Vastly Improved Raspberry Pi Performance With Wayland

New submitter nekohayo writes "While Wayland/Weston 1.1 brought support to the Raspberry Pi merely a month ago, work has recently been done to bring true hardware-accelerated compositing capabilities to the RPi's graphics stack using Weston. The Raspberry Pi foundation has made an announcement about the work that has been done with Collabora to make this happen. X.org/Wayland developer Daniel Stone has written a blog post about this, including a video demonstrating the improved reactivity and performance. Developer Pekka Paalanen also provided additional technical details about the implementation." Rather than using the OpenGL ES hardware, the new compositor implementation uses the SoC's 2D scaler/compositing hardware which offers "a scaling throughput of 500 megapixels per second and blending throughput of 1 gigapixel per second. It runs independently of the OpenGL ES hardware, so we can continue to render 3D graphics at the full, very fast rate, even while compositing."

Comments

A really good demonstration of Wayland.

[Anonymous Coward]

This should be a really good demonstration of Wayland and why (and what) it improves over X.
Hopefully this would put to rest the flame wars.

Re:A really good demonstration of Wayland.

[Anonymous Coward]

I imagine performance would be much worse on older hardware that didn't have the same level of support, like ATI's X1*** range.

Re:A really good demonstration of Wayland.

[Guspaz]

The X1*** range is more than capable of doing basic 2D scaling and transparency in hardware... you'd have to go more than a few generations back from there to be bottlenecked for basic 2D work.

Re:A really good demonstration of Wayland.

[fast turtle]

I'm one of those boundless masses of end-users who's never had to forward an X-Session. When I do need remote access, I tend to use one of the free apps like "Go To My PC". They work and all I need to do is provide them with a link to the connection and the app does the rest. No farting about with a Remote Help session, nor dealing with RDP. Just like using MS/AOL/Yahoo messenger. Easy Peasy and does the damn job. So I ask in all seriousness, what benefit does X-session forwarding give me in the face of alternatives.

Yes I'm a Linux user (Gentoo Ricer "-O2001 -pipe"

I'm Sorry Bob, but that needs admin privelages - Do You want to continue?

Seriously, I need to check Wayland out if it's offering that kind of performance. It may make the difference in Wine and my few windows games. Most are playable but FPS rates stink at time and I suspect it's the x server giving me trouble but no tweaks are available.

Yes, let's bring that back

[Anonymous Coward]

The time when everything needed to be specifically ported to a machine to make it perform bearably or at all. How I missed having stuff not work without that extra length to go to.

Re:Yes, let's bring that back

[multiben]

Haters gonna hate.

Re:Yes, let's bring that back

[Anonymous Coward]

It's all about tradeoffs, and always has been.

Nothing has changed.

Either you write generic support which works everywhere and performs with mediocrity at best (e.g., standard Linux on a desktop), or, you optimize for a particular hardware platform and get more performance.

The thing with RP, is that it's a low-power machine, so the generic mediocre performance is pretty awful and you need to specifically optimize to make it usable.

Re:Yes, let's bring that back

[fuzzyfuzzyfungus]

The time when everything needed to be specifically ported to a machine to make it perform bearably or at all. How I missed having stuff not work without that extra length to go to.

On embedded hardware, that time never ended... And the rPi isn't really fast enough that you can just run in all software, or even with just the relatively feeble OpenGL hardware, and pretend.

Re:Yes, let's bring that back

[fnj]

Cheerleaders always gonna lead the cheer.

There. Did we get that out of our systems?

Re:Yes, let's bring that back

[epyT-R]

Things like low level OS frameworks and related drivers, which require low latency, high performance, and sane memory footprints, must be ported to the architecture in a language whose compiler/linker spits out native binaries. No python/java/.NET here, because the lower the hog is in the stack, the greater the impact on latency and performance it has.

Wayland is a perfect example of this as it sits very close to the hardware with a driver between it and each device. This concept will never change because at some point the software must speak to the hardware directly no matter how the hardware is designed. If anything, the decade of sandboxed apis are a big reason why we need gigabytes of ram and microwave clocked CPUs to do basically the same things we were doing with desktops in the 90s with acceptable performance. The current situation on desktops (regardless of OS) is a sloppy waste of cycles that could either go into greater performance or power savings (or both, depending). Clean, efficient code is not, nor should it ever be, passe.

Re:Yes, let's bring that back

[ThePeices]

The time when everything needed to be specifically ported to a machine to make it perform bearably or at all. How I missed having stuff not work without that extra length to go to.

On embedded hardware, that time never ended... And the rPi isn't really fast enough that you can just run in all software, or even with just the relatively feeble OpenGL hardware, and pretend.

Not to mention the Pi is only $35 and uses a few watts of power, you cant expect current laptop class performance for that price.

The OP ignores the fact that incorporating this tech into the major Pi distros and projects is only work for the developers of said projects, not end users.
End users just wait for the next software update, and then they get vastly improved graphics performance.

I fail to see what on earth is wrong with a major advance in performance to a specific piece of hardware.
I just smell the acrid stench of cynicism wafting from the general direction of the OP.

Re:Yes, let's bring that back

[Lumpy]

AMEN. Slackware with a custom compiled kernel on my laptop utterly decimates anyones Ubuntu install. To the point that people cant believe that it's linux running that fast.

The low grade dog-food that is the popular distros today are causing more harm than good with them being dog slow and broken all over the place for the sake of supporting everything possible.

Re:Yes, let's bring that back

[ADRA]

I'm not philosophically against clean fast code, but to your point my desktops are probably 98% CPU idle when doing a normal workload, and only really pick up when: Playing games, Playing flash, Doing a compile, Running a development server and testing. The age of low level fast optimization is all but dead. For a brief time during the smart phone revolution, pathetic CPU's were a bottle-neck, but with my N4, nothing I throw at it feels slow or choppy. It has 2GB of ram IN A PHONE. Sure limited spec and fit for purpose devices will need fast low level access to optimize, but that takes time, and quite often we're finding that hardware's faster and cheaper than wasting time optimizing for the apex solution.

Take your question again: In 10 years when our entire assortment of devices has as much horsepower as my desktop computer does today, are we really going to need significantly tight processing? I'd say the better long term solution will be making development faster and hopefully more expressive.

Re:Yes, let's bring that back

[sp332]

What kind of changes did you make to your kernel build? And what CPU and I/O schedulers are you using?

Re:Yes, let's bring that back

[smash]

Hardware is a lot cheaper than developer time. This trend is only becoming more apparent.

Re:Yes, let's bring that back

[Unknown Lamer]

Also, languages like SML and F# can theoretically (and sometimes in practice do) generate code that is a lot more efficient than C, and not tied to a specific style of Von Neumann machine. That, and garbage collection has complexity bounds whereas malloc/free do not, not having managed memory leads to a lot of pointless copying (and nowadays the memory controller is what kills you so deep sharing of structure is always a huge win), etc.

And those expressive type systems happen to save programmer time too. I was personally amazed at how quickly I was able to dive into a large SML code base not even knowing SML and actually make changes that were pushed into production and didn't blow up (after figuring out the magic decoder ring for type errors that is).

Re:Yes, let's bring that back

[fuzzyfuzzyfungus]

Arguably, the fact that this specific hack had to take place is a bad sign, just not one specific to Wayland, or even (particularly) to the rPi.

There are, attempts at least, at standardizing the interfaces for the sorts of features that this Wayland modification used to get better performance on the pi; but they certainly aren't anywhere near where OpenGL is in terms of adoption, and so the compositing and windowing modification had to be made specifically for the 'DispManX' API used exclusively on these Broadcom Videocore parts.

At least this isn't a situation where applications have to have much platform-specific knowledge; but it's always nice to keep platform-specifics abstracted in some standard way as low in the pile as you can.

Re:Yes, let's bring that back

[gagol]

Rabbits always gonna chew your wires?

Re:Yes, let's bring that back

[EzInKy]

So what is low level today was high level yesterday? Sure it is nice not to have to peak and poke everything, but it seems that anything above that requires buying into somebody else's idea of the way things should be done. So, in the spirit of thought and freedom, just what does "Wayland" bring to the table?

Re:Yes, let's bring that back

[phantomfive]

The age of low level fast optimization is all but dead.

I keep thinking that, but then keep running into situations where I have to optimize things. My coworker has been optimizing a piece of code for the last two weeks because our customers find it too slow, and this is on a 64-bit i7 with 16 gigs of RAM (some image processing stuff). There will always be things that need optimization.

Re:Yes, let's bring that back

[Anonymous Coward]

Rabbits always gonna chew your wires?

FTFY

Re:Yes, let's bring that back

[Gerzel]

At least until Moore's Law ends. Dunno when it will happen but assuming the continued survival of the human race there will come a time when our computers are not becoming more powerful with each generation.

For the short and moderate term you can risk relying on Moore's but in the long run all good things come to an end.

Re:Yes, let's bring that back

[UnknownSoldier]

Yeah I'm kind of wondering that myself. Hopefully the OP can give more details.

Re:Yes, let's bring that back

[Anonymous Coward]

Slackware with a custom compiled kernel on my laptop utterly decimates anyones Ubuntu install.

The kernel has very little to do with user-preceived 'speed'.

Re:Yes, let's bring that back

[Anonymous Coward]

Don't ever diss rpi without mentioning its price.

Why is it so hard for some people to forget they are talking about 35$ system? Yeah, ok: money and value are hard concepts for the mentally challenged. But, at least try.

Re:Yes, let's bring that back

[Goaway]

Low level optimization is far, far from gone. It's just what you need to optimize that has changed. CPU number crunching is no longer the bottleneck. Memory is, and drawing things on a screen is essentially mostly memory operations. And we have special hardware to handle this for us to try and relieve the bottlenecks, so now the optimizations is in how you use that hardware.

If you tried running a non-optimized CPU-only system on modern hardware, you'd go out of your mind because it was so sluggish.

Re:Yes, let's bring that back

[Goaway]

It's ending right now. Clock speeds have stalled years ago. Memories are running at crawl speeds compared to CPUs. We're just buying time by increasing parallelism now, but Amdhal's law is waiting around the corner to put a stop to that, too.

Re:Yes, let's bring that back

[smash]

How will Moore's law ending result in developer time catching up? Humans aren't accelerating in ability to deal with programming issues; compared to processing capability in hardware, human ability is flat or (many would argue) on a downhill slope.

Re:Yes, let's bring that back

[Anonymous Coward]

And you think the OS running on your phone is not heavily optimized or what??

Re:Yes, let's bring that back

[fisted]

So you don't understand Moore's law. Thanks for making that obvious.

Re:Yes, let's bring that back

[gl4ss]

it's more like 48 euros.

for a hundred you can get a choice of similar systems boxed with remotes and psu's..

Re:Yes, let's bring that back

[SimonTheSoundMan]

Clock speed, Moore's law, what have they got to do with computing power?

Wirth's law is you enemy.

Bremermann's limit is waiting for you Goaway.

