Qt On DirectFB

Ashcrow writes "The feasibility for DirectFB to replace XFree86 just a little stronger thanks Maurizio Monge very first alpha release of Trolltech's Qt library for use in DirectFB. You can check out some screenshots or go straight to the source. And yes, it has been released as Free Software."

Comments

Good start, but not useful yet

[lakeland]

Now I guess we get to find out how much KDE assumes X11. Because there aren't many QT only apps out there.

Re:Good start, but not useful yet

[puetzk]

many KDE programmers have, because of QT/Embedded (and the zaurus).

Konqueror, KHTML et all already have releases built to not need X (kdenox cvs module).

Now, for the more desktop-ish apps this is certainly true, and X11 usage is (unfortunately) rather sprinkled about.

The biggest single piece is probably replacing kwin, followed by the dcopserver.

Before all the flamers get in.

[mindstrm]

Consider this: What do you really NEED X for. Try to think bigger than unix for a minute.

Yes, X has remote display. That's a really useful and flexible feature in some situations, no doubt about it. And from a technical point of view, it's extremely elegant.
In reality, though, to a great many linux users, it's a neat trick that you don't necessairly NEED.

We use QT or whatever and try to design desktop systems (KDE, Gnome) which really just use X as a way to load up graphics primitives... those same systems could equally work on something else, with some great benefits in terms of speed.

From a GUI perspective, if you use all KDE apps, for instance, things have a very nice consistent feel to it. Same with gnome. When you start mixing things, plus mixing in old X apps, you just detract from an overall experience.. so let's come out with a fast, standard display system taht's NOT x.... and use X rootless for those legacy applications we need.

Re:Before all the flamers get in.

[tomstdenis]

so first port motif, pango, gtk, etc... to DirectFB and then you can say "drop X, use DirectFB".

I mean its like saying "drop linux kernel, use QNX kernel" :-)

Tom

Re:Before all the flamers get in.

[mindstrm]

I think you are missing the point.. we aren't saying "This is a drop in replacement for X" .. it's NOT X. I'm saying, to build desktop GUIs, we don't necessairily need to use X as a base.
Yes, that might mean that only apps written for that gui would work.. but that gui could be, say, QT (as the article is about) or something else.

See OSX for an example. Can I run X apps? Sure, if I fire up the X server. ANd they run just how you expect them to, they look the right way, and everything... but the apps that work really well use the native display library, not X.... and they work even better. And no, it's not because the X server sucks, in fact, the X server is quite good.

We are adding so much stuff on top of X we have to question if we really need what X provides, and if it can't be better provided better another way.

Re:Before all the flamers get in.

[Natalie's Hot Grits]

"The good news is that they're not going to do that. If it ain't broke, don't fix it."

Well, considering one of the founders of the XFree86 Project (and board members) has said that the *nix desktop needs to be replaced by a direct rendered model with an X interface on top of it (exactly the same thing but with faster local rendering at the cost of nothing) I would like to call bluff on your "If it ain't broke, don't fix it."

Because according to (at least one) XFree86 founder, it is broke, and it does need to be fixed.

Re:Before all the flamers get in.

[MbM]

When you do run a QT or GTK app over remote connection, the remote (server) library renders the widgets to x primitives which are then sent over to the local (client) computer to display.

Why is it that (to my knowedge) nobody has done this at a higher level? If the client already has a QT library why not simply send over a 'draw widget' command to that library, creating a proxy out of the server's library.

Re:Before all the flamers get in.

[EvilTwinSkippy]

Network transparency is a beautiful thing. I admit, my needs are a little exotic, but I happily run my computer from several dumb terminals (stripped down laptops).

Why maintain a stable of computers when you can have one ubermachine (and of course a few cruddy ones for DNS and webcaching.) The wife has a copy of Win4Lin for Quicken and Office. And I never have to worry about being booted off the "good" computer.

Hell, with my cable tuner in the big computer I can actually watch TV over the wireless. That is of course, if I had cable. I'm practicing living on Internet and DVD's alone. Apparently I missed something called "Reality TV."

Re:Before all the flamers get in.

