Ubuntu Dumps X For Unity On Wayland

An anonymous reader writes "Canonical and Ubuntu founder Mark Shuttleworth has announced that Ubuntu will move away from the traditional X.org display environment to Wayland — a more modern alternative. The move means there is now little reason for GNOME developers to recommend Ubuntu as an operating system. Shuttleworth said, 'We're confident we’ll be able to retain the ability to run X applications in a compatibility mode, so this is not a transition that needs to reset the world of desktop free software. Nor is it a transition everyone needs to make at the same time: for the same reason we'll keep investing in the 2D experience on Ubuntu despite also believing that Unity, with all its GL dependencies, is the best interface for the desktop. We'll help GNOME and KDE with the transition, there's no reason for them not to be there on day one either.'"

Comments

No standards at all

[Anonymous Coward]

WTF

> The move means there is now little reason for GNOME developers to recommend Ubuntu as an operating system.

I'm getting sick of this crap "journalism". if you want to make a comment, add a comment. Don't add your opinion to the summary. Just report the facts. If you really have to, blog about your opinion and add a link to that blog, stating that it's your opinion.

Re:No standards at all

[jgagnon]

What else would a summary be except someone's opinion? Seriously. How exactly do you shrink something down to fewer words without distorting the original meaning through interpretation? And if every summary was just a cut and paste job from the original article, why not just link to the original article and leave it be?

Re:No standards at all

[jhigh]

Since when is /. journalism??!

Re:No standards at all

[nixkuroi]

This article reads less like a news story than an emotional, personal rant by someone who's puckering with contempt because he got his feelings hurt.

Tech companies make crappy decisions all the time. Ubuntu probably thought it would have more time to become the king of the desktop before realizing that soon the desktop would be irrelevant and that *nix alternatives had already beaten it to the punch for being the kings of mobile.

At this point, he should probably start thinking further down the road to gesture and voice computing. My kinect tells me that it's almost time to stop touching devices at all, and I believe it.

Re:No standards at all

[Waffle Iron]

I'm getting sick of this crap "journalism". if you want to make a comment, add a comment. Don't add your opinion to the summary. Just report the facts. If you really have to, blog about your opinion and add a link to that blog, stating that it's your opinion.

You must be new here. The summary of an article is nearly always the *opinion* of whoever submitted it. The "news" part is in the original source to which the link(s) in the summary point (assuming the original source isn't itself just an opinion or troll). The summary IS the "blog" part, and it acts as the root of the entire discussion thread. That's the way it has always worked on this site, and it's not very hard to figure out.

Re:No standards at all

[TheCRAIGGERS]

A summary should be as factual as possible. A cut and paste job from the article, aka, an excerpt, is just fine. It's just like the 'breaks' that many blogs use, and just like the 'Continued on page A3' that newspapers have used for decades- you give a summary of the story up front, and if the reader feels like they would benefit from reading the rest, they do so.

This is opposed to what you describe, which is in my opinion bad journalism. Taken to the extreme it's like seeing a summary in a newspaper that reads 'FREE BOOBIES, continued on A4' and then turning to find an article totally unrelated.

Re:No standards at all

[DNS-and-BIND]

Dude. Have you been alive for the past 10 years? Sneaking a biased opinion into an otherwise factual story is a technique that's gotten a TON of use. Not much incentive to be a journalist otherwise, fact-based reporting is what those hick writers in flyover country do and it certainly won't win you any industry awards.

Re:No standards at all

[bsolar]

In this case it's not even true. The offending part lacks the quotation marks but actually comes from the original article.

Re:No standards at all

[Java Pimp]

That opinion WAS in TFA... if you had read it...

Re:No standards at all

[noidentity]

I think the comment is entirely appropriate. After all, as someone who contributes to Slashdot comments, I have no reason to recommend any other website to anyone. Why would I? Clearly, if it's not Slashdot, nobody could benefit from it. For the same reason, a GNOME developer would never have any reason to recommend any other desktop environment to GNOME to anyone.

Re:No standards at all

[Tharsman]

Although I generally agree with this feeling, if you read TFA, you would find this quote:

There’s now little reason for these GNOME developers to recommend Ubuntu as an operating system.

So as you can see, it's not something the summary writter made up, he just pasted something that was already in TFA, with just one word changed by a short phrace to better fit the short summary context: "There's" with "The move means there is"

If you want to insult the article itself, go for it, but at least in this one case, your insult of the summary is horrendously out of place.

Re:No standards at all

[jemtallon]

From the article: "There’s now little reason for these GNOME developers to recommend Ubuntu as an operating system."

So... slashdot did a good job?

Re:No standards at all

[TheCRAIGGERS]

I was only replying to the parent. The GP was indeed wrong and obviously didn't read the article before jumping to conclusions.

But in his/her defense, it was a first post in an article about linux, by an AC, and didn't contain anything about a 'frosty piss'. It's still a definite improvement.

Re:No standards at all

[Hooya]

> it's almost time to stop touching devices at all

I've been touching devices for quite a while and I am going blind.

Seriously. I'm supposed to pick up my prescription glasses today.. for reading on the computer monitor.

So quit touching your device. You'll go blind.

Re:No standards at all

[thePowerOfGrayskull]

Like this:

Canonical and Ubuntu founder Mark Shuttleworth has announced that Ubuntu will move away from the traditional X.org display environment to Wayland — an alternative display server. Shuttleworth said, "We're confident we’ll be able to retain the ability to run X applications in a compatibility mode, so this is not a transition that needs to reset the world of desktop free software. Nor is it a transition everyone needs to make at the same time: for the same reason we'll keep investing in the 2D experience on Ubuntu despite also believing that Unity, with all its GL dependencies, is the best interface for the desktop. We'll help GNOME and KDE with the transition, there's no reason for them not to be there on day one either"

Re:No standards at all

[Kjella]

The move means there is now little reason for GNOME developers to recommend Ubuntu as an operating system.

I'm getting sick of this crap "journalism". if you want to make a comment, add a comment. Don't add your opinion to the summary.

What about crap commenters? RTFA:

There's now little reason for these GNOME developers to recommend Ubuntu as an operating system.

Re:No standards at all

[Digicrat]

Actually, that is an excerpt from TFA. It's still an inflammatory opinion to be in a news article, but in this case the fault is the original article and not the /. summary (hmm, is that a first?).

Re:No standards at all

[Dog-Cow]

Please kill yourself. That quote is directly from the linked "article".

Re:No standards at all

[chrb]

Indeed. The headline is alarmist - "Ubuntu Dumps X For Unity On Wayland" makes it sound like they just made a huge change without consulting anyone, but Shuttleworth does say they have consulted others, and he predicts that it will take a year to get the first images out, and 4 years or more to shift applications onto Wayland. Shuttleworth is talking about a long-term direction, and it doesn't seem to be a rash decision - Intel and Nokia both appear to be backing Wayland for mobile devices.

Something like this was bound to happen after Google decided not to use X for Android. The Linux world would benefit greatly from a fast and lightweight display server that has a common codebase for mobile devices and desktops, and can be used as a backend for Android, Meego, KDE and Gnome.

Re:No standards at all

[camperdave]

Since when is /. journalism??!

Since the moment it slapped the words "News for Nerds" on its website.

Re:No standards at all

[Anonymous Coward]

Yeah, that bit confused me as well and didn't really add anything valuable.

Re:No standards at all

[jgagnon]

I read the original post as an opinionated piece about two opinionated articles, not someone trying to be a journalist and failing.

I haven't been reading Slashdot as long as many of you, but I can say I've VERY rarely read a summary here that I would remotely consider "professional journalism". So I'm confused as to why so many readers apparently expect something different than what they consistently get?

A summary on Slashdot is like a redneck amicus brief so why try to put it at some higher-level standard that it can never achieve?

Re:No standards at all

[Anonymous Coward]

What else would a summary be except someone's opinion? Seriously. How exactly do you shrink something down to fewer words without distorting the original meaning through interpretation? And if every summary was just a cut and paste job from the original article, why not just link to the original article and leave it be?

There's a difference between compressing the important parts of an article, and adding your own opinion to it. There may be a fine line in some cases, but in this case, it does seem to be a bit more clearly over the edge.

Take out that one line, and it's not bad at all.

Re:No standards at all

[Narishma]

Not to mention that the title is incorrect. They are not replacing X with Unity but with Wayland. Unity is just another desktop like Gnome and KDE.

Re:No standards at all

[Lumpy]

Not really. That statement sums up everything wrong with Gnome and KDE development. Both camps hold the fault of the mess that is the Linux desktop in their hands. It's their fault that Linux is not as useable as it should be because they cant get over their petty and silly differences and join to make a single decent desktop.

I like "choice" but only when choice is between good things not half done betas.

Re:No standards at all

[rwa2]

Ha, first line of that wikipedia article:

Wayland is a lightweight display server for the Linux desktop. Started by Kristian Høgsberg, one of Intel OSTC member,[1] the software's stated goal is "every frame is perfect, by which I mean that applications will be able to control the rendering enough that we'll never see tearing, lag, redrawing or flicker"

Last screenshot on the Wayland site:
http://wayland.freedesktop.org/clutter-flowers.png

Oh well, sounds interesting. Hope they have a good time reinventing all the network transparency and other features...