Re:Yes, let's bring that back

[gr8_phk]

Take your question again: In 10 years when our entire assortment of devices has as much horsepower as my desktop computer does today, are we really going to need significantly tight processing? I'd say the better long term solution will be making development faster and hopefully more expressive.

Yes, yes it will be important to have optimized software in 10 years. Remember X was entirely usable (OMG I'm getting old) 20 years ago. That 150MHz Alpha workstation in the lab was amazingly fast, but here we are today talking about tweaking software for modern hardware 20 years later. And since Moore's law doesn't really work any more, we can expect little progress in terms of single core performance over the next 10 years. Besides, the big thing now is power efficiency to extend battery life and that requires efficient code as well. Optimization will always be important. Why use twice the core count when you can just write tighter code? That argument isn't going away.

Re:Yes, let's bring that back

[gr8_phk]

So you don't understand Moore's law. Thanks for making that obvious.

What Moore originally wrote down was the equations that defined how you scale transistor sizes, which was an enabler for the continued performance increases which are now associated with Moores law. While doubling core counts every 2 years will allow the continued performance increases, the original moores law ran out of steam some time ago.

Re:Yes, let's bring that back

[ArcadeMan]

There's only one law. Brannigan's Law.

Re:Yes, let's bring that back

[Anonymous Coward]

That's not exactly true. I went from getting 85fps encoding with handbrake under a 3.2.0-486 kernel to a whopping 175fps encoding when I went back to the 2.6.32-686 kernel. (Debian's Wheezy only offers 3.2.0 as either 486 or 686-PAE, and the PAE kernel just plain wouldn't run on my machine for some reason)

Re:Yes, let's bring that back

[fisted]

Moore's law is about transistor density (or equivalently, the number of transistors) doubling every other year. Number of transistors and transistor density are equivalent in this case because Moore's law isn't constrained to a single die. Considering the whole machine instead, adding another core roughly doubles the amount of transistors, and hence the density from (say) 1 gazillion transistors per 19" casing to 2 gazillion transistors per 19" casing.

Re:Yes, let's bring that back

[Moondevil]

There are also compilers that spit out native binaries for Java and .NET.

Re:Yes, let's bring that back

[HappyPsycho]

I think you are arguing something different to GP, you are arguing that the transistor count specified by Moore's Law is increasing, GP is arguing that it doesn't matter as much of today's software is single threaded and hence single threaded performance dominates most PC performance today.

If it didn't then OpenCL / CUDA (heck everyone would be going nuts over a 6 or 8-core and dumping their quad cores, most seem happy with dual-cores) should have blown everyone's mind (It does for some people but most of those are in specialized fields), most software isn't even multi-threaded so sure there may be more transistors but for 90%+ of the tasks a normal computer performs only a fraction are in use at a particular time.

Re:Yes, let's bring that back

[fisted]

I'm not quite sure what you're trying to say - however initially all i was arguing against was the claim that moore's law had come to an end. or rather, overtook or got significantly overtaken by reality.

Re:Yes, let's bring that back

[Anonymous Coward]

"Windows/OS X geek"...

Re: Yes, let's bring that back

[Sigg3.net]

" I just smell the acrid stench of cynicism wafting from the general direction of the OP"

Your insight goes beyond you; I often find myself being negative and cynical when I have a gassy tummy.

Re:Yes, let's bring that back

[Anonymous Coward]

for a hundred you can get a choice of similar systems boxed with remotes and psu's..

[citation needed]

Re:Yes, let's bring that back

[Hognoxious]

Did you even read the post you're replying to? You know, the bit about it being in the future, when phones will be as powerful as current desktops, which will make überoptimisation less critical?

As to ADRA's question, I'd say probably. There's already programmable watches that don't have stacks of computrons. In ten years you might get similar power to today's smartphones in, say, an earring or a tie-clip.

Re:Yes, let's bring that back

[james_pb]

The age of low level fast optimization is all but dead.

In the embedded systems world, you can turn more optimization into cheaper hardware into (lower prices/higher profits). It's not dead.

Replaces hardware lag with animation lag

[White Flame]

Hopefully all the swishy fadey stuff can all be disabled, so that the speed improvement actually manifests usably.

Re:Replaces hardware lag with animation lag

[ebenupton]

Yup. We know lots of people don't love the shiny (or love the speed more than the shiny), so we'll be providing the ability to turn off fades and scaled window browsing. Disabling fades has the nice side effect of removing 120Mpixels/s of blending, so you can have more windows on the screen before the back of the stack falls back to 30fps (for responsiveness the front of the stack will always run at 60fps regardless of scene complexity).

Re:Replaces hardware lag with animation lag

[Anonymous Coward]

Seems a shame that the hardware blender hardware can't also do inline blends without having to have a fullscreen layer of a single colour/transparency. I guess that layer could also incorporate a fancy shadow for the topmost window though to make it less of a waste. Alternatively, you could surround the active window with smaller "Mac OS X like" drop-shadow layers to reduce the hit on blending throughput.

However this is a definite usecase for going to Weston over the status quo. I have long wondered why normal graphics hardware doesn't provide more 2D layer/blending hardware, as that is what most computers spend their time doing.

wayland

[Anonymous Coward]

Great, more wayland propaganda. As if exploiting certain hardware features has anything to do with Wayland vs X11. Wayland: Breaking decades of backwards compatibility for no good reason.

Re:wayland

[Jah-Wren Ryel]

Great, more wayland propaganda. As if exploiting certain hardware features has anything to do with Wayland vs X11. Wayland: Breaking decades of backwards compatibility for no good reason.

Exactly. This article boils down to "wayland performance on pi went from suckass to very nice" which is mildly interesting but the implication that wayland rulez and X snoozes because of that is specious. There is no reason X couldn't see the same performance improvement if it too switched drivers.

Re:wayland

[Anonymous Coward]

X11: Being needlessly complex with today's use cases for no reason.

If X11 is so good, why isn't Android using it?

Re:wayland

[zakkudo]

I've seen lots of special ports of packages made to take advantage of the RaspberryPi's gpu. X11 is the most conspicuous one left out. I want to hear you give me this reason.

Re:wayland

[Anonymous Coward]

As if exploiting certain hardware features has anything to do with Wayland vs X11.

Of course it does, fool. Hardware acceleration is huge win.

Re:wayland

[0123456]

Hardware acceleration is huge win.

Yeah, because, as everyone knows, X11 has no hardware acceleration, which is why it sucks and stuff.

Re:wayland

[TheRealMindChild]

You show me documentation on how to write an X video driver, and I'll friggin do it. There is no consistency between the drivers at all to even snag one as a "template". There is no article/paper that I could find that says "Here is how you develop an X video driver".

Re:wayland

[smash]

Why doesn't any device that actually requires decent GPU throughput use it, including the Mac, the PS2/3/4, etc?

Why did those developers see fit to NOT use the freely available BSD-style code out there and spend their time writing their own rendering pipelines?

For fun?

Re:wayland

[Anonymous Coward]

"Cascade of Attention-Deficit Teenagers"

Re:wayland

[Anonymous Coward]

As if Android is a good example for great engineering choices.

Re:wayland

[ebenupton]

As the video and Daniel's post explain, we don't lose backwards compatibility because we can host legacy X applications in a Wayland window using XWayland. We get all of the benefits of doing top-level composition in hardware, none of the pain of writing (and maintaining) a hardware-accelerated X driver. Can you explain why anyone starting from a clean slate today would chose to accelerate X itself instead?

Re:wayland

[ebenupton]

Amen. X seems to have the highest complexity to documentation ratio of any major software subsystem I've ever come across.

Re:wayland

[EETech1]

Mod Up(ton)

Re:wayland

[epyT-R]

NIH syndrome, most likely. The opengl performance on X11 is quite good. The nvidia driver shows that it's possible.

Re:wayland

[epyT-R]

At some point, someone will have to maintain the hardware specific driver. Wayland may or may not be a cleaner api, but the work still has to be done for each device.

Re:wayland

[smash]

If you've been using Linux since 1.0 (I have since 1.2) and have never seen any X11 failings, you're either talking out of your arse or are completely blinded by unrelenting fanboy-ism.

I've seen plenty of X11 failings over the years, ranging from inability to change screen resolution on the fly for about the first decade, poor security, crashes in the video driver taking down the OS, various hacks to get things like multi-monitor or 3d support to work, etc.

Yes, some of those things have been "fixed" via various bodges, in much the same way that the average wannabe Nissan Silvia drifter will "fix" crash damage with a drill and some cable-ties.

High latency, low bandwidth, high security risk stuff like network transparency does not belong in the same process as the rendering engine. It certainly doesn't want to be running as root. Especially when the majority of people simply do not use it, and it can easily be retained via a daemon like every other platform uses.

Re:wayland

[smash]

Just because something is "possible" it doesn't mean it is a good idea. The fact that as per TFA wayland got 20% better power consumption BEFORE they took out a lot of un-necessary data copying should be reason enough for Linux people to sit up and take notice.

Mobile devices are future and a 20% plus reduction in power consumption whilst improving performance is nothing to sneeze at.

Re:wayland

[ADRA]

I'll be generous and say there are probably all of 50 generic display device drivers written specifically for X11, probably the same for Apple, and maybe double-triple that for Windows drivers. It isn't exactly a large playing field for development efforts to just pick up from nothing, which is also why 99% of drivers are written by the manufacturer of the device.

You could always look into http://www.x.org/wiki/Development for guidance, but in the end code is king. X Development is not simple, but neither is graphics development in general, which is why every card on the planet has a large set of incompatible surfaces that all get crushed together behind huge drivers in order to spit out a more or less coherent systems interface that a display server will interoperate with.

Re:wayland

[cduffy]

I'm... sorry?

You think SysV init scripts are in any way, shape or form moderately acceptable?!

I have a very simple refutation to that -- the collection of run scripts behind this link.

Go ahead -- have a look. Keep in mind that systems using those mostly one-line scripts all provide not just startup/shutdown/status, but also the ability to auto-restart on failure and lack the propensity for race conditions that pidfile-based locking almost universally used by SysV scripts is so very, very prone to.

Holding up SysV init scripts as a thing that doesn't have to be changed... it beggars belief.

Re:wayland

[Microlith]

Which is why such drivers should go upstream. That's why the kernel developers want you to push your driver into the kernel - it gets maintained.

Re:wayland

[Anonymous Coward]

Network transparency?

Re:wayland

[smash]

Yes, and the cleaner API is everything. If backwards compat can be maintained (it is) and the codebase can be a lot cleaner (it is) and perform better (it does) then why are people so anti-X replacement?

Open source is supposed to be a meritocracy, yet with all the weston hate around here you'd certainly not get that impression every time a weston thread pops up.

Re:wayland

[Anonymous Coward]

so here's the story:

<rant>
some X11 devs decided that tcl/tk and motif was passe and should die, and that X11 should look like their new iPhones. So they tried to make X11 look like an iPhone and it didn't work very well, all the apps ended up having to be rendered and shaded pixel by pixel and it was a huge bottleneck.

so they abandoned X11 in its now broken state and started anew with Wayland, where they were freer to try and make it look like an iPhone.