[slyall]

I would not be happy to give up remote display. Not going to happen.

There is no reason why you can't run a X server over the top of a directFB desktop. This would enable applications that support the new system to run fast locally while X based Apps (remote and local) can take to the X server.

There are plenty of X servers for Windows and MacOS that plenty of people use already.

Re:Before all the flamers get in.

[Arandir]

In reality, though, to a great many linux users, it's a neat trick that you don't necessairly NEED.

In the midst of citing reality, you're ignoring reality. First of all, Linux isn't the only OS that uses XFree86, X11R6, or another X11 based windowing system. Heck, it ain't even the only free OS that uses it.

Second, even supposing Linux will achieve it's goal of "World Domination", where everyone must use Linux or be branded a luser, it's still ignoring the fact that Linux is a Unix-like operating system, and to confine it to only the home based game machine is to deny it 95% of its potential.

Third, that "neat trick" doesn't cost you a damn thing if you don't need it. The only thing holding back XFree86 performance is the fact that it must operate in userland.

I've heard the phrase "why keep it of 95% of the people don't use it", referring to the remote network capabilities of X11. Well, why not turn that statement on its head? Why support SMP in Linux, if 95% of the users don't use it? Heck, why do I need snowchains for my car if 95% of the time I won't be driving in snow?

Fact of the matter is, most people using Linux, BSD or UNIX outside of the home will want and need the networking capabilities of XFree86. If you want Linux to be confined to home game machines, then go roll your own distro. But in the meantime a lot of us want the capabilities of XFree86.

Re:Before all the flamers get in.

[Anonymous Coward]

People have done things like that, and it sucked. The reason is that it requires any remote machine to have a compatible version of Qt and GTK and every other damned toolkit someone might want to use. This is an administration nightmare at best, and impossible at worst. (What if you have two apps, one on each machine, and they both require incompatible versions of some toolkit? Then you simply CAN'T display either app remotely without jacking up the other app, or adding some overwrought version management system. What if some app you use requires the very latest version of its toolkit for some interesting new widget it adds -- do you really want that to force an upgrade on every machine you might want to display that app on?)

To make this a little more concrete, let me make a little analogy. Imagine you have a web site, and you want to push more processing to the client. So, you change your dynamic PHP code into JavaScript code. Since you now no longer have the power to just say "this must run under version X of PHP, so we're putting that on our server", you have a choice: either make your app run in every javascript implementation that you might encounter, or place restrictions on what version(s) of javascript can be used to visit your site. If you had left it in PHP, there would be some disadvantages (slower, etc), but at least you'd know it would work with any browser.

Now, naturally, the "any browser" thing is a myth. There are quirks with different HTML implementations too. But this brings out an important point. Somewhat unlike HTML, the great thing about X11 is that it does something really simple. It doesn't concern itself with complex details like what a doubleclick means, what a button is, etc. It stays at a nice low level. As such, it's more easily possible to reach a point where it's matured and it's stable. The X11 that some person has on their machine will be compatible with everyone else's, by and large, because it's at a low level and can be stable and mostly unchanging.

So, while X11 isn't a paragon of simple, elegant, modern, clean design, it or something like it (with the same goals of sticking only to universal primitives, but maybe also with support for changing resolutions, etc.) is a very beneficial thing for the reasons similar to why Swiss banks are a beneficial thing.

Re:Before all the flamers get in.

[caseih]

Isn't this backwards, going back to a dumb (albeit accelerated) framebuffer? X does much more than just push pixels. If you take a look at how Microsoft does their gui these days, you'll find that it's a lot more like X these days than a framebuffer.

I own a zaurus and was initially impressed with the Qtopia/OPIE user interface. That is until I hit that one design flaw: They write directly to the framebuffer. This means I can't mix and match gtk programs with qt programs. This means I can't develop any non-free apps at all, since QT is GPL and that's the only thing you can run on Qtopia.

As I disected QT/E, I found that it pretty much had to duplicate many things that X does well, like windowing. Yep. And event handling and exposure stuff. Personally I think the Qtopia guys would have been much better off using the mini KDrive X server and use a modified version of QT/X11.