Re:No standards at all

[idontgno]

If he had read TFA, this wouldn't have been Slashdot. It may in fact have been SPARTA!

Re:No standards at all

[Anonymous Coward]

If you would've taken the time to read the source article on CIO you would see that this "crap journalism" comment your referring to was copy and pasted from the original article. In fact, the article goes into a little more detail of why they see this as an issue.

Assuming that the person who wrote the summary was intending to summarize the articles they linked, then I would argue that not including a reference to those comments would have been adding "their" opinion. However, as it stands, it seems they are just referencing the opinion posted in the original article.

Not a big deal, just something you should be conscious of - In this case its probably better to attack the source and not the copy and paster.

Re:No standards at all

[Anonymous Coward]

Hope they have a good time reinventing all the network transparency and other features

They won't. Wayland is intimately tied to the Linux DRI model and makes no attempt at cross-platform compatibility or other X features.

The idea behind Wayland, for as far as I have understood, is to offer a framebuffer-style (OpenGL) surface for applications to draw on. That's it. There are no intermediate layers, and no abstraction: the application (toolkit) must have an OpenGL backend or it won't work. I believe clutter and Cairo already have an OpenGL backend, not sure about GTK/Qt.

Backwards compatibility is achieved by (optionally) running X(server) as a sub-process of Wayland. I'm actually quite positive about this development: it's not an X replacement, but an encapsulation. It will be interesting to see how it develops. Personally, I think this will remain limited to a handful of applications (WM, decorator, configuration), at least that's how I'd like it to be. As long as there's no inherent slowdown for X apps, I fail to see the downside.

Re:No standards at all

[budgenator]

if I had mod points, I'd be torn between funny and insightful.

Re:No standards at all

[Slime-dogg]

I'm getting sick of this crap "journalism". if you want to make a comment, add a comment. Don't add your opinion to the summary. Just report the facts. If you really have to, blog about your opinion and add a link to that blog, stating that it's your opinion.

You must be new here. The summary of an article is nearly always the *opinion* of whoever submitted it. The "news" part is in the original source to which the link(s) in the summary point (assuming the original source isn't itself just an opinion or troll). The summary IS the "blog" part, and it acts as the root of the entire discussion thread. That's the way it has always worked on this site, and it's not very hard to figure out.

This is pretty much spot on. /. is a tech blog. It always has been. There are very few original /. stories, most of them being book or video game reviews.

I don't know where this idea that this place is a news source came from.

Re:No standards at all

[Anonymous Coward]

He He He. He said "boobies!"

Re:No standards at all

[budgenator]

Ubuntu probably thought it would have more time to become the king of the desktop before realizing that soon the desktop would be irrelevant and that *nix alternatives had already beaten it to the punch for being the kings of mobile.

At this point, he should probably start thinking further down the road to gesture and voice computing. My kinect tells me that it's almost time to stop touching devices at all, and I believe it.

Oh yeah like

Android is a mobile operating system initially developed by Android Inc., a firm purchased by Google in 2005.[4] Android is based upon a modified version of the Linux kernel. It is a participant in the Open Handset Alliance.[5] Unit sales for Android OS smartphones ranked first among all smartphone OS handsets sold in the U.S. in the second and third quarter of 2010,[6][7][8] with a third quarter market share of 43.6%.[9] Android

  well honestly Android, a *nix OS isn't the big kahuna, but

iOS is Apple's mobile operating system. Developed originally for the iPhone, it has since been shipped on the iPod Touch, iPad and Apple TV as well. Apple does not permit the OS to run on third-party hardware. As of October 20, 2010 (2010 -10-20)[update], Apple's App Store contains more than 300,000 iOS applications[1], which have collectively been downloaded more than 7.5 billion times, as per a keynote on October 20, 2010. As of May 2010, it had a 15.4% share of the smartphone operating system market in terms of units sold, ... iOS is derived from Mac OS X, with which it shares the Darwin foundation, and is therefore a Unix-like operating system by nature. iOS (Apple)

a *nix OS certainly is.

Re:No standards at all

[Anonymous Coward]

Since when is /. journalism??!

Not for several years now sir.

Re:No standards at all

[Sloppy]

My kinect tells me that it's almost time to stop touching devices at all, and I believe it.

I challenge you to a narrative contest. I will type, using a keyboard. You will enter text however you think best with whatever equipment you choose.

I don't know if I'll win, but if I don't win, it'll be because you chose keyboard too.

Re:No standards at all

[electrosoccertux]

This article reads less like a news story than an emotional, personal rant by someone who's puckering with contempt because he got his feelings hurt.

NPR sounds the same way to me. But it's professional journalism. You have to rustle feathers, especially on a site like this, to keep people interested.

Re:No standards at all

[mweather]

>What else would a summary be except someone's opinion? How exactly do you shrink something down to fewer words without distorting the original meaning through interpretation? By that definition, all news articles are opinion, because they just sum up things that happened into fewer words.

Re:No standards at all

[thePowerOfGrayskull]

Fair point

Re:No standards at all

[thePowerOfGrayskull]

Hope they have a good time reinventing all the network transparency and other features..

Sounds like they don't need to, for that's not their target. When you think about it, X11 was designed for a very different architecture than its primary use in consumer linux space today.

Re:No standards at all

[Anonymous Coward]

I agree with you, but what about all the stuff that sits atop X, like VNC, Citrix, FreeNX / NoMachine, etc.?

Do we have a fast app client / server / screen scraper to run X and Wayland apps remotely? Or will it be like Windows and Mac, where some apps are remote and multi-user, while others aren't?

One of the great things about Linux / Unix is the ability to remote basically EVERYTHING, with few exceptions. If we lose this with Wayland, people will dump Ubuntu, plain and simple.

Can I X into Wayland apps *and* Wayland into X apps?

I'd rather Wayland be X Improved(tm), instead of X Replaced(tm).

There are PLENTY of improvements that can be made to X. Remote control via the high-performance NoMachine is a good example. (I'm using it right now, in fact).

Re:No standards at all

[jgagnon]

They pretty much are. If I choose which parts to include, exclude, paraphrase, quote, etc., then how is my opinion and position not an influence (at the very least) on the resultant piece?

Re:No standards at all

[gknoy]

The trouble is, they both believe fundamentally different things about what it means to have a Good UI.

If my idea of a good dessert is brownies, and yours is lemon meringue pie, it's really hard to combine the two (or compromise) and yeild something that doesn't suck. I'll look at pies you make and say, "This dessert lacks sufficient chewiness" or "This crust makes a fundamentally problematic dessert interface", and you'll look at my brownie-like concoctions and want more crust, less chocolate, more lemon (!?), and some creamy bits. Sometimes, two very different and competing ideas work well by servicing different niches.

(Though, perhaps Schadenfreude pie would be a good compromise between pie and brownies, so my example may be flawed.)

Re:No standards at all

[MikeBabcock]

Use the English dictionary near you and tell me what the word "summarize" means. It certainly does not mean to altar or add to in any way. It means to reduce in size while maintaining the broad concepts given in the original.

Summarizing an article has nothing to do with your opinions or thoughts on it. A review does, a summary doesn't.

Re:No standards at all

[Noitatsidem]

I'd like to add that because news articles do sum things up into fewer words, they often times add opinionated statements to get their point across: I seriously doubt you'll find many news articles, or someone talking about the news with no bias. Hell, even Shuttleworth's blog posting had his own opinions in it- It's natural for opinions to make their way into any sort of news. It's simply interpretation of the actual facts.

Re:No standards at all

[Anonymous Coward]

It's still a shame. X is one of the strenghts of UNIX and all UNIX-like systems.

Canonical should put their support into improving X.

Now Ububtu will become incompatible with all other versions of Linux if networking apps is needed.

Re:No standards at all

[muckracer]

> So quit touching your device. You'll go blind.

That's the advantage of being raised catholic:

$ touch /dev/myself0
touch: cannot touch `/dev/myself0': Permission denied

Everything's still nice and clear.

In God# mode, however, touch(1) works. Therefore, God is blind and he doesn't see, if you install the Slapper rootkit and give yourself permissions for touch(1) and eject(1). What He don't know, won't blind you! So go ahead and don't worry about it. :-)

Re:No standards at all

[Noughmad]

They're replacing (Gnome on X) for (Unity on Wayland).

Re:No standards at all

[HermMunster]

Personally I think it is a bad move. It attempts to make a change when some other serious issues still are outstanding in Ubuntu. This past release of Ubuntu showed more of the same "making changes for change's sake" It didn't help my opinion of it in that it wound up wiping one of my very important installs.

The news of Unity replacing Gnome was not opportunistic. They are making the change even though there's no reason to make the change. The point here is that there is no real reason to make the shift from Gnome to some other rather immature desktop manager (that Unity is). Then, to remove X for another extremely new and very much untried technology to replace one of the most significant parts of Ubuntu just doesn't cut the mustard. This is a pretty sad tale of one announcement being used to obfuscate the other more significant change. On top of that, too many changes at once is never a good thing.

So, we have a new display manager requiring changes to drivers, on a technology that was developed a short time ago and hasn't been tried in any systems other than Meego (which itself has hardly been used in any hardware), to replace something that may have baggage but has survived the test of time. Add to that a desktop manager which is primitively immature and is essentially a dumb interface that is actually confusing to users because it doesn't hold to the desktop paradigm instead it is more like a DOS menuing system implemented in a GUI environment.