What pisses me off is that they mock X11 for being broken and a huge bottleneck, but THEY are the exact same bloody few people who recently broke it and made it that way.

And thank you very much most of my professional work is done in motif, be it coding in nedit or using the HPC number crunching apps. I could give a shit if my cluster management software has the latest compiz-wobble transparency bling added to it, I'm trying to get work done here not look at cute puppies in 3D with surround sound and a semi-transparent background gently blending into my web browser window.
</rant>

Re:wayland

[smash]

Having used PCAnywhere, VNC, X11, ICA, RDP, and PCoIP - X11 rates last in terms of performance. It rates last in terms of features. And before someone says "oh but rootless mode!", RDP and ICA have been able to do that since 1996 or earlier. If X11 is gradually phased out something better will replace it. Like perhaps something developed in the last 15 years.

Re:wayland

[Anonymous Coward]

Wayland with XWayland has backwards compatibility but no forwards compatibility for new applications. Also, by design, XWayland is a second class backwards compatibility kludge with the declared goal that it will be depreciated. Embedded platforms will probably never include it by default and this will break the Linux ecosystem completely (Android already did a lot of damage). In the unlikely case that XWayland would always be included and maintained for compatibility, overall complexity would be higher because now all X11 code would be maintained in addition to everything Wayland adds. So this is clearly not what is planned. And if the only point would be to have a nice API to write drivers against, this could clearly be accomplished by refactoring X11 internally.

Re:wayland

[ebh]

OK, here's the slightest failing I've felt numerous times: Old laptop whose graphics under WinXP were just fine, if a little slow. Install Linux with your favorite desktop (LXDE works best for me on that machine). Machine is under moderate load. Click and drag a window to move it. When you release the mouse button, the window is still following the cursor, because the mouse-button-up event was not handled properly.

JWZ complained abut this sort of thing (events not being presented to the handler in proper order) when he was working on the Unix/X version of Mozilla. ISTR him saying that it was a problem wih the protocol itself, which would explain why the problem persists almost 20 years alter.

Re:wayland

[electrofelix]

X11 on linux is network capable but really can no longer be classified as network transparent. None of the main rendering engines for X11 on linux are network transparent.

The talk on the state of X11 and Wayland/Weston given by one of the lead developers is a bit of an eye-opener about just how munged up X11 is at this stage.

Re:wayland

[Anonymous Coward]

Amen.

Why don't the same people that screwed up X11 and now working on Wayland just fix X11? I sense a fail on the scale of epic Pulseaudio fail just waiting to happen.

Re:wayland

[Anonymous Coward]

So you've never briefly pressed a key when using VNC and seen key repeat go out of control? I've repeatedly (groan) seen the key press got to the other end, but the packet containing the key release was delayed. Never happens with X.

Re:wayland

[peragrin]

My complaint was simpler. Hot swap monitors in 2003.

in 2003 I could unplug a monitor from my powerbook G4, and plug in a different monitor with a different resolution without causing anything other than window resizing things(and even that was done mostly automatically)

I tried that with linux in 2010 and not only did it crash out X11 but the automatic tool that was supposed to do it wouldn't restart. I didn't want to manually rerwrite x.conf every time I wanted to plugin in a different monitor(something I was doing several times a day).

To this day I miss aspects of transparent network windows. remote desktop/VNC just are not the same. However they are fast/ stable compared to X over anything but a local 100mbit lan.

I truly wish someone would rewrite X from the ground up with some new ideas on how to do the network transparency.

Re:wayland

[kenaaker]

I work every day on 2 or 3 X based systems with all the individual windows coming to a triple-headed X desktop driven by an Nvidia GTX 660.

Even the netbook that's driving my FDM printer runs X clients remotely, very nicely and Cura displays its 3D renders from the netbook to the X desktop system just fine using OpenGL remote. By the way, the netbook has NO OpenGL hardware.

On that same X desktop machine every Linux Steam game that I've tried works without any problem.

You want to re-invent the wheel, go right ahead, don't let me stop you. Just quit trying to displace something that works amazingly well for everything I want with some spatch-cocked thing that you cooked up to scratch your own itches.

And stop making asinine claims that nobody wants to do what I do every day.

Re:wayland

[smash]

You have no idea what I've done mate. I've also used X11 over 10 megabit ethernet and it's still crap.

Re:wayland

[Anonymous Coward]

*Networked* mobile devices are the future and network transparency makes a lot of sense here (if it would be embraced instead of shoved aside into a backwards compatibility kludge). And every improvement in buffer management which can be achieved in wayland can (and will) be achieved by extending the X protocol. Doing it in Wayland is only because of NIH syndrome and has not a single good technical reason.

Re:wayland

[smash]

Can't say I have. Used to run TightVNC as a remote support tool for about 5 years (supporting 500+ windows boxes at the time) and never seen this issue?

Re:wayland

[smash]

Network transparency, if relevant will be provded by an additional daemon, where it does not have to run as root and does not become part of the local rendering pipeline. Networks are slow, any minimal theoretical performance impact moving the networking outside will incur will not be noticed as it will be imperceptible vs. the latency on even 10 gigabit ethernet.

Re:wayland

[smash]

If you think X11 was "recently broken", you're deluded. It's been a steaming turd for a very long time. And whilst you can't polish a turd, you can dump it in sparkly glitter and it will look a little better than before. But it's still a turd.

Re:wayland

[EzInKy]

True, with "X" multi device computing was an afterthought...Wayland takes it center stage. Or am I wrong?

Re:wayland

[complete loony]

http://www.collabora.com/services/case-studies/raspberrypi/

Wayland's protocol and architecture allows it to serve X11 clients, through an emulated server. Improvements made to Weston as part of this engagement with the Raspberry Pi Foundation by Collabora enabled X11 applications to run seamlessly, running faster than under the legacy X.Org server.

Re:wayland

[EzInKy]

Because most users own multiple device, each of which may require their own accelerator? If Wayland is to replace X, it will need to replace all the drivers needed run on all the devices that X runs on.

Re:wayland

[gagol]

Most people are afraid of change...

Re:wayland

[smash]

LOL. I have plenty of complaints with Microsoft. But if something is crap, I'll call it crap - i don't particularly care who is responsible for it. X11 is crap. VNC is crap too, but performs somewhat better than X.

Re:wayland

[lister king of smeg]

If you think X11 was "recently broken", you're deluded. It's been a steaming turd for a very long time. And whilst you can't polish a turd, you can dump it in sparkly glitter and it will look a little better than before. But it's still a turd.

Just to be pedantic; yes you cans polish a turd mythbusters busted that one.

https://www.youtube.com/watch?v=yiJ9fy1qSFI

and by the way x11 is no turd.

Re:wayland

[Anonymous Coward]

What do you mean by "becomes part of the local rendering pipeline"?

Networks are slow... That depends. People implement direct copy from the NIC to the GPU in high performance computing, because an additional copy through main memory adds too much latency.

But this is not the main point. The problem with moving the network transparency into a backwards compatibility kludge is less optimal integration by making remote application second class citizens and loss of forwards compatibility.

Re:wayland

[ebenupton]

It's not a lie at all - way to play the man rather than the ball.

Most of our users care more about local performance than they do about network transparency, so this is where we're investing our (limited) resources. People who care about network transparency can continue to use X either to or from the Pi; I don't think anyone is seriously suggesting that X is going to be replaced by Wayland in all use cases, merely that Wayland meets a need for high-performance, high-quality, low-power (and lower software complexity) local composition.

Re:wayland

[Blaskowicz]

Rootless RDP : need to spend like $1000+ on Windows + pack of CAL licenses + terminal server licenses.

VNC : need to add a vncserver, fuck with bit depth and resolution settings, then you hijack the desktop that is running on the other machine, at the wrong resolution for your local computer. Right, this is totally not what I want to do. Maybe you can sysadmin your way around it (and even fix the lag)

X11 : ssh -X or putty + xming. No configuration, no installing something, no buying Windows Server, it just works at all. You won't be watching videos on it but who cares? It's good enough for even Firefox on a lan as long as you don't go to some page that make it tank. Or load a file manager and music player, drag'n'drop files to the playlist and have the sound come out of the big speakers attached to the remote computer.

Re:wayland

[smash]

100 gigabit ethernet is still slow compared to your GPU. You don't need to make it a "second class citizen" by removing the code and relocating to a user space daemon. The performance figures I mention are to demonstrate that network IO is drastically slower than the GPU, so much so that any miniscule performance hit from relocating said code to a daemon will be imperceptible - with maintainability, security and other benefits.

Re:wayland

[Guspaz]

When I used Linux as my desktop (well, laptop) OS on a day to day basis (maybe four or five years ago), the only option to add a second monitor (plug my computer into my TV) using X itself was to restart X. Since that kills all your apps, that is effectively a requirement to reboot your machine just to connect to a TV.

The only way I ever got around those issues were via nVidia's proprietary drivers and control panel. They, at least, could add additional displays without killing everything. Most of the time. Sometimes it just didn't work, or killed X.

I've not used Linux as anything but a headless server OS for several years now, so I can't claim that this situation still persists, but X forced me to deal with headaches that Windows and Mac OS had solved a decade earlier.

Then again, my experience trying to use the latest version of OS X on a macbook pro on a pro-grade video mixer last summer very annoyingly illustrated that nobody has got multi-display anywhere close to perfect yet. I think Windows gets closest (comparing my Macbook to my Windows desktop), but it has its own set of issues. Annoyingly, a lot of these issues are UI-level or configuration-level deficiencies, not technology-level issues like X has (or had, at least).

Re:wayland

[smash]

I'm not quite sure where you get the idea that I'm a windows fan. I am platform agnostic and use whatever fits the purpose.

I'm actually a FreeBSD + Mac user at home a mixed environment (FreeBSD/Linux/vSphere/Windows) at work to pay the bills.

None of my current platforms of choice can run Wayland, and currently on FreeBSD, X11 is the only option. It's still crap.

Nice try though.

Re:wayland

[metacosm]

I hope people will stop doing that (VNC remotes), as a contractor I have seen VNC installed "in the wild" 4 times. In all 4 cases they used a common password corporation wide. This password was stored weakly encrypted on the individual machines in registry, trivial to decrypt. At that point, it is just a matter of searching the network for the most important sounding user "Bob (CEO) Laptop" -- connect to it, watch 'em work for a minute, then open notepad and write "Can you call me at extension X, Thanks!"

Re:wayland

[smash]

And as to the why behind my choice of home platform? OS X server is crap. Windows (i have an install for gaming) is pretty crap from a UI perspective. Power consumption is bad (I'm on a macbook pro) as well. I bought mac hardware because the trackpad works and being aluminium the chassis still looks and feels like new after 2 years rather than being discloured/worn shiny plastic. Guess what: PC laptop enclosures are crap.

You can piss and moan about my percieved bias against anything open source all you like, but it won't change the facts. There are many reasons why Linux on the desktop has not taken off, and X11 is one of them.

