said, I do wish libjpeg was faster and actually made significant use of SSE. Intel's optimized jpeg routines are way WAY faster.

Re:AGP on your 486?

[MikeBabcock]

In the last application I was writing for a client that used graphics routines I downloaded Intel's performance primatives libraries and they worked very well. Very fast, quite impressive. I'd love to see AMD release something similar and link against either dynamically from within graphics code.

Re:copied?

[As Seen On TV]

If by "borrowed" you mean "bought." Mac OS X was largely based on NEXTSTEP, which we bought from NeXT's shareholders for more than $400 million.

Re:Was I the only one?

[astrokid]

Wow. You must have a lot of enemies.

Flawed argument

[AvantLegion]

The problem with your argument: a desktop that is "really neat" in 1995 is not "really neat" years later.

1995: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
1998: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!
2000: We'll have really neat X11 desktops Real Soon Now(TM)! See, here's a demo!

All of these have been met. Maybe not as timely as would be nice, but met. What you don't seem to understand is that "really neat" is a moving target.

E17 eh?

[BuBu_]

I always planned to be playing Duke Nukem on my E17 desktop running on GNU/Hurd.

Re:E17 eh?

[root_42]

I always planned to be playing Duke Nukem on my E17 desktop running on GNU/Hurd.

Oh man, now even the running gags are presented wrongly. It should be Duke Nukem Forever. ;-)

Advanced graphics programs?

[starseeker]

Window manager effects and all are nice, but the part I find interesting is whether Gimp et. al. will be able to more easily impliment things like layers and transparency now. Anybody know how that would work?

X will forever be X...

[jcomeau_ictx]

Bitmaps are slow, no getting around it. Is anybody ever going to resurrect NeWS? I don't know enough about legal matters (zilch, actually) to know if Sun, Adobe, and/or Xerox would have to agree to it, or if people could just hack up something similar, as the ghostscript people did with the postscript language.

Re:X will forever be X...

[argent]

I've exchanged email with some of the people who actually may have some of the source to some version of NeWS, they're not sure. It's kind of gotten lost, and they don't really think it's interesting any more. It's a shame, I was really pumped.

Re:X will forever be X...

[arethuza]

I did quite a lot of work in NeWS using HyperNeWS and it really was one of the nicest graphics environments that I have ever used. Maybe not the easiest, but certainly the most elegant.

Maybe building something similar in SVG would be the way to go.

Re:copied?

[AKAImBatman]

You forgot the Dock effect, which DOES actually improve usability. There's also a neat dropshadow trick that improves the user's ability to visually figure out what window is active. There's also the Expose effect which helps you dig out the window you want from everything that's open.

In other words, OS X leverages its OpenGL abilities in a rather sneaky fashion. By doing this, it improves the overall experience without screaming the features in your face.

Re:Too Late - I've already bought a Mac

[SirTalon42]

Hes admitting that Unix is God.

did someone say Berlin Project?

[displague]

Or was that Fresco?

Either way, the website hasn't been touched in two years...

Re:copied?

[nickos]

Why don't you do something instead of moaning. Off the top of my head I can think of loads of window managers that are trying new things.

Bloat, bloat, bloat, wafer thin mint, splat...

[suitepotato]

Imagine if you will combining all of Object Desktop's c00ln3ss with the very heart of Windows XP if you will. Simply imagine a heavily modded cybernetic sloth pumped with downers that's not been on the charger in two days.

Now imagine it can actually be named in the same sentence as the phrase "open source".

Now stop imagining because it's almost here. We've gone from "small is better even if it is not correct" to "small is better and we should be correct too" to "large is okay even if it is not correct". So it's neither small nor correct. The two chief winning attributes over Windows and where we were headed at one point.

Proof that the attitudes of Windows coders and designers and users can also be found in the *nix community: whiz bang flash and glitter over the core tasks at hand is cool and better than making it work solidly and stably as a computer.

Okay, so how do I get some eye-candy

[moonbender]

I'm probably going to wipe off XP from my laptop RSN. I've already got Ubuntu on it, but I'll probably re-install from scratch anyway. Not sure which distro yet, pending any other convincing arguments I might just end up re-installing Ubuntu. But that's beside the point.