Re:No standards at all

[Raenex]

I can say I've VERY rarely read a summary here that I would remotely consider "professional journalism".

A lot of them are just copies from professional journalism, so I disagree. Besides that, a lot of them don't editorialize. A few a day do, and I'd rather see complaints than just letting Slashdot slide into Digg.

Re:No standards at all

[spun]

Hey, that's not fair! I know Amicus Brief, he's not a redneck. He may be backwards, uneducated, opinionated, a hick, white trash, a redneck, and an idiot, but the man is no redneck!

Re:No standards at all

[Risen888]

Because god knows it wouldn't be a thread about the free desktop without a "OH NOES I'M PARALYZED BY CHOICE" troll...

Re:No standards at all

[BrokenHalo]

Do I have to be the first to point out that Ubuntu is not an operating system? [Sigh.]

Re:No standards at all

[Anonymous Coward]

You would have really loved it when Michael and Jon Katz were here :)

Re:No standards at all

[Hatta]

The problem is that nobody can make a decent desktop. For every thing you find that is broken in KDE or GNOME or XFCE, I can point to things that are broken in Windows or OS X. The only way to have a decent usability experience on any platform is to use the shell.

Re:No standards at all

[Stregano]

But if newspapers said stuff like that, I would read them everyday, and go to the continued stories. I do not think 'Free Boobies' is opinion though, it is a want and a necessity.

Re:No standards at all

[BrokenHalo]

Backwards compatibility is achieved by (optionally) running X(server) as a sub-process of Wayland.

Isn't this inherently inefficient? I can understand (I guess) why some might feel that x.org might need a re-write from the bottom up, but that doesn't help those of us who actually use the interface. Without clearly defining why x.org should be relegated to a sub-process, we are left with an impression that someone has too much time on his hands, and has the leisure to break a lot of programs for no very good reason.

Most of us who have been hanging around the various Unices for a while will have pet peeves about projects that have been unnecessarily diddled with. While I can see that x.org (formerly xfree86) is big, its supposed "bloat" derives from the fact that it has a hell of a lot to do.

If the Wayland developers can present a convincing case for why we need to introduce another process underpinning X11, I'll be interested to hear it. But from what I've read so far, it looks more like a few people are simply bored with the old interface.

Re:No standards at all

[BrokenHalo]

I wouldn't know (and don't want to) about being raised as an idolater^W Catholic, but at least some things are still simple enough:

$ man woman
No manual entry for woman
$

Re:No standards at all

[Lumpy]

I'm not paralyzed by choice. I like choice.

Problem is currently my choices are half done betaware, and half done betaware.. NOTHING is a complete solution... They both focus on stupid shineys instead of fixing major problems that have existed for a very long time.

Re:No standards at all

[spitzak]

GTK is using Cairo.

Re:No standards at all

[Anonymous Coward]

OSX and Windows I can set everything from a nice config window. ONly very recently did this exist in a working way on Gnome and KDE, and many of those functions are working ,then broken, then working again, then broken...

Re:No standards at all

[Tetsujin]

Use the English dictionary near you and tell me what the word "summarize" means. It certainly does not mean to altar or add to in any way.

Perhaps you should grab a dictionary and look up "altar". :)

Re:No standards at all

[jgrahn]

I'm not paralyzed by choice. I like choice.

Problem is currently my choices are half done betaware, and half done betaware.. NOTHING is a complete solution... They both focus on stupid shineys instead of fixing major problems that have existed for a very long time.

What are these major problems? I make the third choice and ignore both Gnome and KDE. ctwm is my window manager, and I don't really need any of the Windows-ish crap.

Re:No standards at all

[MichaelSmith]

The selection model might be different between X and Wayland apps. I would hate it if I was unable to cut and paste between applications.

Re:No standards at all

[Knuckles]

Except that current Unity in 10.10 is largely still Gnome except for a quite rudimentary (for a desktop) shell.

Re:No standards at all

[Anonymous Coward]

Do I have to be the first to point out that Ubuntu is not an operating system? [Sigh.]

I hope you're the last, because Ubuntu is an OS. It's a renamed distribution of GNU and Linux.

Re:No standards at all

[wizardforce]

Slashdot is supposed to be news for nerds so generally even though we know better, we expect a news/discussion site aimed at the nerd demographic to not emulate the other segments of mainstream media.

Re:No standards at all

[Anonymous Coward]

And hence we're back to point of the grandparent's comment. As far as desktop environments go, there's not just a first choice, second choice, and a third choice like yours. In Linux land, there's a 25th choice, a 57th choice, and even a 100th choice, and all of them suck because they're all 'half done betaware'. It's the belief that such a system is a good thing that's keeping normal computer users away from Linux.

Re:No standards at all

[Homburg]

While I can see that x.org (formerly xfree86) is big, its supposed "bloat" derives from the fact that it has a hell of a lot to do.

This is true, but much of what X11 can do is not used a lot of the time. With recent toolkits (for instance GTK using Cairo for drawing and Freetype for font rendering), the X server mostly just mediates between the application and the compositor. This is kind of backwards, and, indeed, is "inherently inefficient" - the X11 core contains a stuff that isn't always used, while the most frequently used things are outside of X11, with X serving only as a mediator.

If the Wayland developers can present a convincing case for why we need to introduce another process underpinning X11, I'll be interested to hear it.

The idea of Wayland isn't to have another process underpinning X - the idea is to have applications talk directly to Wayland where that makes sense, while the extra features of X can be brought in where they are required.

Re:No standards at all

[Anonymous Coward]

I completely agree, what a load of crap. Especially since if the author had done an OUNCE of research, he would know that GTK+ is already drawing on Wayland. Wayland is a fantastic thing for the Linux Desktop. Ditching X does not mean ditching GNOME. I may have to quit subscribing to slashdot if it is going to be full of this biased bologna.

Re:No standards at all

[ArsonSmith]

I choose cut and paste with my mouse...

Re:No standards at all

[ushering05401]

Jon Katz.

Re:No standards at all

[TheBilgeRat]

You mean I'm going to have to learn to speak vim commands?!

that's great- I get angry, flip off my kinect, and next thing I know its gone and 360dd'ed.

Re:No standards at all

[HGG]

Microsoft holds some of the OPENGL patents So now isn't Ubuntu in danger of "submarine" patent cases?
http://www.zdnet.com/news/microsoft-claim-shakes-graphics-world/124000

Usually these cases don't surface until the transition is well locked-in.

Re:No standards at all

[Anonymous Coward]

wow so many non apparent but obvious redundancies. teheh

Re:No standards at all

[Anonymous Coward]

I want to learn the php, my Ideal is become a hacker so i will get here often to learn the php, I am a lucky people.Because have you,thank you.
in here this is my Black's Web site recently:
Replica watch Replica watches Replica swiss watches Discount watches wholesale Replica cartier watches Replica breitling watches Louis Vuitton Watches Hermes h watches Bell & Ross replica watches a lange & sohne watches Replica breguet watches Replica bvlgari watches Replica gucci watches Replica iwc watches Longines replica watches Franck Muller Watches Replica tag heuer watches Replica chanel watches please give me some suggestion.

Re:No standards at all

[mcclungsr]

before realizing that soon the desktop would be irrelevant

[citation needed]

Sounds like mobile hype.

Re:No standards at all

[turgid]

The Linux world would benefit greatly from a fast and lightweight display server that has a common codebase for mobile devices and desktops, and can be used as a backend for Android, Meego, KDE and Gnome.

Good god, No!

X doesn't need replacing!

This "X needs replacing" meme keeps getting trotted out once in a while, and each time it doesn't get any more correct.

X is lightweight, Open, free, robust, well understood, cross-platform, ubiquitous and fast. This has been argued to death all over the internet and is evident by X's longevity and the fact that it is easily extended, whilst retaining backwards compatibility, to implement the trendy features of the day.

By all means, please let someone's toy Linux distribution throw away X for some new-fangled mobile-phone compatible junk, but the Linux and unix world in general does not need to throw away two and a half decades of success because of what people have on their mobile phones.

One last piece of useless information: X was designed to run on 1 MIPS machines. That's right, computers without graphics accelerators that could do a measly 1 million instructions per second. Your cheap Dell desktop is 10 000 times as fast, plus the graphics is done in separate hardware.

Re:No standards at all

[wlad]

yeah, what they forget for their bout of sensationalism is that Unity is simply a variant of GNOME.

Re:No standards at all

[wlad]

Yes. Unity is based on GTK and Gnome. It's as asimple as that. I don't see what's all the panic about. You will still be using the Gnome applications. It would have been a bigger jump if they went for KDE/QT.

Re:No standards at all

[Anonymous Coward]

When you say "*nix alternatives", do you mean alternatives to UNIX, like Android and iOS, or do you mean alternative UNIXes, like GNU/Linux and BSD/Mach?

Re:No standards at all

[Anonymous Coward]

People notice very fast when a summary on Slashdot is biased, that's because it's obvious to everyone. We're lucky it's nothing like "real" journalism.

Re:No standards at all

[Anonymous Coward]

What, so iOS, that famous BSD/Mach derivative via the OSX it was born from, is an alternative to UNIX? And Android, that famous Linux derivative, is an alternative to UNIX?