Having run multiple platforms other than Linux and used remote display capabilities on all of them - for all the puported X11 remote display "advantage" - in reality it just doesn't perform. EVERY other remote display technology I have encountered has been superior.

No audio, shitty bandwidth consumption, poor security unless paired with something like an SSH tunnel. Then there are the local deficiencies: a massive monolithic blob running as root in kernel space, the convoluted code-base and hacked rendering path (layering nightmare), the fact that it is attempting to do the job of the kernel by providing driver support, etc.

Now, if you want to stick your head in the sand and pretend X11 doesn't have serious problems, be my guest. I'm sure some die-hards will maintain the Xorg tree just for people like you. But don't try and convince me that X11 is perfect and doesnt have major problems because that is just demonstrably false.

Re:wayland

[smash]

Oh nice, call me a shill. It hurts man. I have multiple FreeBSD machines in production (primary NS, sendmail for a 2500 user company) have it running as a ZFS NAS, have run IPSEC endpoints with it. I've previously run Linux exclusively on the desktop for a number of years.

Anyone is free to go through my comment history, it goes back to 1997 or so and I've gone through various phases of fanboyism over the years (starting with Linux) and after exposure to a lot of different shit have become fairly agnostic.

I want to get shit done. If, for a given problem, Linux is best - I run Linux. If I feel FreeBSD is a better fit, I run FreeBSD. For my personal machine, I run OS X.

But I'm never going to convince you of that. You're just a zealot, and anyone who says anything bad about your platform of choice must be a shill. Me? I'll happily list faults with every operating system I use. OS X - the filesystem sucks. The lack of customization sucks. FreeBSD - the lack of virtualization sucks. The upgrade process sucks. Windows - the UI sucks. The third party driver support sucks. The lack of relevant logging sucks. The command line sucks. Linux - the GNU userland sucks, it is different to proper unix for no good reason. info sucks. give me proper man pages. BTRFS sucks. systemtap sucks.

There is stuff that sucks about every platform. You take the trade-offs and you make your choice.

What you DONT do is stick your head in the ground and accept crap as perfectly OK, because if you do, you'll just end up still having crap 10, 20 or 30 years later.

Re:wayland

[smash]

Yup, this is one of the major reasons we canned it. We originally used it because there was no in built remote assistance tool in Windows 2000 or previous.

One of the first things I did when i returned to this company was start phasing out VNC in favour of remote assistance/remote desktop and as soon as I can get it up and running properly, SCCM's remote control.

I've actually put group policy in place to disable VNC on all domain machines (several years ago), but we're pretty much free of it now due to PC attrition/replacement.

Re:wayland

[smash]

That's nice. I have remote machines on the end of shitty 512kbit satellite links in africa. We have enterprise licensing for Windows so the costs aren't that bad. We need some level of windows infrastructure in place in any case to handle Exchange (PHBs want it) and the various mining industry tools out company uses to get minerals out of the ground.

The point is thus: irrespective of what platform you run, X11 performance when compared even to offerings by Microsoft (RDP) is just blasted into the weeds.

That, my friends should be a fucking embarassment. X11 on 10 megabit ethernet performs worse than RDP over 256kbit frame relay. It's a fucking disaster.

IF i want to replace all my end user's desktops with dumb terminals, X11 simply isn't going to cut it.

Now, I'm not saying run windows everywhere. I'm simply saying that by any metric you care to use, X11's remote performance is simply horrible. IF wayland starts a push to phase it out in favour of something that is actually usable over something slower than an ethernet LAN, this is (long term) a GOOD thing.

I'll bet you X11 stalwarts are complaining about the need to convert to IPv6 as well? If not, why not?

Re:wayland

[Jah-Wren Ryel]

> It's not a lie at all - way to play the man rather than the ball.

Are you kidding me? Did you even read what I wrote? Go back and pay closer attention - I accused you of being disingenuous and emplying misdirection, not of lying. And your mischaracterization right there, that only reduces your credibility even more.

> Most of our users care more about local performance than they do about network transparency, so this is where we're investing our (limited) resources.

You know what? That is FINE. Just stop trying to put words in the mouths of the people who disagree with your prioritization. Go ahead and say that you don't care about network transparency, but stop it with the misleading statements about "backwards compatibility" and now this new one about wayland replacing X in all use cases - because those are simply strawmen, useful for misdirection, not useful for educating anyone else reading along.

Re:wayland

[Jah-Wren Ryel]

Yes, it is well known that X11 over long-latency links sucks. So running remote X over a satellite link is basically the worst possible use case. But remote X on 10mbps local ethernet is significantly faster than RDP over a 256kbps satellite link.

Re:wayland

[Anonymous Coward]

I am but a humble anon who has been speaking and reading English as my first language for the last fifty years, so what do I know, but "being disingenuous and emplying misdirection" sounds pretty darn similar to "lying" to me. Would you say that to this guy's face? Or are you just waving your dick around so that you can look like a big boy by publicly chastising an internet-famous person?

Re:wayland

[Anonymous Coward]

Which introduces the problem of maintenance for every OS. Or will the Linux kernel maintainers also maintain the drivers for the Wayland-related hardware in all the *BSDs, *Solaris, etc.? Instead of introducing a driver change in one codebase (of X or Wayland), it may be necessary to change it in dozens of other OS kernels.

Re:wayland

[rastos1]

The SysV init scripts have one huge advantage though: I can read/debug/understand them and all I need to know for that is a bit of sh(1) and coreutils. I have no use for shaving off 10s from the boot process and I don't start/stop services so fast that I could run into a race condition. I like being able to find out whether the service is today called bind9 vs named or httpd vs apache2 by simple filename completion.

Although your /. id is smaller by 3 orders of magnitude, I'll stick with scripts if you don't mind.

Re:wayland

[smash]

For general GUI configuration stuff, RDP over 256kbit is not much different from local. Last time I ran X11 over 10 megabit ethernet (from my Linux box to a Solaris box), I saw flickering (as window content was updated), etc.

Fact is, I regularly do RDP over 512kbit satellite and whilst unpleasant it is fairly usable. I used to do it via 56k modem.

RDP isn't even the best tech out there.

Re:wayland

[Anonymous Coward]

You know who the main devs are for Wayland? They are Xorg devs, who presumably having worked on the Xorg code for many years know it inside out, if they think Xorg can't be fixed and it needs replacing, then they are far more likely to be right than me or some random Slashdot commenter. No doubt you could use kludges to wring decent performance out of Xorg, but that doesn't mean Wayland is the wrong way to go.

Re:wayland

[gbjbaanb]

if you relocate the graphics rendering to the GPU and make it perform better, then using that system and sending the rendered data to clients means not only the performance improves, but the load on the server reduces.

RDP has been doing "network transparent" viewing for a long time, its more than sufficient for all, so if we can improve things using this - we should. No need to run X just because its X.

Re:wayland

[serviscope_minor]

which is mildly interesting but the implication that wayland rulez and X snoozes because of that is specious.

Welcome to the FUDStorm which is Wayland. The irritating thing is that it frequently comes from X11 developers who certainly do know better.

There is no reason X couldn't see the same performance improvement if it too switched drivers.

X11 has certainly supported 2D video scaling through the XV extension for a very long time. In fact I remember specifically upgrading a card to get better XV performance years ago. That was back when xmame (now vanished) needed all the CPU it could get, so a card which would upscale to full screen in RGB (cheap ones only did YUV) for free was a real boon.

And now onto some fun facts:

The "screengrab of a video gives a blue rectangle" problem which has been around since long before "protected video paths" and exists on Linux (or did once) was due to the video scaler. You paint a mask on the screen (pure blue was usually the colour of choice), then via a separate mechanism, pass a bitmap to the video card. It then scales it and paints it over the blue bits.

The pixels never enter the framebuffer, so screengrabs don't work.

The masking is necessary so that video controls etc can be displayed on top of the video.

Rather entertainingly, some old cards weren't able to have the mask done in anything except an unclipped rectangle.

So, if you had an xterm partially in front of the video and did an ls --color, you'd get a tiny bit of video bleeding through the directories.

XV doesn't seem to use the blue masking method any more, and the pixels seem to wind up in the framebuffer now, but nevertheless, X11 has supported CPU free scaling for a while. I was also astounded that it worked seamlessly across multiple physical cards with different drivers in 2003 just fine.

Re:wayland

[serviscope_minor]

Network transparency, if relevant will be provded by an additional daemon, where it does not have to run as root and does not become part of the local rendering pipeline.

You mean just like X11?

Locally X11 uses unix domain sockets (very fast) for the command stream. The even higher bandwidth stuff goes via shared memory (MIT-SHM) if you're rockin' it old school and via DRI if you're doing 3D.

There's a reason why the best framerates for games are in fact present on Linux/Xorg, not OSX or Windows.

Re:wayland

[serviscope_minor]

Sure, there are some pretty dodgy init scripts out there, but my god the replacements aren't half terrible.

Don't get me wrong, they are all shiny and great and work excellently when they work, which, in fairness, they mostly do.

But when they do go wrong (and they do eventually), they are so horribly opaque that fixing them is far, far harder.

Honestly, I do not really see the advantage of things like upstart and systemd. They haven't significantly altered the boot time of systems (well written init based system booted fast anyway) and they've just added layers of complexity.

Also, regarding pidfile based locking: open(..., O_EXCL | O_CREAT) is atomic, which is what you need for a lock.

Re:wayland

[DrXym]

Breaking decades of backwards compatibility for no good reason.

Run rootless X over the top of Wayland. Oh the humanity.

Re:wayland

[serviscope_minor]

I tried that with linux in 2010 and not only did it crash out X11

Yeah looks like you had bad drivers and/or hardware. Fortunately that only happens on Linux, so you should be safe on other operating systems /sarcasm.

On the several machines I have, I can freely switch monitors just fine (intel and nvidia. No AMD laptops here for now) with no crashes.

But the plural of anecdote is not data.

I truly wish someone would rewrite X from the ground up with some new ideas on how to do the network transparency.

A ground up rewrite is not needed. NX actuallyu keeps much of the X protocol, but has a number of tricks to reduce latency and round trips, much of it by making sure information is pushed to the clients asynchronusly, so they can then query it as if it were local.

It's not a ground up rewrite, but it basically fixes the problems.

Re:wayland

[serviscope_minor]

why are people so anti-X replacement?

Because X has lots of features we really like and even ones which aren't truly Wayland features are making it into weston (client side decoration, for example), even though they're a terrible idea for a variety of reasons.

Open source is supposed to be a meritocracy, yet with all the weston hate around here you'd certainly not get that impression every time a weston thread pops up.

Or, Wayland is worse and all the people are shouting about it because we don't want something worse forced upon us. So, the meritocracy is working just fine.

Re:wayland

[DrXym]

X11 is so shit since these days because it's basically passing around pixmaps or long sets of xrender instructions. And it's doing that because apps are pushing rich modern UIs that X11 is incapable of supporting through primitives.