As soon as I can, I'm blowing away Qtopia and not going to OPIE but rather to GPE, which is based on X. A much better solution. Check out their screenshots if you don't believe me how well X fits a handheld.

My point here is that I don't think this directfb idea gives me any more advantage than X does. Sorry. Furthermore, we'd just need an X server on top of directfb anyway to run our main apps, and that is essentially duplicating the drivers that X11 already uses to talk to the framebuffer.

The best improvement I think we could bring to X11 would be a special mode where each window is a live opengl surface. That way we wouldn't need to do "exposure" events and other things. Window dragging would be silky smooth. Other 3d effects could follow. Forget the frame buffer.

Re:Before all the flamers get in.

[buysse]

I don't want a whole bloody desktop with my remote X, thanks. I want a window. I want my emacs window, or my gvim window. (I actually do use both). I want an xterm, or a commercial product like ArcGIS with a single (fast) remote window. VNC is *not* an acceptable substitute for network transparency.

Re:Before all the flamers get in.

[eakerin]

And to a little more fuel to this one, using VNC requires a Framebuffer on the server(in this case the Farm running appliations, not the Display in normal X terminology), drastically increasing memory requirements.

With X the only memory needed on the server is the memory needed for the application itself, since all of it's drawing routines are done directly to the computer running the display.

Give it a try for yourself, Fire up the VNC Server on a unix system (Windows Users can't have 2 desktops on VNC anyway, so they don't get to complain), and check the memory usage for that, it should be approx Width X Height X Bits Per Pixel. (plus the overhead of VNC) Now compare that with the None needed for X, and I'll take X any day.

Re:Before all the flamers get in.

[SN74S181]

Yes, but I want to run the same apps on an X server on any other machine on my network. The directFB apps won't do that. Once we start down an 'applications bound to a single framebuffer' path we're all stuck using KVM switches, VNC, and similar kludges; or we have to plant our butt in front of each machine to run graphical software on it. And both options suck.

I have found that people who hate the X Window System are people who don't understand it. Many come from a PeeCee background and still don't get the idea that the network means more than 'that drive over there on a server' and HTML.

err

[Anonymous Coward]

What's up with all the "Hot Babe" backgrounds? Makes all Open Source developers look like horny teenagers. Do you want a horny teenager writing your production Apache server??

Wowsa!

[tds67]

The screenshot looks HOT!!! And oh, yeah, the desktop looks okay, too...I guess...

Sounds like a plan.

[Meat Blaster]

While I've grown accustomed to X-Windows' ideosynchronities, I've always thought that it would be a good idea to reengineer the whole system from scratch to take advantage of today's hardware and UI concepts. X-Windows 4 has been a vast improvement, but I'm talking about something more like OS X where the whole thing is rewritten to be very smooth and responsive to user input.

If this is a step in that direction, and it sounds like it is, I'm all for a decent alternative that isn't slowed down by having to be a swiss army knife. Especially if it makes resolution switching, 3D graphics, and direct screen drawing less of a hassle.

Re:Sounds like a plan.

[Snoopy77]

Yeah, I'm currently running X-Windows 4 on my Linux 9.0 box.

Re:Sounds like a plan.

[warrior]

I've always thought that it would be a good idea to reengineer the whole system from scratch to take advantage of today's hardware and UI concepts.

Good idea, I've thought the same thing. I wrote a GUI toolkit for X, and a window manager, so I've got a good idea of how the whole thing works. I quit working on it as I was frustrated that I couldn't do some of the neat things I see in OS-X on X (that sounds funny, doesn't it?). Soooo...

I started from scratch writing an OpenGL based display server. I'm using a lot of ideas from X, but throwing out a lot of cruft and adding lots of enhancements. All of the drawing is double-buffered -- no more Expose events!!!! :) All of the drawing is also hardware accelerated. I've figured out a way to do this very well, without context switching the gfx hardware. One possible method will allow many clients to draw at once and keep a constant framerate (by not context switching/swapping buffers within a certain timeslice, these are very costly operations).