I mean, yes, Linux isn't UNIX and we're all nerds and know that, but still, try and have a clue what you're talking about, yeah?

Re:No standards at all

[Anonymous Coward]

Lemon meringue pie topped with brownie chunks? Sounds quite tasty.

Re:No standards at all

[commodore64_love]

He may be "confident" we'll be able to retain the ability to run X applications in a compatibility mode, but I'm not. MY PS3 doesn't run a whole host of PS1 or PS2 games even though Sony claimed it would. (So I bought a space PS2 instead.) Windows Vista and Seven doesn't run old 3.1 or 95/98 apps. Mac OS X doesn't do well with classic PPC or 68000 apps.

Nope. Not confident at all.

Re:No standards at all

[gilgongo]

"My kinect tells me that it's almost time to stop touching devices at all, and I believe it."

Voice control (or the easy possibility of it) on mobile devices has been around for years. What's changed recently that you should think it's going to become something that's worth doing?

Re:No standards at all

[jbolden]

X's killer feature was network transparency which was wonderful in the days of low security networking. Now that we have higher security its main feature can't be used effectively. We have two and a half decades on windows and mac and their model won for a reason.

Re:No standards at all

[Have Brain Will Rent]

Yeah... who'd put any value on fact based reporting...

Re:No standards at all

[turgid]

Their (Windows, Macintosh) model didn't win. Everywhere I go the Windows (and Mac) heavyweight, fragile, non-solutions are being replaced by X (on Linux).

People have been telling me that Windows is the future since 1992. Everywhere I go, I see Windows (and other proprietary OSs) being replaced by Linux.

How on earth can you argue with a straight face that Windows has a superior model, especially regarding security, let alone reliability?

Have you ever tried to get a week's work done using Windows Terminal Server Client? I'm not sure just what they meant by "terminal."

Re:No standards at all

[t2t10]

As long as there's no inherent slowdown for X apps, I fail to see the downside.

X apps aren't just these isolated things anymore drawing a bunch of lines on the screen. We have window management, keyboards, and tons of other functionality. Much of that still relies on X11 and won't be supported by Wayland. If we keep going down this path, you'll only be able to write any kind of useful desktop apps with Gnome and Qt. It's a huge step in the wrong direction.

Re:No standards at all

[Anonymous Coward]

Here's the thing: it's not going to take "years" to move to wayland. People using Free Drivers and Native Ubuntu Software, like me, will be using Wayland experimentally in six months to a year, and everyone else is going to have worse power consumption and RAM usage than Windows XP like always. Then people like me are going to get all self-righteous about our choices of hardware drivers and software and start telling other people that they wouldn't have the problems they do if they just didn't use Firefox/nVidia/OpenOffice/Catalyst/whatever, while Abiword mangles our documents like the Elric brothers' undead-not-quite-mother and Ease saves our presentations in some file format no one else has ever even seen before. Then, Canonical will add a Native Ubuntu + Launchpad hosted certification to the software center along with donations and developers are going to get all pissed because suddenly the ones who listen to their users get the money. They'll go on some rant about how capitalism is dangerous and a threat to our ideals and way of life but no one will care, and then Canonical will sue Apple over Launchpad to get legal and stable ITMS integration in Banshee and free unlimited .ac3 licenses for Ubuntu users. Ubuntu will blow past OS X in market share and three months later some Gentoo user will assassinate Mark Shuttleworth, Jimmy Wales will be hit by a bus, Torvalds will OD on heroin, Obama will be impeached because he pirated a KISS album, Jamendo servers will suddenly go offline forever, and Free Culture will go down in history as a cautionary tale.

Re:No standards at all

[jbolden]

People have been telling me that Windows is the future since 1992. Everywhere I go, I see Windows (and other proprietary OSs) being replaced by Linux.

On the desktop? Linux as a desktop OS in terms of percentages has been relatively flat in terms of usage for the last decade.

Re:No standards at all

[Risen888]

Seriously? You're going to look at Gnome 2.3 and KDE 4.5 and call them "half done betaware" and not a "complete solution?" As far as I am aware, they are the only "complete solutions" presently being offered in desktop software, for free or not. Every other vendor out there (by which I mean, of course, Microsoft and Apple) offer piecemeal half-assed stuff that you buy seperately and pray it all plays nice together. Use Windows 7 out of the box, then go KDE out of the box, and come talk to me about "half done."

Re:No standards at all

[ThePhilips]

If we keep going down this path, you'll only be able to write any kind of useful desktop apps with Gnome and Qt.

Ask anybody who programmed using Xlib and you would (1) find very of fans of it and (2) be immediately advised to use GTK/Qt/wxWidgets/etc instead.

Do not forget: Xlib doesn't even support widgets, one has to use toolkits for that or draw all the stuff manually.

Re:No standards at all

[turgid]

Yes, on the desktop, but I work in engineering.

Re:No standards at all

[jbolden]

OK well you are some sort of niche. That's not in any way reflective of what is happening in the wider community even the engineering community. Niches can do all sorts of odd things because of application availability / pricing models.

Re:No standards at all

[drinkypoo]

Haven't you ever run an X server on Windows? They work fine, as long as you're not using Cygwin's X.Org... which is pretty great for free, to be fair.

Indeed if we can get something that permits compositing features that actually work, supporting X through a compatibility layer might actually result in a better user experience.

Ok great for beginners

[betterunixthanunix]

...but I still know a LOT of people who forward X over SSH, and there are still a lot of professors who are advising their students (at least in the engineering schools I have seen) to do the same. I guess this is one of those times that just saying, "I use Linux!" will not convey what people think.

Re:Ok great for beginners

[Anonymous Coward]

"I use Linux!"

Slackware? Debian?

RedHat? Mandrake? SuSE?

Fedora? Gentoo?

Windowmaker? Blackbox? Enlightenment? KDE? Gnome? Bluecurve (lol)?

What? What?

"I use Linux!" has never conveyed what people think it means.

Re:Ok great for beginners

[Anonymous Coward]

I use a Mac, and I can still forward X over ssh. It only takes a simple app to handle the client side to make it work.

Re:Ok great for beginners

[somersault]

I still know a LOT of people who forward X over SSH

I don't think that's relevant. They'll still be able to use encrypted VNC, or some other solution of their choice.

Re:Ok great for beginners

[sd.fhasldff]

Wayland is a display server, like X. Why wouldn't it be possible to forward Wayland over SSH?

Re:Ok great for beginners

[AaxelB]

Fortunately, X is not Linux, and Linux is not X, which might help clear up some of your confusion. When someone says "I use Linux!", it means exactly that, and you can't really assume a priori what else they've got. Of course, most Ubuntu users will say "I use Ubuntu!", which should make things easier.

Personally, I rarely do anything that really depends on X being X, so my reaction is essentially "huh, I wonder how that'll work out."

Re:Ok great for beginners

[betterunixthanunix]

Because not all the applications people want to forward are written for Wayland; one that comes to mind is a VLSI tool from Cadence, which is proprietary software that is often encountered in EE curricula (for VLSI courses and whatnot), which I doubt will be updated to Wayland any time soon. People have come to rely on an X server, specifically, being available to them.

Re:Ok great for beginners

[jedidiah]

Linux is Unix, and X is a part of any proper Unix.

Although I am more concerned with performance going to sh*t because some idiot thought it would make sense to break with the past and all of the driver development that's already been done.

Shuttleworth is really pretty irrelevant here.

Re:Ok great for beginners

[betterunixthanunix]

Fortunately, X is not Linux, and Linux is not X

Yes, I know that, but for a lot of people, "I use Linux!" implies the availability of an X server.

Personally, I rarely do anything that really depends on X being X, so my reaction is essentially "huh, I wonder how that'll work out."

Speaking for yourself; I see plenty of people who depend on X being X, and plenty of people who are being advised to depend on X being X. A move to Wayland will create all kind of confusion for those people, who implicitly assume that any Linux distro which has a graphical environment is using X.

Re:Ok great for beginners

[Anonymous Coward]

1. Linux is not Unix

2. X is neither part of Unix nor required for it.

Anything else you'd like to add to this discussion?

Re:Ok great for beginners

[Culture20]

I still know a LOT of people who forward X over SSH

I don't think that's relevant. They'll still be able to use encrypted VNC, or some other solution of their choice.

Forwarding an X session (ie running Firefox on the remote machine and having it display on your local X daemon appearing as a local program) is far different from running VNC (using a full desktop environment on the remote machine), even though both can be run via ssh tunnels.

Re:Ok great for beginners

[Anonymous Coward]

It's always conveyed, "I use some strange computer program that the listener understands nothing about. I hear lately, though, that it sort of runs Windows." Anyone who actually uses Linux understands that almost nobody's Linux boxes are the same.

Re:Ok great for beginners

[Joehonkie]

Linux is NOT Unix. That's been pretty important since day 1. And Unix doesn't need X, since OSX is Unix (real Unix, not Linux) and only runs X under it's main (non-X) window manager as needed (just as Shuttleworth is talking about doing with Wayland). As a matter of a fact, you can have a Unix or Linux box without any windowing system on it at all! It's amazing, I know, but totally possible.

Re:Ok great for beginners

[Culture20]

Linux is Unix

GNU Linux is GNU is not Unix Linux

X is a part of any proper Unix

Always? I was under the impression that having X on some servers is a waste of resources.