As such any advantage it may have had in the past is moot at this point. It's just a bottleneck for local desktop performance *and* for network performance. While it's likely that X11 will live on in a Wayland powered desktop for some time to come, it's likely over time to become less relevant

Most apps run over APIs like GTK, QT, Cairo, Mesa anyway so they don't particularly care what backend they're hooked up to. I expect the libraries themselves could intelligently work out which backend to use for the context and apps will need minimal work to make the transition. The gnarliest are likely to be things like web browsers which host plugins and where the NPAPI on Unix has always been X so they have to shim the worlds in some way.

Re:wayland

[Anonymous Coward]

This isn't about video scaling. It is about using 2D blending hardware on the Broadcom SoC in the Raspberry Pi to accelerate a 2D desktop environment sensibly and smoothly, and retain full ability to use all of the GPU for OpenGL use, and all of the CPU for processing.

Such fixed function hardware is common on ARM SoCs, so it seems a shame to not use it when you have an explicit use-case - a desktop environment.

Re:wayland

[olau]

Did you actually read the article? The article in which an X developer points out that this is really hard to with X?

Re:wayland

[Anonymous Coward]

Why?

Because people have grow suspicious when someone shows up and wants to take away what they are using and promises a land flowing of milk and honey.. sometime in the future. Smells like communism. "You'll have to starve for now, but in the future..."

Analogous, you're never going to win any popularity contests when you actively remove features a lot of people with the current user base need and relies on, and try to placate them with some hand-waving about that "nobody needs that, and anyhow someone will fix the problem.. in the future". Any attempts will only label yourself as a syncopate or a flat out idiot.

Of course people get angry. If you think people are going to buy that "someone will fix it" line, you clearly display your contempt for their needs and their intelligence. You think you'll gain customers if someone says "I'm not buying that, I don't trust what you say", and you counter with "U R STOOPID, IZ KNOWZ BETTAH and spitting in their face?

Re:wayland

[cduffy]

The SysV init scripts have one huge advantage though: I can read/debug/understand them and all I need to know for that is a bit of sh(1) and coreutils.

This sentence being included in your reply implies -- strongly -- that you didn't follow the link and read the scripts behind it.

Hint: run scripts are far, far simpler and easier-to-read than SysV init scripts. (They also exist on-disk -- in a directory matching /service/${servicename} if you're following DJB's conventions -- so filename matching is absolutely supported).

If you gained something by your race conditions, I'd see it. Using big, overcomplex SysV init scripts adds nothing but bugs.

Re:wayland

[Jah-Wren Ryel]

> Would you say that to this guy's face?

Absolutely. When people tell "true lies" to avoid addressing my points in real life I call them out on it just as much as I do online.

Re:wayland

[Jah-Wren Ryel]

> Did you actually read the article? The article in which an X developer points out that this is really hard to with X?

Yeah, I read a complete gloss-over about that topic. I'll even give it to you that the X.org implementation could use an overhaul to better match modern hardware characteristics (presumably stuff that was designed with directx in mind, even if it is arm). But clearly there is nothing about the X protocol that is inherently hard to use with this hardware since the X/Wayland server runs on top of it.

I'd rather see the effort put into overhaul X.org, or even start from scratch with a new X11 implementation instead of starting from scratch with an entirely new system that lacks the features X11 has.

Re:wayland

[chuckinator]

You should look into using disper.

Re:wayland

[socrplayr813]

I really haven't followed any of the X11 vs Wayland crap, but are you against a new system in principle? Or is it just about Wayland?

In my case, I've rarely used anything but the most basic functionality of X, but I can see some nice benefits to a new system built specifically for desktop Linux that does exactly what I need and nothing more.

Honestly, I get excited about a little competition in the open source world. Far too many projects start and get to the point where they do what the developers wants, then stagnate. That's great and all (yes I know that the projects aren't necessarily for everyone else), but when it's something that hangs out in most or all Linux distros for years after while hardly changing with the times, maybe it's time for something different.

Re:wayland

[gr8_phk]

I'd rather see the effort put into overhaul X.org, or even start from scratch with a new X11 implementation instead of starting from scratch with an entirely new system that lacks the features X11 has.

Many of the X developers disagree with you.

Re:wayland

[kcbnac]

No, the worst possible use case would be remote X11 over a RFC 1149 compliant connection. (Although 2549 adds QoS, or RFC 6214, for those that use IPv6)

Re:wayland

[jedidiah]

So? GNOME3 is ample demonstration that the developers are not a magically valid appeal to authority.

Re:wayland

[jedidiah]

> If X11 is so good, why isn't Android using it?

Probably the same reason that Android doesn't have the sort of built-in printing support you would expect from a real OS. You have kids so eager to run from the past that they ignore anything they might learn from it.

I've had PCs less powerful than my phone that ran X quite happily. The same goes for CUPS too.

Re:wayland

[jedidiah]

> inability to change screen resolution on the fly for about the first decade,

No. That was always possible. The problem was changing the color depth. If you are going to spread FUD about this sort of thing you should at least get the basic details right.

> hacks to get things like multi-monitor or 3d support to wor

When I bought my first 3D card for Linux it "just worked". That was back when people might think that BeOS might be a contender.

As nifty as it was, the fact that Linux had working 3D acceleration was more useful.

Re:wayland

[unixisc]

Welcome to the FUDStorm which is Wayland. The irritating thing is that it frequently comes from X11 developers who certainly do know better.

But how do you know better than an X11 developer? Are you one yourself?

Re:wayland

[Anonymous Coward]

You apparently don't. In other words, you're part of the reason corporate new-speak and blatantly lying politicians exist. Have you ever thought about that?

A lie is a lie, no matter who tells it. And not daring to speak up against such because it's $FAMOUS_PERSON telling it makes you nothing but a coward, part of the problem and unfit for living in a democracy.

Re:wayland

[jedidiah]

> Network transparency, if relevant will be provded by an additional daemon,

Which will SUCK.

Just use a Mac for awhile and you will experience this kind of stupid design firsthand.

Ironically, Microsoft has spent the last 20 years trying to recreate the network transparency of X. It turns out network transparency is a pretty useful feature despite what the X haters want everyone to think.

Re:wayland

[JonJ]

Did it cease to be a turd when they polished it?

Re:wayland

[Jah-Wren Ryel]

I really haven't followed any of the X11 vs Wayland crap, but are you against a new system in principle? Or is it just about Wayland?

My problem is with a new system that does not even match the features of a 30+ year old system.

I can see some nice benefits to a new system built specifically for desktop Linux that does exactly what I need and nothing more.

The problem is that what you need isn't necessarily what other people need. There are people who don't use the mouse, would you be happy if Wayland skipped implementation of mouse support?

Re:wayland

[Jah-Wren Ryel]

Many of the X developers disagree with you.

And many of them agree with me. Whatsyourpoint?

Re:wayland

[maestroX]

I truly wish someone would rewrite X from the ground up with some new ideas on how to do the network transparency.

I truly wish the network transparency feature would be left out of this equation for once.
When I start this X thing, just use the balanced resolution/timings/colors like win2000/xp over 12 years ago did without any tweaking, proper gl support, even egl, plugin monitor support, dual anything, video, acceptable latency, suspending and ONE default theme.
Please, make the mistake to put it in the Linux kernel and get this thing FIXED.

Re:wayland

[Anonymous Coward]

His point is that the statement "There is no reason X couldn't see the same performance improvement if it too switched drivers." isn't necessarily true, though you posited that it definitely was.

Since you apparently knew that there was controversy, your failure to point it out is intellectually dishonest, and signals that you cannot objectively argue the point. Understand, or should I use smaller words?

Re:wayland

[Anonymous Coward]

remove your bitmap heavy theme

Re:wayland

[Anonymous Coward]

His point is that the statement "There is no reason X couldn't see the same performance improvement if it too switched drivers." isn't necessarily true, though you posited that it definitely was.

Funny how that's not the line he quoted and responded to. Good thing we've got you here to tell us what the other guy was thinking.

Re:wayland

[smash]

You mean a stock install of FVWM?

Re:wayland

[cduffy]

Honestly, I do not really see the advantage of things like upstart and systemd. They haven't significantly altered the boot time of systems (well written init based system booted fast anyway) and they've just added layers of complexity.

Please respond to the post I wrote, not the one you think I wrote. I didn't espouse upstart or systemd, both of which are far, far more complex than runit or daemontools (the DJB tool on which it's based).

Also, regarding pidfile based locking: open(..., O_EXCL | O_CREAT) is atomic, which is what you need for a lock.

Sure, it's atomic, but it's not fit-to-purpose.

Okay: You can create a file atomically. Now, what good does that do you if you want to clear the lock when the process holding the lock is dead?

If you're using POSIX advisory locking, you don't do anything: Lock is released the moment the process closes its file handles. Reboot? No locks are held on the way back up. kill -9? No locks held. You don't have to deal with stale locks, because they can't ever happen.

Re:wayland

[serviscope_minor]

But how do you know better than an X11 developer? Are you one yourself?

A nice example is Keity Packard talking about why Atoms are bad, because to initialise N atoms you need N server round trips and incur N latency costs.

If you flip to the XCB documentation, it confirms what I know to be true about the X protocol and actually gives example code for how to initialise N atoms incurring only 1 latency cost.

So yeah, being an X developer does in no way imply that you know what you are talking about/will tell the truth.

I'm not a developer but I do actually know my way around the protocol well enough to know when the developers are talking crap.

Re:wayland

[socrplayr813]

I truly don't understand why people are getting worked up over this.

My problem is with a new system that does not even match the features of a 30+ year old system.

The problem is that what you need isn't necessarily what other people need. There are people who don't use the mouse, would you be happy if Wayland skipped implementation of mouse support?

Of course I wouldn't want them to skip mouse support; you're being deliberately obtuse. Did you know that what you need isn't necessarily what other people need? A plain-old desktop user doesn't need all the features of X. If Wayland can be better in any significant way while maintaining the features that I need, why shouldn't I be happy about it? (Side note: I don't think this is what you're saying, but just in case - supporting 30 year old features just for the sake of feature parity would be assinine and a complete waste of effort. If they're features that really are used, then OK.)

This is the beauty of open source. You can continue to use X all you want while those of us who don't need the features can move on to something else. If X is truly better for a given use case, then people will use it. Or not. There is no harm in having both. Drop the torch and pitchfork and let people vote with their feet. If us lesser beings are all using Wayland or whatever, you can focus development on X on the 30 year old features that make you so special.

You might make an argument that development time should be focused on X, but that's always been a flawed view. The people that go start a new project do so because they don't want to continue with the old project. Again, the power of choice and beauty of open source.

Re:wayland

[Jah-Wren Ryel]

> you're being deliberately obtuse. A plain-old desktop user doesn't need all the features of X

Yeah, like what exactly?

Re:wayland

[cupantae]

You are the one who is complaining about lost features, but hasn't given an example. The main example being repeated is network transparency, but I honestly can't work out why I should care. What other differences convince you of X's superiority?