All those demos are nice and all, but are there any usable ways of getting cool eye-candy in a working, moderately stable Linux install today? Without all the hassle of checking out code from a VCS? Is Enlightenment a sensible choice for an install that should primarily just work? For instance, I'm going to install OpenOffice and to stuff for the university on it - is working with OOO better supported in Gnome or KDE than in E, or is there no difference? I like some eye-candy (if it doesn't get in the way, XP-style), but it's no use if the prerequisite is a system too geeky or unstable to do any work on.

Re:Okay, so how do I get some eye-candy

[veg_all]

Well, since you're distro-shopping anyway, try here.

Actually, just insert your package manager controls in place of "emerge," and it should be applicable in ubuntu or whatever.

Re:Okay, so how do I get some eye-candy

[poofyhairguy82]

I hear you, I love me some eye candy. Its quite possible in Ubuntu. Currently in my Ubuntu desktop I have wonderful fading and drop shawdows. The only time I experiance bugs is when I use the logout applet in Gnome or when I try to do things that are opengl (such as screensavers or games). You have to pick one at a time, eye candy or games. I use have two different user accounts- one for games and one for desktop use (eyecandy). Gnome doesn't official support this stuff, but other than the logout thing it works fine. KDE 3.4 is actually very compatible I've heard, but I only use Gnome and XFCE.

Next you must find out if you have the correct hardware. The best case scenario is that you have a newer Nvidia card. Any modern (at least a geforce 2) Nvidia card can do the coolest eyecandy trick ever- fading! Each window fades into whats below it. Its hard to describe, and it cannot be expressed with screenshot or low res movie. Let me say that fading is worth every penny of a new Nvidia 5200fx card I bought just to be able to do it.

If you have a newer Nvidia card (anything more recent than the geforce 4 series), you can add a sweet trick to the fading- drop shadows! I personally believe these are overrated, but many people LOVE them. To get any Nvidia card to do this fun stuff, you must have the driver installed (trivial to do in Ubuntu) and you must get the driver to accerate composite (I give instructions how at the end).

If you don't have an Nvidia card...well...think about buying one. Mesa drivers and ATI's flgrx drivers can do Drop Shadows (it takes a performance hit to do it though), but nothing but an Nvidia card can fade worth a damn. As I said, some people like the drop shadows, so try them out if you want. Anything faster than 1ghz and its almost free eyecandy. But-in the long run- if you want the best eye candy Linux has to offer- you need an Nvidia card. ATI is the devil.

Directions of how to do everything (including how to enable acceration with Nvidia drivers) can be found here. Have fun!! I know I do...

Re:Okay, so how do I get some eye-candy

[moonbender]

I'll check it out, but the laptop I'll be doing this on only has the "extremely" integrated Intel GPU, I'm afraid. And I haven't figured out how to cram an Nvidia PCI-E card into it, yet.

Re:Okay, so how do I get some eye-candy

[poofyhairguy82]

That "extremely" integrated Intel GPU" might actually be better than an ATI card. The Luminocity devs use an old Intel GPU to demo their cool wobble effects. If it can do that, it might can do some other cool stuff. Google is your best friend on that one. Something like "xcompmgr" and "Intel."

Re:copied?

[justsomebody]

Ok, let me bite this one. I was talking about 3d effects done as they should.

Dock effect: Try this
enable dock effect (very small dock, large icon when mouse over). Start some job. Move mouse over dock left and right. Measure time neaded to finish job with or without this. If this would be implemented over 3d there wouldn't be a difference of 300-400%. This is obviously pixbuff effect not 3d.

Drop shadow:
Same goes for drop shadow effect. try moving over menus left and right, although CPU load is smaller. Job is smaller too, then again.

Expose:
Ok, I give you this one. It is done as it should. Even video plays when minimized without loss of performance

But if you were only talking about visual effects (not limited to 3d and their 3d implementation), then I agree with all mentioned features.

btw. I tested this just now on my 10.3

Isn't it too complicates

[wereHamster]

Is it just me or do you also have the impression that the whole X11 server architecture is way too complicated.

Xorg (the xserver), dri, drm, kernel modules, Xgl, mesa.

What amazes me, Xgl rund on top of the xserver (Xorg), because theres something with dri that Xgl can't do directly.