Re:Ok great for beginners

[Anonymous Coward]

Tons of people use Linux. They watch their new HD TV.

Re:Ok great for beginners

[makomk]

Wayland is a very minimal display server. It requires clients to access the graphics card hardware themselves using the DRM kernel API in order to actually render anything, and you can't do that over a network. Basically, Wayland only works when the display server and its clients are running on the same machine, and that's a deliberate design decision that can't be changed easily.

Re:Ok great for beginners

[jellomizer]

Forwarding X isn't really an ideal way to do things anymore.

VNC/RDP and other protocalls are MUCH faster then using X. The professors would still have students do their code with punch cards if they had their way.

most X applications have gotten so graphically robust that their design isn't optimal anymore. Back in the old days where it was just vector graphics where CPU were fast compared to bandwidth meant a some simple box drawings made a robust X app. Now almost every element today has some sort of bitmapped graphic tied to it. And making it slow for remote use.

It is not saying the X doesn't have any advantages over others... It does however if you weigh the tradeoffs you will find that people are suffering more then they are being helped.

Re:Ok great for beginners

[computational super]

X is a part of any proper Unix.

Proper Unix doesn't have any graphical display capabilities at all.

Now get off my lawn.

Re:Ok great for beginners

[rpopescu]

You have no idea what you're talking about, do you?

Re:Ok great for beginners

[dpilot]

You mean like the fact that I need to use the same Cadence you're talking about as part of my day job, as well as a whole host of other X-based VLSI CAD applications. Every now and then I need to work from home, and X lets me do that. To be sure, sometimes I use VNC, but sometimes I run the X tools native on my home system, too. Different tasks call for different approaches.

Leaving work out of it, sometimes I just like to run some GUI tools on my server, with the display exported back to my desktop. My server doesn't even have an X server installed.

I strongly suspect that the people who pooh-pooh the networking capabilities of X never got used to using them.

Re:Ok great for beginners

[Anonymous Coward]

OSX isn't Unix. "UNIX certified" is not UNIX. z/OS and OS/400 are not UNIX for the same reason.
 
  To the best of my knowledge, the only UNIX that is still on the market is Solaris, HP-UX, AIX, and SCO OpenServer/UnixWare.

Re:Ok great for beginners

[2.7182]

Nice troll! You managed to choose a topic that is probably as complex and volatile as Kirk vs. Picard, but yet is not as familiar.

Re:Ok great for beginners

[Digicrat]

Not true. There are plenty of people that view certain (ie:headless servers, embedded machines, etc) Linux/Unix systems as a pure shell with no GUI.

Re:Ok great for beginners

[swillden]

People have come to rely on an X server, specifically, being available to them.

I had the same thought, but after looking at the Wayland architecture I'm less concerned. Here's a relevant quote from the Wayland architecture pages:

Wayland is a complete window system in itself, but even so, if we're migrating away from X, it makes sense to have a good backwards compatibility story. With a few changes, the Xorg server can be modified to use wayland input devices for input and forward either the root window or individual top-level windows as wayland surfaces. The server still runs the same 2D driver with the same acceleration code as it does when it runs natively, the main difference is that wayland handles presentation of the windows instead of KMS.

So it sounds like application developers will have a choice of using the Wayland window system directly, or using the X protocol to talk to an X server which uses Wayland to display its output. In practice, of course, no one will do either. Application developers use toolkits like Qt, GTK, wx, etc., so what will probably happen is that the toolkits will choose either the X or the Wayland protocol, perhaps dynamically based on the available options.

I was pretty sure when I went to look at the Wayland stuff that this is a bad idea. After reading about it a bit, though, I'm not so sure. Wayland is designed around the notion of compositor-based display, which is clearly where everything is now or is going soon, while the compositor is a somewhat-klunky add-on to an X server. If Wayland can retain X's network transparency, streamline and simplify the graphics architecture, provide a cleaner and less...bizarre... protocol, and also allow native X apps to continue running without issue and to be gradually ported from the X protocol to the Wayland protocol as it becomes convenient... I think it may be a very good idea indeed.

Re:Ok great for beginners

[Anonymous Coward]

The Open Group (owners of the infamous UNIX(tm) trademark) say OS X, Z/OS and OS/400 are UNIX. Anonymous Coward says they're not.

Well, thanks for clearing that up for us...

Re:Ok great for beginners

[jedidiah]

> You have no idea what you're talking about, do you?

No. I just happen to do this for a living.

Re:Ok great for beginners

[Anonymous Coward]

I don't think that's relevant. They'll still be able to use encrypted VNC, or some other solution of their choice.

Which is slow as molasses in January over slower connections.

(Although a better solution like NX works extremely well, but that's based on X as well. Even Microsoft's RDP is faster then VNC.)

Re:Ok great for beginners

[AaxelB]

Yeah, my last comment was basically to say that I don't have a dog in this fight, so I don't particularly care what Ubuntu does. I'll probably continue using Ubuntu on my laptop, as long as it does what I want and there's nothing else that's demonstrably better.

Re:Ok great for beginners

[RCGodward]

X is absolutely required for Unix.
uniX
See?

Re:Ok great for beginners

[Anonymous Coward]

the people who pooh-pooh the networking capabilities of X never got used to using them.

These people are usually the ones too stupid to realize that X doesn't use the network when it's used locally.

Re:Ok great for beginners

[Anonymous Coward]

And Unix doesn't need X

I would tend to disagree. I mean, it's right there in the name.

Re:Ok great for beginners

[AaxelB]

1. Linux is not Unix
2. X is neither part of Unix nor required for it.

Anything else you'd like to add to this discussion?

Nice troll! You managed to choose a topic that is probably as complex and volatile as Kirk vs. Picard, but yet is not as familiar.

Nah, it's pretty well known and accepted that Linux is not Unix. Linux is certainly and undeniably Unix-like, but it's not Unix.

Not really complex. Not really volatile. Not a troll.

Re:Ok great for beginners

[somersault]

Having had a look at the Wayland site, I'd say you could do the same thing with the Wayland architecture if you really wanted to. These days considering even mobile phones can run full featured web browsers, I don't think it's much of a selling point though. It is nice from a security standpoint, but soon enough it will probably be feasible to run browsers in a virtual machine on your phone if you really want that level of security! :p

Re:Ok great for beginners

[Crudely_Indecent]

Linux is Unix, and X is a part of any proper Unix.

INVALID

I use Linux exclusively. At work, where I'm required to use Windows, so I run several Linux virtual machines. My kids run Linux. My servers run Linux. 90% of my Linux machines do not have X or any other windowing system. In my previous job, 100% of all Linux systems (several thousand) had no X server, hell, 99.99% of them were headless.

Linux is not Unix, it's LIKE Unix. X is not Linux - X runs in many environments - including OSX and Windows

Re:Ok great for beginners

[camperdave]

> You have no idea what you're talking about, do you?

No. I just happen to do this for a living.

Ah! A consultant.

Re:Ok great for beginners

[jhol13]

Exactly. VNC is usable over low latency links, X is not.

Re:Ok great for beginners

[abigor]

Sun shipped with NeWS. NeXT shipped with Display Postscript. SGI shipped with MEX and later 4Sight. I guess none of these were "proper" Unixes in your godlike eyes - someone better alert The Open Group.

Wayland reuses X's drivers. It can also host X with a negligible performance loss.

All in all, this is a great thing for desktop Linux, which needs all the help it can get. With commercial vendors rallying around Qt, which already has good Wayland support, the future looks hopeful.

Are you some kind of junior sysadmin? A little knowledge is a dangerous thing.

Re:Ok great for beginners

[Lumpy]

Yggdrasil.

Re:Ok great for beginners

[Anonymous Coward]

In (2D) X, clients tell the server what to draw, using a series-of-bytes network-transparent protocol that's been around forever.

In Wayland, clients tell the graphics hardware what to draw to an off-screen buffer. The Wayland server only knows about where to draw those buffers and how to composite them.

So, yes, Wayland could become network-transparent--once someone makes a network-transparent DRI, deals with the inherent security issues, and convinces the increasingly conservative kernel maintainers to burden their ultralight direct hardware access facility with a bunch of networking crap of dubious security.

Not gonna happen.

Re:Ok great for beginners

[Anonymous Coward]

Linux is Unix, and X is a part of any proper Unix.

Although I am more concerned with performance going to sh*t because some idiot thought it would make sense to break with the past and all of the driver development that's already been done.

Shuttleworth is really pretty irrelevant here.

I thought "GNU is not Unix."

Re:Ok great for beginners

[Anonymous Coward]

Yes far different. X sucks for remote WAN stuff, vnc and freenx are better for that.

http://www.vanemery.com/Linux/XoverSSH/X-over-SSH2.html

Re:Ok great for beginners

[squizzar]

Or wobbly links. I had to run X over an SSH tunnel on ADSL that reset every now and then. Result: All the applications close immediately. So what you needed was another machine at the remote end you could run your VNC server on, and then run the X apps from that. I don't know if I missed some kind of session persistence option but it was a pain in the arse.

While I'm on about it there's a utility for compressing the X protocol on the fly that made a world of difference to usability over ADSL speed connections. Stopped it soaking up all our bandwidth too...