Here's a challenge: tell me what is missing, and why it's bad. I am a fair man, and if your reasons are good you'll win me over. But this:

The problem is that what you need isn't necessarily what other people need. There are people who don't use the mouse, would you be happy if Wayland skipped implementation of mouse support?

is total nonsense. If you cover all the usage cases of the old system, why are you replacing it? Some features are more important than others, and it is perfectly sensible to talk about "the average user", since the features needed by different people overlap, with some features needed by very few people (e.g. network transparency), and others by almost all (e.g. mouse support).

Re:wayland

[socrplayr813]

Cupantae covered my thoughts pretty well on that, but:

Even if the feature set of Wayland were exactly the same as X, in what way could it possibly hurt you for there to be two projects that do the same thing. Maybe Wayland is better is some way. Why are you worked up over it? I truly cannot think of a good reason and I really am trying to see your side in this. I posted my original question because I actually wanted to understand your point of view. The only things I can think of at this point is that you're either in it for the argument or you really just can't stand to see anything compete with your favorite project.

Whatever your issue is, you should either provide a reasoned argument or lighten up. If not, you're really living up the stereotype. How sad for all of us.

Re:wayland

[BitwizeGHC]

RDP is fast because it caches pixmaps. It can do this because it's basically a Windows GDI driver on the client side, and whenever there's a CreateBitmap or similar GDI call that creates a new HBITMAP, it can cache the bits on the server side, and further graphics calls on that HBITMAP do graphics operations on the server.

X11 works this way too; in fact, it was designed to work this way from the start. The problem is, in this era of kids throwing out and replacing critical bits of system functionality, the current toolkits don't actually use this functionality. They are written like a VGA game from the early 90s. They use their X11 window as a dumb frame buffer, allocate pixmaps within their own memory space, and then any time they want to draw they ship the entire goddamn pixmap over the network. Maybe as an optimization they'll ship a rectangular section of the pixmap.

NX gets around this problem by a form of aggressive optimization that actually looks at the pixmap data that's being drawn and caching the samey bits on the server side automatically. But if the l33t KDE and GNOME kiddies actually used the functionality that every Joe Blow Windows coder uses -- if they were aware of the issues -- a lot of X11's latency problems would go away.

And this is a big part of why I eschew "modern desktop environments".

Re:wayland

[BitwizeGHC]

The modern X11 protocol is more than capable of doing modern desktop graphics. The XRender extension can cache and composite pixmaps on the server side with full alpha blending. This can be used to, for instance, render antialiased fonts by caching one copy of each glyph to be drawn and then sending commands to composite the cached pixmap IDs.

The problem is nobody uses it, because attention-deficit linux kiddies think that they can do so much better than the infrastructure that's as old as their dads.

Re:wayland

[Jah-Wren Ryel]

Here's a challenge: tell me what is missing, and why it's bad. I am a fair man, and if your reasons are good you'll win me over. The main example being repeated is network transparency, but I honestly can't work out why I should care.

So, you clearly know the point under contention and you decide to prematurely declare that you are unconvertible. Nothing says "fair man" like prejudging.

Some features are more important than others, and it is perfectly sensible to talk about "the average user",

The problem here is that ignoring network transparency is not a case of trading off heavily used versus rarely features. It is a case of dropping shit the developers just don't care about.

You want to make an argument that something like drawing polygons or stippled lines aren't necessary anymore because 99.9% of X11 software does not do that, that's reasonable. But practically ALL X11 software runs on remote displays - dropping such pervasive functionality is in an entirely different league.

Re:wayland

[Jah-Wren Ryel]

> Even if the feature set of Wayland were exactly the same as X

Hhhm. Just what part of "My problem is with a new system that does not even match the features of a 30+ year old system." leads you to believe your statement applies to my position? Don't be that guy man, don't be that guy.

Re:wayland

[olau]

X is not really network transparent. It only works over a limited set of networks because the stack isn't geared for high-latency networks.

Meanwhile, Wayland developers have actually worked on network support. The main developer demoed it a couple of months ago.

I too would like to be able to SSH into a box, start a GUI program and get a window up. And I think that's going to happen at some point natively. In the meantime, X on Wayland allows you to continue to use X's network support. I'm looking forward to a day where I can SSH over ADSL and still get a usable window back. This sort of works with NX, but NX is just too buggy for me.

By the way, "network transparent" is in itself a stupid notion, and X is the perfect example of how stupid it is. Other similar flawed systems include CORBA and .NET Remoting. Going over a network is so fundamentally different that trying to gloss over it is guarenteed to fail.

Re:wayland

[olau]

I think you missed something here. It's not the bandwidth that's the problem, it's the latency. X is waiting for far too many requests. You could put it on a 10 Gbit high-latency network and it would still suck.

Re:wayland

[Jah-Wren Ryel]

> X on Wayland allows you to continue to use X's network support.

Watch your word choice. X on Wayland does nothing for native wayland apps. Which is about as meaningful as saying X allows you to continue to use X's network support.

Re:But Wayland sucks!?!?

[smash]

LOL. Pretty much the typical ignorant crap you get whenever anything wayland is posted. Top work summarizing the expected content of this thread :-)

Nice. Let me know when Wayland has networking

[LaughingRadish]

Wayland working on Pi is nice, but it's still a non-starter as long as it lacks networking.

Cool

[readingaccount]

But now that Ubuntu's moving towards Mir, Wayland is pointless. I mean, Wayland can't possibly be any good, otherwise why would Mir be in development? At least that's what people have told me.

Re:Cool

[gmueckl]

At this point it's not about being better or worse. It's about which of the two alternatives gets the important closed source drivers first (high-end AMD and nVidia GPUs as well as mobile GPUs). *No* open source driver for this hardware is remotely good enough to put it through its paces properly.

Re:Who gives a shit about the raspberry pi?

[Anonymous Coward]

Its because the hardware is economical and affordable in the global market on a massive scale.

THAT is just one of the many reasons why.

Fuck backwards compatability

[ArchieBunker]

99% of Linux users want desktop performance, not remote desktop performance. Put that legacy remote shit into a module if you want.

Re:Fuck backwards compatability

[PPH]

I smell a gamer.

Not that there's anything wrong with a gaming platform. But if platforms are being driven single-mindedly toward an architecture optimized for one class of app. I say: Go get your own platform (an Xbox). Desktop means we may need to minimize your precious full screen game. Maybe open another app on top of it. Maybe even a client from a remote system. So, if we have to strip everything out to satisfy your performance requirements, forget it.

If Wayland can manage multiple client windows and handle remote X clients, then maybe its worth a shot. We'll see.

Re:Fuck backwards compatability

[EzInKy]

No reason to run Linux if all you are worried about is desktop performance. Why in the world would you even consider using Linux if you can't think outside your own box?

Re:Fuck backwards compatability

[lister king of smeg]

"network transparancy is legacy" my ass. i remote into my debian box regularly

Re:Fuck backwards compatability

[Anonymous Coward]

Your sentences aren't connected by logical link. Your geek card is revoked remotely.

Re:Fuck backwards compatability

[Anonymous Coward]

Desktop means we may need to minimize your precious full screen game. Maybe open another app on top of it. Maybe even a client from a remote system. So, if we have to strip everything out to satisfy your performance requirements, forget it.

*Facepalm*

I suppose CAD isn't real work either, right? It's easy to make slow crap on top of fast code, it's kind of hard to write fast code on top of slow crap. This is why we have layered systems, the core services should be minimalist and fast and then you pile your slow turds like X11 remoting on top as modules/daemons.

Wayland can host an X11 server on top of it so you can remote X11 programs to the hosted server on top of Wayland which is something you could have learned by just looking at the damn Wikipedia page.

Re:Fuck backwards compatability

[Viol8]

Oooh , an X11 server ON TOP of wayland - so another layer of abstraction. That'll be efficient!

Not.

Here's a clue - X works. If someone has some pet project they think is superior let them prove it by showing how well it works in a 24/7/365 remote server scenario. Just shouting loudly about how shit the old system was just marks them out as a know nothing assÂ

Re:Fuck backwards compatability

[serviscope_minor]

Oooh , an X11 server ON TOP of wayland - so another layer of abstraction. That'll be efficient!

Thankfully you could even then run a Wayland compositor on top of that so every Wayland program just gets forwarded to X11. That way, you can still have X at the centre and benefit from the quality window managers and so on.

Of course, the copy/paste/DnD still won't interoperate properly but since it can work either war round you can have a shouting match about who is "legacy".

Here's a clue - X works.

Not just works, but works really well.

Re:Fuck backwards compatability

[jones_supa]

Not that there's anything wrong with a gaming platform. But if platforms are being driven single-mindedly toward an architecture optimized for one class of app. I say: Go get your own platform (an Xbox). Desktop means we may need to minimize your precious full screen game. Maybe open another app on top of it. Maybe even a client from a remote system. So, if we have to strip everything out to satisfy your performance requirements, forget it.

Why would this kind of scenario even be a problem? When we detect that there is a full screen OpenGL-accelerated application (game) running, we keep other apps buzzing in the background but not let them actually draw anything. We can give all the resources to the game, and popping another app on top of a full screen game is not very important. Then when we quit the game, normal desktop operation is restored.

Re:Fuck backwards compatability

[Anonymous Coward]

All the users that I know use Linux on servers remotely. Windows is still preferred for local desktop environment.

Re:Fuck backwards compatability

[Anonymous Coward]

Heh, servitroll_major. I knew I'd catch you defecating on this submission soon enough. What I love so much about X is that it's without a doubt the slowest, most inefficient, most problematic, most obsolete component of GNU/Linux, and rather than push onward to greener pastures, you'd twist, spin, lie, and do whatever it takes to safe face. Anything to avoid dealing with the fact that X is a dead end. That's the price you pay for getting too personally involved with your computing preferences. I've watched you troll, slander, and piss all over everyone who you perceive as standing in your way. Enjoy it while it lasts.

Re:Fuck backwards compatability

[PPH]

and popping another app on top of a full screen game is not very important.

Really? Can you speak for all users?

Re:Fuck backwards compatability

[PPH]

I suppose CAD isn't real work either, right?

OK, lets go with CAD. Back 20 years ago, we had CAD systems on X11 terminals with remote clients. By remote, I mean 30 miles between the data center and the engineering desktops. And we could have multiple users working on a single CAD model concurrently. And rendering was fast enough that we could walk through a model (not just wire frame, but rendered surfaces) at useful speeds.

Back then, it wasn't the gamers that screamed and pulled the whole thing down. It was Microsoft. They didn't have a piece of that pie, remote clients and a networked display model not being in their architecture. The pissing and moaning back then sounded a lot like what is going on now, for different reasons. Oh, yeah, and Microsoft's lap dog, Autodesk.

CAD system development has been driven by the same licensing fee per seat/display agenda that doesn't want that to propagate to other areas of computing.

Re:Fuck backwards compatability

[ArchieBunker]

CAD has progressed way beyond simple 2D drawings. Let's see you run Solidworks over an X11 session. I'm not a gamer but I do like a responsive GUI. Run VNC if you want remote desktops.