I've created a simple (and by simple I really mean simple) 'xserver': the basic idea was to take an existing API and build the 'xserver' on top of it, the API is OpenGL. Windows are special objects in memory that can be shared between the client and the server. The client creates an window and tells the server 'hey, there's a user and he/she wants to see this window, please put the window with ID xyz onto the first monitor so he/she can see it'. and the server loads the object and puts it into the frontbuffer. and the client can draw (write) to the window and the server reads from it. and both the server and client use the same API (OpenGL) to draw things, the server into the frontbuffer, the client into the window object. Of course there's a tiny API for handling these objects, but that's only very few functions, maybe 15, that's enough.
And it works. Now if someone writes a driver that makes use of the GPU/video ram, this could be really fast. Currently it supports the mesa software opengl library for rendering.

Re:copied?

[Imazalil]

This is one of my 'problems' with open source. Generally it feels like everything is a copy of windows/os x. (yes, I know there are a ton of projects underway, but nothing too mainstream) It's great we're getting transparency, fancy window effects etc, but really we're just copying os x, and a bit what longhorn will bring to the table.

We can code, no question, but what we need is a vision of what the computer is that goes one step ahead of os x/windows for people to take notice. Right now we are just sreaming 'me too' os x has nice transparency, me too! longhorn will have bland animations, me too! We need to get one step ahead, so that we can say, yeah os XI stole that from us, that's right, longhorn xp 2013 did copy that from us.

Im.

Re:Meanwhile Today On Earth...

[RealAlaskan]

>> come by any Apple Store and pick up a mini

>This is illegal where I live. Here, we have to give money to a store ...

Actually, you can pick up an ipod mini in a store where you live. But, buddy, if you don't have the bucks, you'd better put it right back down.

Plan 9's 8 1/2 Windowing System - Light and Fast

[billstewart]

Rob Pike gave a talk at a Usenix a decade or so ago about his 8 1/2 windowing system for Plan 9. His basic intro was "Ken and I spent a decade figuring out what things a windowing system shouldn't do and wrote one that doesn't do them." It was something like 64K lines of C code, and when he typed 8 1/2 (in Unicode :-) into a shell and hit return, he had a running window system up in about the amount of time it normally takes to get a $ prompt back - it was way sub-second. The UI he was running on it was his Acme tiled window manager - I much prefer systems with overlapping windows (but he'd already written one of those for the Blit, so this was new usability research for him.)

Why I want what I want

[RealProgrammer]

I want to have several terminals running 'tail -f' on my log files, each turned so I can just see the shape of the output and maybe read it a litle. This gives me more information than just that some file changed - I can see a little bit of the info without having to know the details.

I want to be able to leave my log files in a little group and look at them when I want to go look at them.

I want to have several processes running on a particular machine and leave that group in a little village while I go look at other things. Alternatively, several processes related to a particular task (e.g., editor, compiler, test files) all grouped in a village, independent of which machine they're on.

Splitting a screen into two vertical halves, or using two monitors, allows for true 3D.

There are LED screens coming out soon (I think) with the pixels laid out on a fanfold or accordion arrangement, such that your right and left eyes see only one side of the fold. In other words,

(left)..(right)

./\/\/\/\/\/\/\

The left eye sees only the "/" characters, and the right eye sees only the "\" characters. With the correct video driver and X server, you could have realistic 3D on commodity hardware.

Re:Why I want what I want

[glwtta]

I want to have several terminals running 'tail -f' on my log files, each turned so I can just see the shape of the output and maybe read it a litle. This gives me more information than just that some file changed - I can see a little bit of the info without having to know the details.

Sure, I do this all the time - code editor in left monitor, running app and trailing log files in right.

I want to have several processes running on a particular machine and leave that group in a little village while I go look at other things. Alternatively, several processes related to a particular task (e.g., editor, compiler, test files) all grouped in a village, independent of which machine they're on.

That's what logical/virtual desktops are for. I am not sure what you mean by "independent of machine" in this case.

Splitting a screen into two vertical halves, or using two monitors, allows for true 3D.

Depends on your definition of "true".

[snip] With the correct video driver and X server, you could have realistic 3D on commodity hardware.

And still no effective input devices. (there are other problems with that method, like the ridiculously narrow viewing angle, it's hard to imagine this working well on a monitor larger than 15" even if you are right in front of it)

My problem with this isn't the 3D display itself, though; but what that actually gives you. You mostly talk about logical grouping of applications, I don't see how a 3D interface would do this more effectively than a 2D one. Though it would be very nice if the existing 2D ones did a lot more with this sort of thing.

...erm, actually, YOU CAN

[x2A]

Slashdot - never let the facts get in the way of a good argument! I love it when people say that "xxx OS doesn't even support simple feature yyy", just because they don't know how to do it. Especially when it's as easy as this one!