Re:Ok great for beginners

[knewter]

There are an unbelievable number of use cases for forwarding X. Even the web browser example you gave - perhaps I don't know how to tunnel my IP traffic (but srsly, ssh makes it easy) and i want to modify my routes at home....it's painless to just tunnel X for kicks.

Granted, most people that know how to do one can do the other, but it's extremely useful and I will be sad to see such an architecture go, if it goes out of common use. X is great.

Re:Ok great for beginners

[Third Position]

I forward X over SSH too - on a Mac, which doesn't use X as it's default display server either. The X server runs only when I'm using an X app. I'm not familiar with Wayland, but I don't why the same arrangement couldn't be supported on Linux.

Truthfully, this is probably a good decision on Ubuntu's part. X does some cool tricks, but it's still a rather clunky, disorganized interface compared to almost any other modern one. Better to support it as a legacy option, and kick the Linux/Unix desktop into the 21st century. You can still use X when you need it, without making it your default display manager.

Re:Ok great for beginners

[jelle]

It doesn't have to be limited to only local use. It's an api, which means somebody can make an implementation that forwards the api calls into some kind of rpc calls (or work on a higher level with all sorts of roundtrip and bandwidth optimization), allowing networked use. Even the drm api can be implemented as a virtual graphics device with a network backend, but that's probably not even necessary because from the architecture pictures I don't see a reason why the compositor wouldn't be able to support a network environment. It would need a local compositor and a remote compositor that know how to talk to each other (and with each some significant code to 'hide' the network), but it would be transparent to the local application. As long as the wayland protocol allows multiple wayland compositors to operate concurrently on a system (where the client apps run), and if it lets a wayland client choose the compositor to use (with some type of access protection in there), it should be possible to implement transparent networked displaying.

Maybe it will make the networked displaying more complex than a tcp connection from the X client library to the X server, but it's certainly not impossible. The X protocol isn't ideal for all networked use (anymore) either, because it's so sensitive to network latency and it doesn't help in any significant way to be bandwidth efficient, especially with the increasing amount of client-side rendering and 3d stuff that is happening these days (and not all X protocol features work on a remote connection (for example (and afaik): xrandr, dri, etc). Perhaps a good networked remote display protocol that optimizes and compresses all that when/where necessary will work equally as well (with as much complexity) in the wayland framework as it will in the xorg framework.

So perhaps in order to support modern displays and diverse networks, remote display has to get a little more complex than a simple remote X display anyway. I don't wayland is going to make that much different.

http://wayland.freedesktop.org/architecture.html

Re:Ok great for beginners

[icebraining]

Why would you need to put in on the Kernel? The "DRI-server" can run in userland just like X or Wayland.

Re:Ok great for beginners

[swillden]

Forwarding X isn't really an ideal way to do things anymore. VNC/RDP and other protocalls are MUCH faster then using X

Bicycling isn't really an ideal way to do things any more. Automobiles are much faster than bicycles.

Okay, so not a perfect analogy, but you get the point. VNC/RDP can't forward just a remote application, they have to bring the whole desktop. They can't integrate cut and paste as seamlessly as forwarded X. They don't allow your local window manager to choose where things are placed, how they're moved, etc.

There are plenty of cases where VNC/RDP make perfect sense and forwarded X connections do not. There are also plenty of cases where the reverse is true. Different tools for different situations.

Re:Ok great for beginners

[doogledog]

Too right! Forget 'the year of Linux on the desktop', this is 'the decade of Linux in the thing'.
A catchier name is required though.

Re:Ok great for beginners

[cayenne8]

I forward X for many things too.

For instance..I set it up for Oracle installs to use the Java GUI...forward it to my desktop, etc..

I too would hate to see X and its usefulness go away.

Re:Ok great for beginners

[JayJay.br]

ROFL, I just spit my coffee all over my desk.

And here I thought that no one remembered that anymore.

Kudos to you, sir. Five-digit Slashdot UID megabonus included.

Re:Ok great for beginners

[Hatta]

Worst case scenario we should be able to run an X server on top of Wayland. Much the same way Cygwin X runs on Windows or X11.app on OS X. Then X clients can talk to the X server over the network, and the local X server can speak directly to the graphics hardware.

Re:Ok great for beginners

[Burz]

And the reason why those two similar scenarios use different software and procedures is that X does not allow for efficient sharing of windows/desktops. Nor does it have NX-like strategies for efficient use over Internet links. Mac and Windows already accommodated these features years ago, while X doesn't give a damn if you want to share an app window while you're in conference... you are out of luck and must resort to bitmap-scraping (VNC) and suffer horrendous performance.

Re:Ok great for beginners

[Unequivocal]

FTW!

Re:Ok great for beginners

[tixxit]

What's stopping someone from running an X server on top of Wayland. Most of those engineering students aren't even running linux, but running an X server on top of Windows or OS X that the clients on their school's server talk to. If you are talking about developers, I'm assuming most software will still use GUI libraries that abstract the actual underlying display tech and they will probably support X and Wayland (and may be Windows and Cocoa), so those will still have no problem using the student's X server. If there are any programs that use sufficiently advanced features of Wayland, then they're probably far too graphical to be running on a thin client anyway.

Re:Ok great for beginners

[gid]

As someone else stated. VNC works just as well even better. If there's a network disruption, you don't want to have to start up the Oracle install from scratch again do you?

I've used Linux for 10 years or more, and tunneling X for me has never been anything more than a novelty. The performance of the program is never going to up to par with running it locally, especially over the internet.

I don't doubt that some people really use X's network transparency. But I'd imagine 99% of the people won't need it, and those that do, won't be running Ubuntu. Ubuntu isn't trying to satisfy everyone, they're just trying to make a solid, easy to use Linux distribution aimed at the majority.

Re:Ok great for beginners

[Anonymous Coward]

No. I just use the command line for everything on a server. Why go through forwarding an X session when that works fine?

Re:Ok great for beginners

[WindBourne]

Wayland will NOT retain X's network transparency. The reason is that it is a local arch ONLY. It was not designed to be networked by itself.
OTH, It DOES run X11 and VNC on top, just like Windows and Mac do. As such, it will still be possible to do all the same things. The difference is that speeds, reliability, driver devl, security, etc will improve.

Re:Ok great for beginners

[electrosoccertux]

X is a part of any proper Unix.

Proper Unix doesn't have any graphical display capabilities at all.

Now get off my lawn.

Just install emacs, that's what I use to play Crysis on my pc.

Re:Ok great for beginners

[Compaqt]

The funny thing is that even Nokia's N900 runs X. So if a mobile device can run X, why can't a desktop or netbook run it?

Re:Ok great for beginners

[swillden]

Yeah, reading some more after my post, I came to that conclusion. Though I suppose it's always possible to slap a network proxy API on top. Not being networkable itself will reduce the value of native Wayland apps, vs X11 apps.

Re:Ok great for beginners

[Abcd1234]

Well, as long as Gtk/QT are built supporting *both* protocols by dynamically switching their backend rendering logic, I couldn't care less. In theory, you'd just get X if it was remote, and direct integration with the Wayland server if it was local (not unlike how, today, you get transport over a socket or mitshm).

Re:Ok great for beginners

[thePowerOfGrayskull]

Wayland is a display server, like X. Why wouldn't it be possible to forward Wayland over SSH?

The RFC for SSH (section 6.3) has specific provisions to allow X11 forwarding which includes authentication and session information; this is implemented separately from tcp port forwarding.

If Wayland uses basic TCP and does not have the same display/auth features, then this isn't an issue - standard port forwarding would work just fine. Further -- reading quickly through the site doesn't say if it's a networked provider at all -- but given the info that *is* present, I suspect not.

Re:Ok great for beginners

[PraiseBob]

So a Desktop Oriented linux won't run a specialized tool?

Isn't that the traditional argument for what makes linux great? You can find a specific distro for your specific needs. Should ubuntu really try to be all things to all people, or should it work on making the desktop experience the absolute best it can be?

Re:Ok great for beginners

[An ominous Cow art]

I got my start in Linux in the early 90s using Yggdrasil. I remember getting it on a CD, shrink-wrapped to a manual the size of a phonebook, which was mostly just various HOW-TOs.

Re:Ok great for beginners

[couchslug]

"I strongly suspect that the people who pooh-pooh the networking capabilities of X never got used to using them."

Ubuntu is intended for mass adoption, not as a professional tool for power users. Ubuntu has been great for Linux adoption, but never lost sight of its original intent.

Anyone needing power user capabilities can run Debian instead, or run a different distro in a VM.

Re:Ok great for beginners

[MikeBabcock]

VNC isn't the same and doesn't work just as well at all.

VNC is not a per-application system in any sense.

When I'm sitting at my desktop running the GUI Xen interface off of one server, the Java HP interface from another server, and a remote database app off a third server all on my screen, I'm using X the way it was designed to be used. If I wanted to do the above with VNC, I'd have a bunch of annoying desktops on my screen to flip through, instead of applications. I'd also have to have vncserver installed on those machines, which is a bloated X server running in memory. No need for any of that with X forwarding.

I have lots of headless servers with no GUI interface that I can run GUI apps on remotely using X forwarding and SSH X tunnelling, and have no intention of installing VNC on any of them for no net benefit, and a horrible UI result.

Re:Ok great for beginners

[shutdown -p now]

Okay, so not a perfect analogy, but you get the point. VNC/RDP can't forward just a remote application, they have to bring the whole desktop.