Re:Fuck backwards compatability

[PPH]

CATIA over X11. About 20 years ago. No problems.

We could 'walk through' a solid 3D model of an aircraft section without problems. Until some 'smart' people decided to make the reference implementation for NT. That slowed it down considerably. It may never have recovered.

Re:Nice. Let me know when Wayland has networking

[Alex Belits]

Especially considering that Pi would be a perfect example of a device that benefits from X11-style remote applications -- being based on a video decoder SoC, it has somewhat nice GPU but tiny CPU.

ok cool

[Osgeld]

but it doesnt change the fact the pi is a circa 2000 computer doing 2013 tasks ... poorly and hacky

Re:ok cool

[Microlith]

Yup, "poorly." Not simply at speeds to be expected of an ARM11 CPU, poorly. Not "well", like a $500 system. And "hacky," whatever the hell that means.

Re:ok cool

[VortexCortex]

Makes a great brain for my MAME cabinet... Apparently there was a LOT of "poorly" made arcade hardware up until just a few years ago.

Re:ok cool

[Anonymous Coward]

Your mame cabinet is a piece of shit in a wooden box if you have an HDMI tv or monitor mounted in it.

Be a man and own real arcade machines.

Re:ok cool

[Osgeld]

its not even a 500$ system in 2001 dumbshit

Re:ok cool

[gagol]

I don't own a warehouse and like to play more than 3 games?

Re:ok cool

[spire3661]

I was totally serving small personal websites on a $35 computer at 5 watts with 247 days of uptime, yep yessireebob.

Re:ok cool

[Osgeld]

you can do the same with an arduino running millawats, clap clap

Re:ok cool

[AC-x]

its not even a 500$ system in 2001 dumbshit

O RLY? HP Pavilion 6835; better CPU, less ram and considerably worse 3D acceleration compared to the Pi. Price was $699 without a monitor in 2001.

Re:ok cool

[AC-x]

Good luck doing that in 2001 (FYI Arduino was released in 2005).

Re:ok cool

[varmfskii]

Still basing the "quality" of the CPU on clock speed I see.

Re:ok cool

[AC-x]

Still basing the "quality" of the CPU on clock speed I see.

You should do some basic research before posting, it's well known that ARM11 doesn't compete well on a per-mhz basis with even older chips. Don't take my word for it tho:

Whetstone
Celeron 733mhz: 598 MWIPS, 185 MFLOP, 162 MFLOP, 116 MFLOP
ARM 11 700mhz: 270.5 MWIPS, 97.8 MFLOP, 100.8 MFLOP, 85.7 MFLOP

Dhrystone
Celeron 450mhz: 720 VAX MIPS
ARM 11 700mhz: 847 VAX MIPS

Re:ok cool

[varmfskii]

Funny the figures I have seen do not agree. The figures I have seen shown a pi @ 700MHz to be faster with integer than an 800MHz Celeron and slower when it comes to floating point. In addition the overall UnixBench figures seem to give the pi a slight edge as well

It is hard to compare apples to apples since ther are so many different Celeron models

Re:ok cool

[AC-x]

Source? The figures I've linked seem to be as like for like as possible and both Celerons listed are getting similar performance to PII and PIIIs of the same clockspeed (so are likely the Pentium based cores rather than new more modern cores).

Re:ok cool

[varmfskii]

600MHz Celeron UnixBench 4.1.0
Raspberry Pi 700MHz UnixBench 5.1.3
For Dhrystone 35% better performance per clock for the Pi and 23% better for the Celeron on the Whetstone. It also shows the Pi with an overall of 53% faster for the Pi (though much of this is due to I/O performance not directly related to the CPU).

Re:ok cool

[AC-x]

I don't trust those figures at all; First of all I can also pull a result of 86.0 and 62.3 for the RPi (which is it?). Secondly in the same source as the Celeron result there is also a 3 ghz Xeon that only gets a score of 160.8. I can find an 866mhz PIII that scores 198, which given that's almost 4x faster than your quoted result for an 800mhz Celeron (and faster even than a 3ghz Xeon) makes the figures you're using seem suspect.

Re:ok cool

[varmfskii]

I don't know about the celeron figures, but the Pi figures are consistient with what I have gotten myself.

Re:ok cool

[varmfskii]

Also, those figures may be for using soft floating point. And the over scores can be very dependant on the I/O systems.

Re:ok cool

[varmfskii]

Having trouble on this end. I would ignore the overall figures; they are too dependent on factors other than the CPU. The figures there are consistent with what I get for the Pi in my own tests. I don't know about the Celeron. The low and conflicting figures you are getting are probably results done using soft rather than hard floating point.

Re:ok cool

[Anonymous Coward]

I think he meant get a real cabinet, like a used The Punisher jamma compatible game cabinet and disconnect the jamma connector. Now read on the jamma pinout and make yourself a hdmi to jamma converter using and hdmi to rgb+lr converter. Make a dual usb controller using the original switches and and 2 controller from a specialized dealer. plug that in the pie and now you have a real cabinet. !!! You might want a Thunder & Lightning cabinet for vertical shooters if that is your thing.

Re:ok cool

[varmfskii]

Sorry to keep at it. I was finally able to look at those two RPi links (the sites were blocked) . The difference between the two of them has to do with the SD card in use (I/O speed) and they are both soft float benchmarks so the floating point performance is awful. My figures were for hard float (explaining why they got about 30 MWIPS and the one I referenced got 270 MWIPS).

Re:ok cool

[AC-x]

What about the wildly varying PC speeds? The Xeon score from your source seems unrealistically low, and the PIII score I found should be roughly the same as a Celeron 800 (at least for the CPU metrics)

Re:ok cool

[Osgeld]

um ok fine the avr, which was released well before the arduino, happy?

Re:ok cool

[Osgeld]

yes, and I didnt buy my computer at circuit city, with an athalon 800Mhz cpu and a gefore256 and 512 megs of ram I just broke 400 bucks

Re:ok cool

[varmfskii]

The Xeon number do indeed look wrong, but the PIII numbers (core wise) seem consistent with what I listed (using indices for simplicity 91.9*866/600=132.6 vice 156.2 [17% difference] and 54.5*866/600=78.7 vice 87.5 [11% difference]). Unfortunately I do not have a Celeron based machine set up right now to do a run myself.

Re:ok cool

[gagol]

Well, personally my laptop is fine for my annual "back to child" moment and like to have a house that dont look like a teenager's dream from wayback. If I had the time to put this kind of project together, I would start by automating my house... more practical and WAF friendly!

Re:ok cool

[AC-x]

And how are you going to network or add storage to that without Arduino's network/storage shields? Sure I suppose you could have hand-built an entire micro-controller based computer that ran on less power if you were really desperate (don't forget to include the power requirements of the extra components btw).

Re:ok cool

[AC-x]

How did you manage that? Reading reviews the price of the CPU and GeForce alone should have been at least $350, and that's without the cost of the motherboard or ram.

Re:Who gives a shit about the raspberry pi?

[Anonymous Coward]

you forgot, for 2004

Re:Nice. Let me know when Wayland has networking

[Anonymous Coward]

Good luck doing anything remotely bandwidth intensive or latency sensitive over the flaky as fuck USB-Ethernet on the B.

Re:Nice. Let me know when Wayland has networking

[Alex Belits]

Good luck doing anything remotely bandwidth intensive or latency sensitive over the flaky as fuck USB-Ethernet on the B.

Evidence?

Another one with no clue about wayland

[dbIII]

none of the pain of writing (and maintaining) a hardware-accelerated X driver

Of course not, since wayland uses those hardware drivers that were written for X. That's a good reuse of good code, avoiding re-invention of the wheel, and you really should have heard of that if you'd spent more then ten seconds learning about wayland instead of spouting "X sux" bullshit.
Once wayland hits new hardware that X doesn't support you get "the pain of writing (and maintaining) a hardware-accelerated driver". There's no point in pretending that drivers happen by magic.

Re:Another one with no clue about wayland

[ebenupton]

Fascinating points, except that once you've offloaded top-level composition to hardware you've claimed 90% of the benefit that you would have gained from full X hardware acceleration; even on the Pi it makes sense to use the software fallback path for all in-window rendering. I did bother to look into this a bit before opening my checkbook.

Re:Another one with no clue about wayland

[Guspaz]

So, let me get this straight... You're telling one of the founders of Raspberry Pi, who is a technical director at the company (Broadcom) writing those hardware video drivers you mention, and who was likely one of the people pushing the development work mentioned in TFA, that he needs to "spend more than ten seconds learning about wayland"?

Yes, I'm sure he's completely ignorant of such things.

Re:Another one with no clue about wayland

[Guspaz]

Are there many apps that have hardware accelerated in-window rendering on the Pi? Web browsers or photo slideshow apps or whatnot? I'm considering using a Pi for a project (display live room schedules for a convention on televisions), and while that sort of hardware acceleration isn't required, fading between screens/images/whatnot would certainly look nicer.

Re:Another one with no clue about wayland

[ebenupton]

Not at present, but we're expending quite a lot of effort on getting hardware-accelerated Webkit running at the moment; Wayland is a key enabler for this.

Fuck backwards rumours

[dbIII]

Sometime in the 1980s people thought of that problem in X and solved it - local stuff runs at high speed with sockets and the remote stuff only kicks in for remote windows.

Re:Fuck backwards rumours

[Alex Belits]

with sockets

...and shared memory.

Really, it's done already, and better than anything current "architects" can ever do.

Re:Fuck backwards rumours

[smash]

"solved" it. lol. If you have a very, very generous definition of "solved", sure.

Re:Fuck backwards rumours

[Anonymous Coward]

Sockets don't really solve this problem, shared memory transports do. As far as I know, those were only available in SGI's Xsgi (and libx11), and never made it into any other X11 servers. As a result, on comparable ~100MHz era hardware, Xsgi was significantly faster than Xfree86 or XSun for anything that required lots of round trips, e.g. starting Netscape 4 with its bazillion Motif widgets. Nowawadays almost all X11 gui libraries avoid round trips whenever possible.

Desktop is forgotten

[Brandybuck]

Wouldn't it be great if this kind of effort was applied to the desktop?

Re:Desktop is forgotten

[Osgeld]

there is no need to, your not running a 32 bit 700Mhhz machine trying to compete in 2013

Re:Desktop is forgotten

[EzInKy]

Wayland is specifically for the desktop, multi-device network graphics are the least of its concerns.

Re:Desktop is forgotten

[tlhIngan]

there is no need to, your not running a 32 bit 700Mhhz machine trying to compete in 2013

Except said 700MHz machine is running a fairly modern and high end GPU.

The processor in it was designed for media tanks and media players - think Roku, WDTV, AppleTV, Popcorn Hour, and other such devices. The CPU load for those things is low (just enough to display a UI and handle streaming the media to the GPU). The GPU is capable of handling decent 3D performance at 1080p resolution as well as video decode and other tasks.