Or is it so advanced that it requires that display #1 be on the left and display #2 on the right -- that is if I want the mouse cursor to appear in the right place when moving it from display to display?

Right click desktop -> properties -> advanced.

You see both your monitors there. You can drag 'n drop them around each other where they are physically located, monitor 1 can be to the left, right, above, or below monitor 2. Top left pixel on monitor 2 doesn't have to be by top right pixel on monitor 1. Just drag them around!

-2A

Why ?

[AT-SkyWalker]

Can someone please tell me why I would wanna Skew my windows the way they are shown in the screen shots ?

I can't seem to find any practical use for all these features...

Re:Why ?

[smash]

Because there's other things you can do with openGL in the X server. They're just a demonstration or "test" to see how performance is going, no doubt.

As to what GL would be useful for? Live previews on your pager for desktop switching (ie, a little mini copy of your other desktop is rendered there live, in as much detail as your pager is big enough to display), or hell - get rid of the pager altogether, and just make a scrollable, zoomable virtual desktop.

Kinda like working with images in whatever photo retouching package you care to name. Key modifier + Mousewheel up = zoom in, mousewheel down = zoom out.

The windowmanager BS they're showing off at the moment is a toy, nothing more - but by no means is it a demonstration of the full potential of a GL desktop.

smash.

Re:copied?

[labratuk]

The 'effects' are just examples to demonstrate the underlying architecture.

...should work

[x2A]

Dunno if gentoo does anything different, but to tell X to allow connection from a host, just run:

xhost +client_address
eg: xhost +192.168.0.1 -will allow connection from that IP address (needs to be run on the machine running the X server)

Appologies if tha's not what you were askin.

-2A

Re:...should work

[jericho4.0]

A 4 digit UID (user id, after his name) kind of implies that wasn't what he was asking.

Oh, also the lack of a question. :-)

Re:...should work

[vsync64]

I hope you're kidding. Try xauth, or in most cases tunnelling through SSH does that automatically and encrypts the X11 display stream.

And a lot of distros nowadays turn off TCP connections in the X server. There's a knob to turn in xorg.conf somewhere.

Re:...should work

[x2A]

I rarely use the feature anymore (use xvnc server now so I can reconnect and take my session around with me), and when I do, is only ever across private LAN's so encryption 'n secrets would be a bit dumb. (Plus haven't used distro's in years, so dunno how they set things up).

Sometimes the simple solution's sufficient. YMMV.

-2A

Re:copied?

[BigBuckHunter]

When OSX stop using X11? Is this something new in Tiger?

you should try a mac

[ashpool7]

They have these problems licked. I can reboot my PowerBook, put it to sleep, reconnect and disconnect my external monitor and it *always* remembers the correct resolution and position.

linux?

[chegosaurus]

You know, I have a feeling X might even run on some other OS than linux.

In fact, now I think about it, isn't X11 actually /older/ than linux?

Re:linux?

[Derleth]

You know, I have a feeling X might even run on some other OS than linux.

In fact, now I think about it, isn't X11 actually /older/ than linux?

Right on both counts, but so what? Nobody in the summary mentioned Linux, and the article isn't about how this will impact Linux. It happens to be in the Linux section, but I don't think Slashdot has a specific X section.

And realistically speaking, most people will be using this on a Linux machine of one form or another.

oh, also...

[x2A]

imagine moving one window over another. The X-server can do this within it's own process. If each app took care of it itself, that's two processes that needs CPU time to perform the same action.

And imagine the nightmare of transparency!

-2A

summary

[ImaLamer]

The report is including many screenshots and five videos.

Ja!

Re:copied?

[mrchaotica]

Huh? OS X has more than one 3D effect built in. For example, the various transitions you can use to switch between users (with Fast User Switching) or desktops (with Desktop Manager). Haven't you seen that cool "cube" transition before?

OpenGL Desktop is great but...

[shaitand]

If it ever got here the OpenGL based xserver would be great because of the speed increase of offloading the desktop overhead the GPU. If you would be so kind, please keep the eyecandy and just let me keep the speed/responsiveness increases.

What I would really like would be an input device, not an X server. How about an EEG coupled with a neural net to send keypresses and mouse movements to the screen? This way applications do not have to be redesigned and our brain takes care of hotkeys automatically. Much in the same way an experienced typer no longer thinks a-n-d but rather just the word "and" the brain macros the keystrokes.