RDP can perfectly well forward individual windows, which, from user's perspective, is the same thing. That's how RemoteApp and XP Mode desktop integration work in Windows 7.

They can't integrate cut and paste as seamlessly as forwarded X.

Why not? Cut-n-paste works just fine for me in RDP in Windows. Heck, it actually lets me copy a file in a file manager in the guest, and paste it in the host - and it gets copied.

They don't allow your local window manager to choose where things are placed, how they're moved, etc.

That one is true, but is it really a major feature that can justify the flaws that come with X approach to remoting?

Re:Ok great for beginners

[Zo0ok]

Not because you cant run it, but because you dont want to build for it and configure it. Connect a projector to a Linux computer and share the display. I wouldnt do that for my presentation without trying it first (and when I did try recently the computer crashed). Even Microsoft got this right now with Windows 7.

Re:Ok great for beginners

[2.7182]

Ack! I totally agree. I meant to reply to the original statement that stated the opposite.

Re:Ok great for beginners

[mr_bubb]

This has always worked for me, right out of the box, for years, on Linux (Ubuntu).

Re:Ok great for beginners

[Compaqt]

One thing I'd like is to be able to restart the display server (for whatever reason), and have the apps keep running.

What happens for me under Ubuntu (Karmic, Lucid) is X bulks up to 101 MB after a couple of days of usage.

People say it's the pixmaps in apps that cause that, but even if you close out all apps, X still stays huge.

And why does it go from 3 to 6 to 28% CPU, and then back? It goes to 37% when Alt-Tabbing under compiz with a lot of windows open.

After reading some more of the discussion below, it's possible (hopeful?) Wayland might be the answer, but with X compatibility, please.

Re:Ok great for beginners

[Magic5Ball]

X does not magically disappear when other options become available. In the worst case, you could roll your own from source.

Re:Ok great for beginners

[dpilot]

I know who it's intended for, and I agree. I'm more worried about the unintended consequences - like crowding out some of the things that are great about traditional X - like network transparency.

That's a nifty description - transparency. VNC can let you do the work, but it just isn't transparent, at all.

I understand that Wayland is necessarily local, but I strongly suspect that if you were willing to give up performance - and drop back to something more like X - it could do remote, as well. The big thing is the unified access. I know you don't want to run a first-person shooter over the network. Heck, I don't even want to run VLSI CAD tools over the network - but sometimes I have to, and when you need it, you need it.

If the most popular distribution goes local-only, I fear the coming round of popular never-transparent applications running on it. People talk about "too many distributions", which is mostly a red herring, because there is so much in common. But two non-interoperable display technologies is true balkanization - a truly dangerous split.

Re:Ok great for beginners

[somersault]

That's an awfully long way of saying "I fall into the 1% of users you mention that don't want this".

Out of interest, do you even run Ubuntu right now?

Re:Ok great for beginners

[somersault]

Surely it would be pretty easy to write a Wayland module to allow sending of app displays to other machines anyway? Don't render them on the server, and instead send the display buffer to the compositor on the client's machine?

Re:Ok great for beginners

[spitzak]

It looks to me like the "compositing" step does not necessarily mean "composite all the windows onto a certain screen". It could instead composite windows onto several screens, including putting the same window on several. And it could composite onto a remove wayland compositor by some VNC-like protocol, sending the changes to the buffer over the network so the remote one can recreate the window.

Re:Ok great for beginners

[Tetsujin]

Five-digit Slashdot UID megabonus included.

Damnit, damnit, damnit, damnit, damnit, damnit, damnit...

Re:Ok great for beginners

[rockNme2349]

...but I still know a LOT of people who forward X over SSH, and there are still a lot of professors who are advising their students (at least in the engineering schools I have seen) to do the same. I guess this is one of those times that just saying, "I use Linux!" will not convey what people think.

You mean, "I use GNU/Linux!"

Re:Ok great for beginners

[Tetsujin]

Exactly. VNC is usable over low latency links, X is not.

You're confusing me here. A low latency link is exactly what you want for a networked display. By "low latency links" do you actually mean "high latency links"?

Re:Ok great for beginners

[Anonymous Coward]

X is actually the only reason I even *use* Ubuntu. Otherwise, I would just use MobaXterm 100% of the time. Ubuntu as OS isn't really all that 'special', such that it can stand alone when the number one reason to use it (super fast and simple X11 desktop with a modest app selection) is compromised.

They dropped Gimp, and I had to start re-installing it. They went to Empatcy (which is junk) from Pidgin, so it made OCS integration a pain, and I had to re-install Pidgin. Now they are moving away from X11, so my X environments that I spend the day in (DBCA, SAP installers, etc) plus who knows what other tools I use will start requiring more config, overhead, and setup.

Think I'll just leave my Ubuntu images off and stick to MobaXTerm for SSH and X, and Portable Apps for all the rest. Certainly reduces my overhead keeping Evolution up to date with Exchange. DAVmail is nice, but it still wasn't 100%.

Re:Ok great for beginners

[Anonymous Coward]

*I* never got used to using them because they're slow as shit.

Even over a LAN, remote clients are noticeably laggy, and over the internet you can completely forget about it. Even something simple like clicking the 'File' menu involves 5 seconds of lag, then watching the server slowly draw the thing bit by bit before your eyes, and this is between machines on half-decent connections.

Screw network transparency. I'd have more sympathy for the die-hards if it were actually usable in the real world.

Re:Ok great for beginners

[Anonymous Coward]

You-knicks.

Problem solved with creative spelling.

Re:Ok great for beginners

[Cluelessthanzero]

Exactly, hardcore X users will be left out in the cold. Like this:

next_ubuntu_desktop $ su
next_ubuntu_desktop # apt-get install X11
E: You can't do that! Mark has told you not to do that!
E: You're a bad, bad person!

Re:Ok great for beginners

[smallfries]

You could trim your toenails with an axe - but it is not the best way to get the job done. It's slow. It's clunky. Animation and effects aren't synced properly to updates. The battery life is terrible and anything to do the same but with less power would be a blessing. When Wayland comes out on MeeGo I'll try it.

Re:Ok great for beginners

[Anonymous Coward]

X is a part of any proper Unix.

Proper Unix doesn't have any graphical display capabilities at all.

Now get off my lawn.

Or virtual memory, noob. ;)

Re:Ok great for beginners

[dpilot]

You must have a pretty bad connection. My remote X sessions on my LAN are just fine, especially for simple non-graphically-intensive stuff like clicking "File". At work I routinely run Cadence over a network connection, without problems or noticable lags. Even through the VPN over the WAN it's a bit sluggish, but usable. You must have a problem somewhere.

Re:Ok great for beginners

[Burz]

X is a part of any proper Unix.

Proper Unix doesn't have any graphical display capabilities at all.

Now get off my lawn.

You mean like the Altair 8800? That was a bit of UI heaven.

Re:Ok great for beginners

[maestroX]

Nah. Network transparency should have been a userapp as soon as clients were powerful enough to run an os and display server by themselves.
The network requirement really complicated development, stiffled innovation:

• Drag and drop took AGES

• No widgets in memory server

• Security (Hello IRIX, Hello Kerberos)

• ICCCM anyone

• Latency problems on a $$$$$$$$$$$ Sun multicore, win3.1 486sx responds faster,

• Read the source Luke

• Gnome KDE wars

• oops I lost my kerberos ticket again

• GL, DPS extensions (yuck)

Point is, clients will never have insufficient power anymore to justify the need for network transparency (most phones nowadays are faster than computers 15 years ago), response time, speed, features such as video, gl and compositing are far more important than maintaining network transparency.
Besides, X is just terrible for networked display in the current age. Just compare the bandwidth of streaming a DivX with streaming DivX over X.

Re:Ok great for beginners

[Blakey Rat]

VNC/RDP can't forward just a remote application, they have to bring the whole desktop.

Untrue. Don't confuse the limitations of one implementation of one client of the protocol for the protocol itself. (Hell, Microsoft even uses this feature themselves in Windows 7's XP-mode. So not only is this feature in the protocol, you can download a working implementation right now.)

They can't integrate cut and paste as seamlessly as forwarded X.

What's RDP missing?

Question: can X transmit sound like RDP does? That's a critical feature when I want to check my voicemail (which ends up in my email box) remotely.

Re:Ok great for beginners

[Anonymous Coward]

Bicycling isn't really an ideal way to do things any more. Automobiles are much faster than bicycles.

This really depends what you're doing and where you are doing it. In some places, for some purposes, bicycles are faster than cars. And that's a good thing.

Re:Ok great for beginners

[Anonymous Coward]

Non-beginners should be plenty capable of typing:

sudo apt-get install xserver-xorg

It's still Linux and will still run Linux software, regardless of which software it ships with by default.

Re:Ok great for beginners

[DrBoumBoum]

I do that too and I absolutely hate X windows. Have you ever tried to run an Oracle installer on a server from home, i.e., with limited bandwidth and some non-negligible network latency? The reaction times are horrendous! I've actually spend quite a few hours at night waiting for 1 to 2 minutes between successive clicks. God forbid I clicked twice on a button. It would have been literally faster to just drive to the site and do the thing locally.

Man, running an X application over the Internet. I can't think of many worse experiences in computing. Perhaps something with coding in VB on a cmd window could match it.