Some of the ideas I am keeping are the idea of "internalizing" graphics buffers to the server where they can be shared among other applications. I'm also keeping the idea of a replacable window-manager like shell.

For fonts I'm using Freetype. Standard image format is png. The display is also hardware-resolution independent and colordepth independent. Right now I'm being setback by the fact that I can't get X working on my new laptop (anyone know a modeline for WUXGA+ 1920x1200@60Hz, for Compaq X1000?). For communication I'm using named pipes/shared mem.

So far, my numbers are better than these.

I'd also like to implement creating server-side macros so a client can pass one command to the server and execute a whole set of drawing routines atomically. Oh, and the source is definately going to be open. Any of this sound like a good/bad idea?

Cheers,
Mike

Re:Sounds like a plan.

[Natalie's Hot Grits]

Restrictions on smoothness and responsiveness to user input are due more to driver and kernel performance characteristics than issues with X itself

This is a myth, and an XFree86 developer, board member, and one of the founders seems to agree:
I've even pissed off Keith and many others on the Core Team by pointing out that X is obsolescent. I've been working in the Windows world for years now, and client-server display systems are utterly irrelvent to the majority of real-world computer users. X needs to be replaced by a direct-rendered model, on which a backwards-compatible X server can be reasonably trivially implemented.

Nobody except people who use X over the network extensively are making claims such as yours. There are many people who do extensive GUI research and programing contradicting what you are saying. the KDE and GNOME projects have both showed interest in direct rendering models. There is a HUGE project of people doing exactly what the above link says. Implementing a direct rendered GUI with an X layer atop for remote display. There is no reason that X should treat everything (including local rendering) as a network socket connection when it can talk to the hardware directly. It is just too much overhead.

People making claims that the UNIX SOCKETS for local display don't involve overhead haven't made their evidence available. if this is true, explain yourself. There is real world proof that the DirectFB model is faster for local rendering, and until XFree86 either gets its own direct rendering model built into it for 2d rendering, and all the bells and whistles that DirectFB has (alpha blending with hardware acceleration, desktop/screen resloution switching on the fly, etc), you people claiming X's faults aren't with the protocol and implementation but with drivers are all blowing hot air.

the unix desktop CAN be faster. But X/XFree86 either needs to grow with the modern desktops, or it needs to be replaced on the desktop with something that works better. Either way, competition is a good thing, and I'm glad that DirectFB is making some headway. Porting QT/Free edition to DirectFB is going to make this competition even better, and the users will win out in the end.

I'm not bashing X here. I'm simply saying that there are better methods to locally render 2D applications. They do exist. They are being developed. The X protocol and XFree86 was designed for UNIX in a client/server networked environment. This is not how most modern computers use on their 2D desktop. I'm not saying XFree86 or the X Protocol needs to go. But if it wants to be _THE_ unix desktop for everyone, it needs to take into account the growing popularity of pure-local rendering environments. (There is no argument from me against it being _THE_ unix desktop in a client/server networked environment)

Re:Sounds like a plan.

[IamTheRealMike]

This is a myth, and an XFree86 developer, board member, and one of the founders seems to agree:

As has already been pointed out to you, said XFree86 developer has been out of touch for so long that he thought Red Hat hadn't contributed to X, and now works on Windows entirely doing AOL stuff. When he took part in the XFree-forum list, he got flamed, badly.

Nobody except people who use X over the network extensively are making claims such as yours.

Don't be so short sighted. Maybe the reason nobody uses network transparency on Windows is because it blows so many goats? Have you thought of that? Yet, Microsoft have still taken architectural backflips to make it work, see W2K Terminal Services etc. It's a lame, poor imitation of actual network transparency, but they sell it anyway.

Simply having worked in tech support, I can think of MANY times when being able to have an xterm launched to me, would have been a godsend, especially as said person could continue working while I also worked.

There is a HUGE project of people doing exactly what the above link says. Implementing a direct rendered GUI with an X layer atop for remote display

Which desktop uses that then? Cos AFAIK both MacOS X and Windows use a model similar to that of X internally. None of them are direct rendered.