Re:OpenGL Desktop is great but...

[smash]

As an X user, you should know that all the zooming windows, etc is determined by the window manager, and not the X server.

All an open GL X server will do is make it EASIER for a window manager to accomplish such feats. P. smash.

Re:OpenGL Desktop is great but...

[shaitand]

Yes, that is to state the obvious. What I am saying is that I would appreciate the xserver much more of the window managers SKIPPED the additional eyecandy bloat they could blow the newly gained speed increases on.

Re:copied?

[shaitand]

damn and here I thought the OS was almost entirely borrowed from BSD and then apple slapped a GUI on top of it.

Useless Vs Useful fx (and other improvements)

[master_p]

Am I the only one that turns off GUI effects like windows zooming in and out, menus folding and unfolding etc after about 5 minutes of use? From the screenshots in the presented link, we can see zooming windows effects and transparent windows. Where is the usefulness in that? it's still down to working with xterm and the apps like the Gimp (with all the (possible ?) usability problems that it has).

The useful effects are:

a) window shadows; it really enchances the depth perception;
b) zooming from and to icons; it really gives a sense of connection to the source for each window;
c) transparent notifications (for example when new e-mails arrive)

I don't think X-Windows need more effects than the above.

But what the X desktop really needs is the following:

a) a way to programmatically specify new server-client protocols in order to minimize round trips. For example, an application's GUI could live entirely on the display server, and the application is simply reduced to using the available protocol to build widget trees. This can also be used for rich WAN applications, including the internet.

b) a widget toolkit endorsed by the X-Window committee (whatever that means), that comes as default with X, is simple and easy to use, using one of the new protocols specified above.

c) the ability to do macro-commands, either by recording them using mouse and keyboard or by entering them via an X11 script language.

The above will make a killer combination...if coupled with an information-management application like TreePad as the desktop shell, then X11 will be a true winning desktop environment.

The Joys of Permanent Beta

[tekcsound]

Wow, so in five years maybe X will have half the features OS X had 2 years ago! ...only minus the standard interface design that makes sure "preferences" is always in the same place.

I mean, seriously, how long have they been showing off E17? And isn't it odd how E14 had cooler themes than E17 (which just looks like MetaCity with cool FX)?

When will the *nix community catch a clue and go the OS X route: Write a coherent, well thought out UI that takes advantage of all our 3d chips... and throw an X server in just for backwards compatibility.

Oh wait, wasn't that supposed to be Fresco? Too bad it went the way of the dodo. Apple created Aqua by souping up the Display Postscript in NextStep. Maybe the *nix world can write a new UI by souping up the Display Postscript in NeWS?

On a positive note, Looking Glass does look rather promising.

Re:The Joys of Permanent Beta

[Slashcrap]

Wow, so in five years maybe X will have half the features OS X had 2 years ago! ...only minus the standard interface design that makes sure "preferences" is always in the same place.

Yes, but what you Mac zealots tend to forget is that I've got an extra $129 to spend on cocaine and hookers whereas you gave yours to Apple. Of course, Steve Jobs will spend it on cocaine and hookers on your behalf which probably makes you just as happy.

Re:The Joys of Permanent Beta

[tekcsound]

Actually I'm a Linux zealot who hates X. Since no good replacement is in sight, I just made a compromise with Steve and keep a Mac on my desk.

Coke makes me too paranoid, so I don't mind giving him my share. And since OS X is still cheaper than Windows I have enough spare $$ leftover for hookers. Plus, since I'm not having to configure infinite config files (on the desktops at least), I can spend a little more time with them!

Actually it's not too bad of a business model if you think about it. You're not paying for the OS, since Darwin is OSS. You're paying for Apple to make a coherent experience for your desktop users.

I'd pay Redhat/Novell/Sun/etc. a little $$ if they could work together as nicely as Apple does. Then maybe I could finally get our entire shop running the same OS. Until then, I'll have to settle for Linux on the servers, Windows on the desktop, and my little Mac for me.

Re:copied?

[Senjutsu]

While OS X has an X Server available for it, its primary GUI has always been implemented with its own GUI system that had nothing in common with X11.

Mobility, persistence

[adam_j_bradley]

As a consumer of X the issues I face most regularly revolve around session mobility and perisistence. Sure I could use NX or VNC Session Manager (http://vncsessmgr.sf.net) but I'd rather a native X method of accomplishing the same.