It's why the processor is weak and why XBMC is stuttery, but when using a GPU accellerated task like video playback, it can play 1080p video.

All these guys have done was to exploit the power of the GPU - including the 2D accelerator, to improve performance even more. The less CPU devoted to graphics task, the more for your tasks that can't use the GPU.

Re:Desktop is forgotten

[Blaskowicz]

This uses special 2D hardware you find in cell phone chips and some gaming (handheld or home) consoles. It offloads scaling, color space conversion, maybe rotation, JPEG decoding etc., maybe encoding the output of a digital camera; on a PC's graphics card you try using the video scaler but it's more limited and "fixed function".

For instance you can look at "Video Display Controller" and "Image Processor" on these diagrams (not too sure about the first one)
http://images.anandtech.com/doci/3912/boxee-02.gif
http://images.anandtech.com/reviews/SoC/NVIDIA/Tegra2/tegra2blocki.jpg

On a PC they just send everything to OpenGL instead, which is more wasteful but because the computer is so powerful and can afford power to be wasted, the Wayland/Ubuntu/Gnome 3 devs think it is okay for you (of course it is disastrous if your OpenGL driver is not up to par, doesn't exist or if you want to run a desktop in a VM)
Intel Quicksync could probably be used, it's a kind of DSP you find in recent Intel CPUs but it's maybe not supported on Linux and Intel disables it on Celeron and Pentium (assholes).

Re:Desktop is forgotten

[gl4ss]

there is no need to, your not running a 32 bit 700Mhhz machine trying to compete in 2013

Except said 700MHz machine is running a fairly modern and high end GPU.

The processor in it was designed for media tanks and media players - think Roku, WDTV, AppleTV, Popcorn Hour, and other such devices. The CPU load for those things is low (just enough to display a UI and handle streaming the media to the GPU). The GPU is capable of handling decent 3D performance at 1080p resolution as well as video decode and other tasks.

It's why the processor is weak and why XBMC is stuttery, but when using a GPU accellerated task like video playback, it can play 1080p video.

All these guys have done was to exploit the power of the GPU - including the 2D accelerator, to improve performance even more. The less CPU devoted to graphics task, the more for your tasks that can't use the GPU.

...for varying definitions of decent. it's crap. it's got a video decoder though.

however the grandparents "wouldn't it be great if this was applied to desktops"... well duh doh fucking puh. it is. that's what the 2d acceleration craze and installing drivers for windows 3.11 was all about!

Re:Nice. Let me know when Wayland has networking

[EzInKy]

Wayland isn't about networking, it's about being pretty on a single device. Perhaps in the end their efforts might be incorporated into a proper networking graphical system like X though, so I earnestly encourage them to push on with their work!

Easier fix

[Blaskowicz]

Configure your window manager to not show the windows's content when you move them.
Job done! my 386 could do that. Dunno where's Openbox's setting for that but xfwm4 has it as a checkbox in a GUI tool.

Re:Easier fix

[Anonymous Coward]

Also, stop trying to use these "graphic" and go back to a proper text mode!

Re:Who gives a shit about the raspberry pi?

[Blaskowicz]

I can't afford one because of the need for a memory card and HDMI monitor.

arguably your N4 is a palmtop computer

[Chirs]

Modern smartphones are small computers that happen to have suitable hardware for accessing the voice network. It really is disingenuous to call these devices "phones", because you can still get feature phones that do basic voice/text/web with far less than 2GB (albeit with much less flexibility).

Sample base

[dutchwhizzman]

Your statistics are most likely based on a sample base of 1 person, yourself. Standard deviation of your survey results make scientific value of said results a bit questionable.

Re:Who gives a shit about the raspberry pi?

[Guspaz]

I want to put a real-time room schedule display on televisions for a convention I work for. I could spend a few hundred dollars on a laptop for each television, or I could duct tape a $25 raspberry pi to the back of it and accomplish the same thing at a significant cost savings...

I can see a value in this sort of ultra-low-cost hardware, is it really so hard for you to? Not every use case requires high performance. In my case, cycling through a bunch of pre-made images (or perhaps I'll throw up a fullscreen web browser and do it live with scripting) on a television doesn't require much more than that. Well, except an RTC, which the Pi lacks and will require a $20 add-on, but the overall cost with accessories still ends up at a third of a crappy refurbished netbook.

What are you waiting for, Christmas?

[jones_supa]

Shouldn't Wayland be out already? The R-Pi port is cool but those demos including terminal windows and flowers are the same ones we saw a couple of years ago already. Let me guess that Canonical has Mir mostly working in a year from now.

Re:What are you waiting for, Christmas?

[Anonymous Coward]

Let me guess they are going to have Mir "mostly working" the next decade.

Re:What are you waiting for, Christmas?

[nekohayo]

What do you mean, "shouldn't Wayland be out already"? The fact that it has had twelve public releases to this day, that it is available to be installed from recent distros' repositories and that desktop environments and window managers plan to have something based on Wayland by next fall?

Please stop pretending to be stupid

[dbIII]

A reply to a post is a reply to a poster and not to some random person linked to the issue.
Of course you know this and are merely pretending to be mentally ill in order to manufacture a strawman. Why do you think such behaviour is acceptable?

Re:Please stop pretending to be stupid

[smash]

And the poster you replied to is....

Re:Please stop pretending to be stupid

[dbIII]

And what he wrote was misguided since it is the same driver for both, which is the same whether I checked the name of the person or not.
Since this place is full of clueless Wayland fanboys that know almost nothing about the topic but fill this place with a dozen repetitive Wayland stories a month I just assumed someone else with a gap was one of those instead of somebody giving it a try. Best of luck to him.

Wayland is totally irrelevant

[OhANameWhatName]

Since Ubuntu are no longer using Wayland, how could it possibly matter any longer? And given that the Raspberry Pi can't get decent graphics performance without Wayland, it would seem that the decision to bother to build any more Raspberry Pi's is completely redundant.

It's flawless logic.
... M.S.

Re:Wayland is totally irrelevant

[21mhz]

Since Ubuntu are no longer using Wayland, how could it possibly matter any longer?

I guess it's a troll, but I'll bite.
In case you are simply deluded, there are other Linux distributions in this wide world, and not all of them are driven by a businessman with a huge control motive. Canonical's NIH solution will be largely ignored outside Ubuntu, as have been Bazaar and Upstart.
Hope this helps.

Re:Nice. Let me know when Wayland has networking

[Jaruzel]

Good luck doing anything remotely bandwidth intensive or latency sensitive over the flaky as fuck USB-Ethernet on the B.

Flaky as what-now? I regularly stream 1080p blu-ray rips onto my Pi under RaspBMC and have never yet seen a problem...

-Jar

Re:Who gives a shit about the raspberry pi?

[AC-x]

SD cards too expensive? You can't afford an extra $5 on top of $35 for the Pi?

BTW The Pi also has composite video output if you're desperate.

Re:Who gives a shit about the raspberry pi?

[AC-x]

Well, except an RTC, which the Pi lacks and will require a $20 add-on, but the overall cost with accessories still ends up at a third of a crappy refurbished netbook.

If it's networked (wired or wifi) it'll pick the time up over NTP automatically

Re:Who gives a shit about the raspberry pi?

[Grench]

Not sure if trolling, but I'll assume you're not (benefit of the doubt) and answer anyway.

HDMI-to-DVI and HDMI-to-VGA adapter cables mean you don't specifically need a monitor with HDMI input. The RPi has RGB outputs as well, so you can connect it to a TV (even if your TV only has a SCART input and not RGB, you can cheaply buy an RGB-to-SCART connector that will let you use the TV's SCART input).

You don't need a big memory card either, especially if you happen to have a spare external USB hard drive - you can have the RPi boot the bare essentials from the card and run everything else off the external disk.

Re:Who gives a shit about the raspberry pi?

[arkhan_jg]

Because it's a full linux computer about the size of a credit card for $35 that runs on a couple of watts? That can do far more than say, an arduino (though arduino still rules for hardware interfacing).

For your average dev who can plop down the cash for a macbook pro or the like, it's terribly underpowered. For someone on a tight budget of money and/or power, such as for maker builds or students, it's pretty awesome. I have one running as a headless personal web server/gitbox because I can and it beats the hell out of the electricity cost of running a full blown x86 server for that job.

Getting a much snappier hw-accelerated GUI is no small thing either.

Re:Who gives a shit about the raspberry pi?

[Anonymous Coward]

HDMI is digital and VGA is analog, unless the hardware supports outputting VGA through its HDMI port (which AFAIK the Pi doesn't) then you need a not cheap active converter cable for HDMI-VGA. Also this is the first I've heard of the Pi having RGB-out, the last I checked it was composite out, which is inferior to RGB.

Yeah , right.

[Viol8]

Get back to me when Wayland is network transparent. You know , that core principal of X which is kind of important to a lot of people. Even MS eventually woke up to the fact that only being able to do stuff on the console was a bit of a show stopper for remote admin. However pixel scrapers are bloody inefficient compared to graphics primitives being sent over the wire.

Re:Yeah , right.

[ardor]

The problem is that there are no graphics primitives to send over the wire. This is part of the core principal of Wayland which is kind of important to a lot of people. It reflects today's graphics hardware much better.

Perhaps a network transparency standard that can be moved into the toolkits could help. Typically, X11 forwarding is used to show individual applications, not the entire desktop. This way, something RDP-like can be established, which not compromising one of the biggest features of Wayland.

Re:Yeah , right.

[morgauxo]

I use X11 forwarding for the whole desktop every day! I'll be interested if and when there is a way to do that with Wayland. Until then I really hope it doesn't become popular enough that applications stop being written and maintained for X.

Re:Yeah , right.

[Anonymous Coward]

This is part of the core principal

Can I talk to his deputy?

A sub $50 graphics accelerated computer

[TomGreenhaw]

that provides acceptable web browsing performance with flat screen, keyboard and mouse will hasten the transition of the desktop market as we know it. The biggest problem with the Raspberry Pi for me has been the slow desktop experience. The impact of using the hardware accelerated graphics to eliminate this issue cannot be underestimated.

Re: Who gives a shit about the raspberry pi?

[Anonymous Coward]

Yeah so is every other piece of shit sbc...you can buy better and more powerful hw (cortex a8/a9 multicore) and sbcs from any the Chinese ARM vendors (e.g. Rockchip, Allwinner, Telechips) for the same price or less that includes HDMI and maybe even a non shit (e.g. non Synopsys) USB controller.

Or, you can be part of the retard trove that continues to wank on over a pile of shitty old chips on shitty pcb designs that were obsolete before they were released.

Re:Who gives a shit about the raspberry pi?

[Guspaz]

The venue doesn't provide free wifi, so either I've got to pay for the RTC or the wifi. For the proof of concept model that won't be internet-connected, I'll probably just grab an RTC. For a potential future full-blown implementation with a bunch of units, I'd need wifi for live schedule updates anyhow, so could use that.

The problem with the rpi lacking an RTC is that many potential uses for it have it in places where internet access it not readily (or at least affordably) available.