People making claims that the UNIX SOCKETS for local display don't involve overhead haven't made their evidence available. if this is true, explain yourself.

Why don't you find out for yourself, instead of ranting on Slashdot? You know, there are indeed studies, performed under controlled conditions, that show the overhead of UNIX domain sockets is negligable. They even tried replacing them entirely with SHM segments at one point, but it made no difference. Domain sockets are one of the most heavily optimized IPC primitives in the kernel, and you are quite free to perform speed tests yourself. The only area that it makes any difference is when throwing large amounts of data through them, such as pixmaps (which is why we use the XSHM pixmap extension), and the memcpy may not be completed in the available timeslice. For small messages, ie the bulk of X traffic, there is no speed gain to be had.

There is real world proof that the DirectFB model is faster for local rendering

Er, that's so wrong. For most people, ie anybody without a Matrox G series card, it's far far slower.

until XFree86 either gets its own direct rendering model built into it for 2d rendering

.... which it has....

and all the bells and whistles that DirectFB has (alpha blending with hardware acceleration, desktop/screen resloution switching on the fly, etc)

..... which it also has .... oh, unless you mean the ability to pointlessly make entire windows semi-transparent so you can't read what's on them, an ability that's useful primarily for screenshots with hot babes in the background

you people claiming X's faults aren't with the protocol and implementation but with drivers are all blowing hot air.

There are faults in every area of XFree/X11, nothing is perfect. The protocol needs some changes, which are being worked on, the driver interface needs to be broken to support proper save unders, and the scheduler is a dog. Needless to say however, DirectFB isn't a perfect work of art either. It's certainly useful, but right now it's not even competitive in terms of speed or features for 90% of users.

I'm not bashing X here

LMFAO. Yes you are. You've made many, many assertions, that you would know were wrong if you had actually sat down for a couple of weekends and done some basic research into the matter, like I have done. I have read boring reports, mailing lists archives, and chatted to various people involved, and so I'm pretty sure the impressions I have are accurate. There is nothing wrong with X as a local rendering mode

Don't be betting on it either way...

[Svartalf]

1) DirectFB supports GTK+ as well- I suspect Fltk's on the way as well.

2) You CAN have X apps under DirectFB with XDirectFB.

3) They're posting rather impressive framerates under Quake III:Arena with the DirectFBGL layer code.

4) Qt's ALREADY in the embedded space- QtEmbedded is what they're using on the Zaurus.

Re:Don't think so....

[DreadSpoon]

DirectFB has a multi-application core, and also a specialized X server that runs on it. You can run GNOME on it already, adding Qt/KDE to the mix only _increases_ the number of apps that can be run on it natively.

And so far as the "features" of X... the only feature X has that DirectFB doesn't is network independence, which very few users need, and those who do can use VNC or the DirectFB X server.

Mirror

[keesh]

Site is kinda slow... one, two, three, karma please?

DirectFB Inherently Insecure?

[istartedi]

Not being familiar with it, the first thing I did was read the FAQ:

Q: Whenever I try to start a DirectFB application, I get the error message

Error opening /dev/tty0
A: You have to be root to run DirectFB apps. The main reason is that only root is allowed to change virtual terminals.

So. In order to get the supposed benefits of DFB, you have to run apps as root? I guess maybe you could log on as a user and su the DFB apps, but that's a pain. Why should a graphics lib muck up security? That seems inherently broken to me. If it really just abstracts graphics then there should be no problem with user apps running it.

This isn't really my area of expertise. Perhaps there's something I'm missing. Can anybody clue me in?

Re:DirectFB Inherently Insecure?

[RelentlessWeevilHowl]

Your X server also needs root access, and for much the same reasons. X needs to muck with the registers on your video card, for example. Nowadays, there's a little setuid program called "XWrapper" that gets access to everything it needs, then drops its privileges and loads the main X server on top of itself.

There is at least one project (KGI) that attempts to rationalize all this. It puts the privileged functionality in kernel space, then exposes it all in a safe manner. Linus has not been receptive to this design in the past, preferring the X mechanism.