Re:copied?

[As Seen On TV]

Isn't it neat to learn new things?

Re:English Link

[Tharkban]

It's really not that interesting. Go look at the pretty pictures on the site, that's what I did, even though I read German. :)

And on that point. Is anyone else bothered by the fact that "neat" warping and transparency effects are being applied to windows and those same windows don't have a correct looking see-through rounded corner?!

Slow Zone

[argent]

And The Blabber.

And all Banks' Culture books I can get my hands on.

And Greg Egan's stuff.

Re:copied?

[bani]

jesus christ. osx's dock effect is one the crappiest and most useless effects ever. a simple highlight/outline/arrow would suffice. instead, you have a crazy dock with icons changing size and sliding all over the place.

this is not even counting how crappy osx's dock is overall compared to say, kde or gnome's panels... functionality wise, osx's dock is horrible.

and the dropshadow effect is near useless, considering that ALL windows have dropshadows. the active window's dropshadow is slightly more pronounced than other windows, but it's only really noticeable when you have windows overlapping each other. the easiest way to tell which finder window is topped in osx is to look for the colored buttons. of course if you're colorblind, you're screwed.

interestingly enough, the use of color as 'active' indicators is a violation of apple's own gui design rules from original macos.

aqua is a step backwards in usability in many ways. apple favors eye candy now over usability. and to think apple users criticize microsoft for the same thing that osx is guilty of now...

expose is needed because osx doesn't have a pager by default. virtual desktops is much better for productivity than juggling tons of windows in a single crowded desktop. in that case, expose is better than nothing -- but only barely.

and now the legions of rabid apple kool-aid drinkers will flame me and mod me down into oblivion.

The NT gui:

[Ayanami Rei]

Someone correct me if I'm wrong here...

CSRSS.EXE is the equivalent to X.
Only in this case X is using all kernel drivers (framebuffer driver, GPM/HID-style input).
The framebuffer driver if you will is a combination of WIN32.SYS and whatever components are provided by your video card.

Each windows process with an application window makes (what I understand) to be a kernel call directly into WIN32.sys to get a handle so it can talk to CSRSS.exe and register itself to receive events from there, and to post things. It uses some kind of IPC that is abstracted by the linked-in portions of the Win32 API in user space.

Not only does CSRSS handle drawables and events however. It also supports the creation of user threads, spawning processes and syncronozation objects and some other things I can't remember.

Anyway the important thing is that it's not magic. It works just like X11 does just that the API doesn't have the option of being network transparent.
The whole winsock thing was added "on the side" and has nothing to do with these APIs.

it has been done

[cahiha]

You can use shared memory as a transport, just like any other. It turns out not to be any faster than UNIX domain sockets.

People have also hacked Xlib to basically interface to a drawing library. But that just turns out to be not very useful (and it isn't really fast either).

Perhaps better IPC, but never loose the net.

[Vengeful weenie]

Solaris supports an IPC mechanism called doors, which I know was ported to Linux. The basic concept is that the system can handle threading through the mechanism. According to Solaris docs: "offers a fast reliable synchronous RPC mechanism between processes on the same host and between the kernel and a user space process." This leads me to wonder if having a plug-in to replace the std socket interface with a doors implementation would yield better throughput. (I would never trade the small gain in efficiency for losing X's networking ability. It's not even a debate in my mind.)

facts and fiction

[cahiha]

I think the real problem is, everyone knows X Windows is broken, but nobody knows what to do about it.

There is nothing to be done. Despite the persistent put-downs from Apple and Windows fan-boys, despite attacks from every proprietary vendor of window systems under the sun since day one, the X Window System (which is actually a protocol) has outlived NeWS, SunView, NeXT, MacOS, and just about every other window system there ever was. It is supported by many vendors and has multiple independent implementations. Chances are, X11 will be around long after OS X, Aqua, and Quartz will be dim memories.

As a Mac OS X user, a previous Windows user, and a current Linux Desktop user, I will not be the first to tell you that X is slooow.

On comparable hardware, X11 runs rings around both Windows and OS X.

Lastly, the problem comes with there being absolutely no good drivers available. Honestly, even though NVidia/ATi tries, they're not up to par with what they've got on the Windows platform, and Apple developers have had the luxury of seeing the developer's specs, so their drivers are just as impecible.

X11 isn't just XFree86. X11 has had commercial, hardware accelerated drivers long before OS X even came out.

Re:Next generation, you say?

[Guy Harris]

Does next generation mean it'll support copy and paste?

To what does "it" refer? Copy and paste is largely a matter of toolkits supporting the ICCCM conventions for the PRIMARY and CLIPBOARD selection and the X clipboard conventions, and, for non-plain-text selections, supporting appropriate conventions for non-plain-text items (or creating them if they don't exist).

you must be kidding

[cahiha]

So in another 4-5 or more years X will have the same stuff that OS X has had for a while?

You are confusing visual fluff with functionality. The fact is that X already has more than OS X will likely ever have.

This highlights the problem with Opensourcesoftwaredevelopmenten.

It took Apple until the 21st century to offer reliable multitasking, a reasonable kernel, and a decent file system, and they managed to do that only by (1) making extensive use of open source software and (2) buying the stuff outside. Apparently, Apple has trouble doing the "un-fun stuff" if it takes them this long. See, that's the problem with closed source.

major protocol version

[cahiha]

The "11" is the major version of the protocol; it changes if the protocol changes incompatibly. If the protocol doesn't change incompatibly, then the major version doesn't change. Right now, it doesn't look like the protocol will ever need to change incompatibly anymore, so X11 it is.

Re:copied?

[tokabola]

I used windows for 9 years, and don't remember ever seeing anything even remotely resembling Enlightenment desktop, with multiple and virtual desktops. Not only is it sweeter eye candy than anything windows put out, it also revolutionized my workflow and made me much more productive.

I've noticed that the same people who are complaining about Linux's "lack of originality" are the same people who complain that "Linux can't even do [insert MS Windows gimmick here]".

Tommy

MOD PARENT UP Re:Why isn't this already out?

[Derleth]

The design of X and Linux might not be the best of all possible designs, but we use them because they work, and nothing else does. For all the bitching about X's network overhead, I, for one, could not live without it.

This should probably be posted in large, red text right at the top of this discussion. If I had mod points, I would be modding this up right now.

The people who think X's networking is slow and inefficient need to realize that, first, on most machines it isn't noticeable, and, second, it's vital to making X usable in a lot of cases. X's networking is the difference between being able to run a graphical program on a remote machine and being limited to running text programs.

Frankly, if X wasn't built around the networking model it would be necessary to reinvent it at a higher level, thus making it orders of magnitude less flexible and, ultimately, less usable. That ignores the inevitable standards wars that would surround any such reinvention process, probably leading to dozens of subtly and individually broken ways of putting a window on a remote machine.

The people who focus on creating a GUI for a desktop machine that only ever runs local programs aren't creating a credible replacement for X. They're trying to reinvent the 'graphical shells' of the 1980s, back when DOS and the original Macintosh ruled the microcomputer world.

Re:Open Source philosophy

[tekcsound]

Finally someone who isn't too high from their overclocked AMD's fumes to realize how outdated X is.

Something of this magnitude would probably require cooperation from all the big-name vendors, otherwise we'd have 50 forks of the same code because one guy wanted purple icons and the other wanted green.

Sadly, they've all put their resources into adding more bloat to GNOME... the equivalent of painting your AMC Gremlin and hoping it'll turn into a Lexus.

Unless there's a GUI-obsessed Linus clone out there, we're probably stuck with X for quite some time.

Re:They call this 3d?

[tekcsound]

http://www.sgi.com/products/visualization/realityc enter/

Yours for a small fee.

Re:X11 must die

[fvwmfan]

Yeah, well when you've come up with the code that actually does what you're talking about (won't bother rehashing that old tripe about bloat and speed), let us all know.

In the meantime, I will continue to use something that works, from people who know what they're doing, rather than listen to the half-baked thoughts of an ill-informed lazy arabica ("too archaic to be fast" - yeah, like 'ls') . X11 will die when something better replaces it (.. and everyone stops using X applications).

Joystick!

[Roadkills-R-Us]

There are plenty of ways to build a joystick that lets you easily navigate in three dimensions.

And in fact, it *can* be a useful paradigm, if it's done right. But I'm not sure how useful it is without a VR helmet, or some other 3D visual implementation.

Would that be better than 2D? It deneds on the individual and the application.

Was there at one point

[Roadkills-R-Us]

I recall seeing examples of just this during the 80s at the X Conferences MIT held. But it either never got out of the labs (Xerox, I suspect), or the people doing it got reassigned.

I really wish I knew why it never got fed back into the core libraries.