The best solution

[Anonymous Coward]

Plan 9's. The graphics, mouse and keyboard devices are standard devices that can be mounted from a remote filesystem; the advantage being the windowing system does not need to handle the network layer. And since each process has its own filesystem namespace, you have a bunch of different consoles and each program accesses its one at /dev/cons.

The entire system, including the default program that runs in the window the equivalent of xterm [Far89] with `cutting and pasting' between windows is well under 90 kilobytes of text on a Motorola 68020 processor, about half the size of the operating system kernel that supports it and a tenth the size of the X server [Sche86] without xterm.

The components of Plan 9 are connected by a common protocol based on the sharing of files. All resources in the network are implemented as file servers; programs that wish to access them connect to them over the network and communicate using ordinary file operations. An unusual aspect of Plan 9 is that the name space of a process, the set of files that can be accessed by name (for example by an open system call) is not global to all processes on a machine; distinct processes may have distinct name spaces. The system provides methods by which processes may change their name spaces, such as the ability to mount a service upon an existing directory, making the files of the service visible in the directory. (This is a different operation from its UNIX namesake.) Multiple services may be mounted upon the same directory, allowing the files from multiple services to be accessed in the same directory. Options to the mount system call control the order of searching for files in such a union directory.

8½ serves a set of files in the conventional directory /dev with names like cons, mouse, and screen. Clients of 8½ communicate with the window system by reading and writing these files. For example, a client program, such as a shell, can print text by writing its standard output, which is automatically connected to /dev/cons, or it may open and write that file explicitly. Unlike files served by a traditional file server, however, the instance of /dev/cons served in each window by 8½ is a distinct file; the per-process name spaces of Plan 9 allow 8½ to provide a unique /dev/cons to each client. This mechanism is best illustrated by the creation of a new 8½ client.

From here

Windows or holes in the walls?

[Yaa 101]

This reminds me of a long going project that was once called Berlin and is renamed Fresco along the way...
Though their ambitions were higher with making a new windowing system...

They still exsist at:

http://www2.fresco.org/

For all those whining about XFree86....

[DaCool42]

1) As many have said over and over, XFree86 is not slow. It runs great on a 486. Try using a faster WM.

2) Transparancy/hardware rendering. For some reason people think XFree86 needs to be tossed out completely in order to get this. Check out this interview statement from David Dawes (XFree86 developer):

David Dawes: There has been some work on a new rendering model for XFree86 that provides some more advance composition techniques (including transparency), this currently being implemented in software. For XFree86 5.0 we'll be investigating this as part of our review of rendering models, and seeing if a hardware implementation would not be more appropriate.

For shame!

[UnknowingFool]

How dare he exploit that young woman like that?! Does he really think that putting a picture of a young, attractive, nubile woman in a suggestive pose is going to attract the attention of a bunch of nerds?

. . .

Never mind.

X is not slow, some WM's for linux are.

[miffo.swe]

X in itself is very fast and pretty slick. Try yourself by kicking gnome/kde and trying OpenBox or some other fast WM. The difference on slower machines is pretty big.

I have a feeling that some n00bs confuse X with their Window Manager and Docks and Panels etc.

Re:GDK vs. GTK

[ambrosius27]

Not exactly. GDK is an *abstraction* layer with multiple backends, the X11 one being the most prominent. To quote from the GTK/GNOME developers' website: "Instead of directly building on top of the X Window System, GTK+ introduces an intermediate layer, GDK, which isolates GTK+ from the details of the windowing system. This simplifies things for the programmer and increases portability." See the webpage. Through GDK backends, GTK has been ported to MSWindows as well as DirectFB(see also here).

I hope that helps.

Re:The *Real* Problem with X11 and the unix deskto

[John Meacham]

um. actually X selections are more powerful than other systems at allowing cut-n-paste and drag-n-drop of non-text. X selections let the pastee and paster negotiate on a prefered file format based on what they have to offer and what they can accept.

Just because people who write apps for X don't seem to use this functionality, don't blame X11. if the app writers are too lazy to use the power of X selections, I don't see why they would suddenly for some new system.