As somebody said (too lazy to find the reference), it's hard to believe that the X11 authors have not been shot in the street already :-/

Re:Ok great for beginners

[dch24]

Mod parent up! This whole article is a non-issue.

Re:Ok great for beginners

[jwhitener]

There are many server administration tools that can only be accessed by forwarding an X session to your management desktop.
Sun's (now Oracle's) Server 5 apps for instance (ldap, messaging, etc..). The servers themselves have no gui (although they technically could).

Re:Ok great for beginners

[Anonymous Coward]

you use the wrong tool for that - you don't need a browser and X tunneling to change your "routes" at home.

linux has /sbin/route for that.

Re:Ok great for beginners

[Jherico]

Its unclear if you're complaining that you can't launch remote X apps on a Wayland Ubuntu desktop or if you can't launch X apps from a Wayland Ubuntu desktop on a remote machine. However, neither is the case. You can run an X server on top of Wayland (that covers #1) and you can still launch X apps on the Wayland machine as long as the libraries are there (that covers #2 and I do that all the time to run an XTerm on my windows box).

Re:Ok great for beginners

[fast turtle]

From what I see, Wayland is an independent project though I suspect it will become part of the kernel sometime within the next 2 years. Look at what Linus has been pushing for in the kernel. A working framebuffer and no dependency video drivers for the GUI. That's what Wayland is working towards. Total hardware agnostism and I hope to hell it gets there soon.

Re:Ok great for beginners

[Anonymous Coward]

> VNC/RDP and other protocalls are MUCH faster then using X

Not by a long shot.

To get hardware acceleration using VNC/RDP, you must have graphics hardware on the application server (the system running the client), and the client must have access to the hardware. If the application server is hosting multiple users, it probably doesn't have enough graphics cards for all of them. Many servers don't have *any* graphics hardware whatsoever (or if they do, it's limited to displaying BIOS messages during boot), in which case all rendering is performed in software.

With X, the client uses whatever hardware is available in the X terminal. An X terminal has to have some kind of graphics hardware, and nowadays you'd be hard-pressed to find a card which doesn't include some form of acceleration.

VNC/RDP are simpler, and also more flexible (i.e. you can mirror the desktop onto multiple displays, connect and reconnect at will), but they're no substitute for X.

Re:Ok great for beginners

[Anonymous Coward]

I use xpra for that purpose. It's screen for X11 over SSH.

Re:Ok great for beginners

[masterme120]

I've still got the disc and floppy. It was my first Linux.

Re:Ok great for beginners

[jgrahn]

Sun shipped with NeWS. NeXT shipped with Display Postscript. SGI shipped with MEX and later 4Sight. I guess none of these were "proper" Unixes in your godlike eyes

I'm old enough to have used NeWS for a few months in 1990 ... yeah, that was Unix. But every Unix I've used since then (twenty years) has had X and applications like twm, xterm, Emacs ... can you blame people like me for thinking a Unix without these things is a bit ... odd?

Re:Ok great for beginners

[Daengbo]

X is a networked display server. You use a networked audio server to forward sound. I have forwarded apps + sound to thirty clients simultaneously from the same machine before. Not really an issue. (There was a proposal to add networked sound to X a couple of years ago, but it never picked up steam.)

Re:Ok great for beginners

[Anonymous Coward]

For the record, GNU is also not Unix.

Re:Ok great for beginners

[mdmkolbe]

Yeah, but then even when both the local machine and the remote app are Wayland aware, you still fall back to X.

I can accept needing X for compatibility with old apps, but if Wayland doesn't get some network transparency then it will never be a replacement for X.

Re:Ok great for beginners

[mdmkolbe]

It matters not whether it is "possible" to bolt networking onto Wayland. It matters whether networking "is" or "will be soon" in Wayland.

Imagine a future distribution where Wayland has replaced X. I won't care about whether networking "could" part of Wayland. I'll care about whether "ssh -Wayland" tunnels properly and efficiently.

The problem is that SSH can't do this on its own. Sure it could ship entire buffers from one side to the other but that is frightfully slow. In order to efficiently use the network, SSH has to be tunnelling higher level drawing command like it does with X currently. But that requires the apps to be rendering in terms of a common higher level drawing API. This is anathema to the Wayland philosophy as they want to be rendering agnostic.

Wayland "could" get great networking support. But the attitudes ... er ... design goals expressed by the developers tell me that isn't likely.

Re:Ok great for beginners

[mdmkolbe]

VNC/RDP and other protocalls are MUCH faster then using X.

Prove it. Programs with vector based rendering will always be faster sending the vectors (as X does) rather than the bitmaps (as VNC does abet compressed). On the other hand a program with lots of bitmaps should send those bitmaps only once when over X, but many times over VNC.

Re:Ok great for beginners

[Anonymous Coward]

> forwards the api calls into some kind of rpc calls (or work on

No.

X11 protocol sends drawing requests to the display (line x1/y1 to x2/y2 ...). Wayland requires the client to allocate and paint to a local buffer. See

http://wayland.freedesktop.org/architecture.html

You can send (a probably delta-compressed) image of the buffer to the compositing manager. But for simple apps which do not require compositing, this is a step back. It will slow down the UI, at least on slow network connections.

Re:Ok great for beginners

[Anonymous Coward]

Well, yes.

To piss you off and repeat what the guy a couple of posts above me said, how about OSX? A descendant of NeXTStep, a certified Unix (on an Intel machine, at least), which you can use today by going and finding someone with a Mac. Sure, it includes X11 in case you use some X11 apps - which I do, I run xmgrace and GIMP daily - but Apple's X11.app is totally optional. You can trash it and your genuine certified Unix will happily go on running just the same, using its Quartz etc. windows manager. (Also I use Aquamacs, which is basically just a recompiled GNU Emacs with graphics calls replaced with calls to Apple's APIs. And I don't see why anyone would rave about xterm over, well, any other terminal emulator out there. They all do the same. If it can go to any directory on the disk and run any program, what do I care?)

Re:Ok great for beginners

[badkarmadayaccount]

And we can name it... hmmm... X11?

Re:Ok great for beginners

[badkarmadayaccount]

X client side compressing proxy server.

Re:Ok great for beginners

[jbolden]

Yes one of the 3 things will happen:

1) Some apps will be graphical and Wayland only while more generic Unixy apps will be X11.
2) Wayland dies out
3) Wayland essentially becomes X12.

Re:Ok great for beginners

[somersault]

Sure, if you want to have hack upon hack to get all the cool stuff that Wayland can do natively. At some point it's better to start from a nice consistent base. Sending a UI over the network is the one feature that people are bothered about losing in X, and it's pretty easy to replicate on any other windowing system - especially compared to all the other crap people have to do on top of X to try to get more friendly APIs and UI features.

Re:Ok great for beginners

[makomk]

Even the drm api can be implemented as a virtual graphics device with a network backend,

Which would require kernel support that's unlikely to be implemented, and would have a snowball's chance in hell of getting into the mainline kernel even if it was. Part of the reason OSS was killed off in favour of ALSA was to avoid this kind of nastiness inside the kernel.

but that's probably not even necessary because from the architecture pictures I don't see a reason why the compositor wouldn't be able to support a network environment.

Could be done. It'd involve essentially reimplementing XVNC at the client level, though - you'd need some way of transmitting changes to the images being composited, plus some way of doing rendering in software - and wouldn't be terriby efficient. In practice, if Wayland applications ended up actually taking advantage of the performance advantages it offers, they'd be completely unusable with this setup.

Anything better would require a way of performing OpenGL operations over the network - i.e. some equivalent of X's indirect rendering support, complete with all the supporting functionality and all the code in Mesa. Not exactly easy.

Of course, the real problem is who's going to implement all this. The Wayland developers aren't - the entire reason they designed it this way is because they wanted a simple compositor without this kind of cruft. Ubuntu has never done this kind of fundamental low-level development before, and I can't see them starting now.

(Oh, and if the network connection fell over, you wouldn't be able to close, move or minimise the remote applications. Client-side window decorations for the win!)

Re:Ok great for beginners

[MikeBabcock]

Actually it was an attempt at being informative and refuting the oft-cited argument that using VNC is equivalent to remote X connections. The fact that many people don't know how remote X connections could benefit them was the majority of the rationale behind being so verbose.

To answer your redundant secondary question, yes, on laptops and at others' homes. I find in general that Ubuntu live discs are the fastest way to demonstrate Linux to others, and the easiest to get up and running with the subset of software I need on the road.

Re:Ok great for beginners

[somersault]

It would only be a redundant question if it was obvious that you use Ubuntu.

Even if the whole of the Linux world moves to Wayland as their primary display system in the next decade or so, it will still be very possible to use X for running remote apps. So I really think that people have no need for complaining here, at all.

Re:Ok great for beginners

[Anonymous Coward]

I can forward X to my Mac's X11 implementation (XQuartz) and apps don't care that this is a layer on top of Quartz/whatever Apple use. There's no reason Wayland won't be able to do the same.

Re:Ok great for beginners

[jelle]

If you are not seeing the solution, then that does not mean there isn't any.

How many apps on the Ubuntu desktop today do you think still use things that benefit from things such as server side font rendering. And which of those will actually switch from the X11 protocol to wayland, when wayland has backwards compatability for x11 client connections?

For example, the old apps will still be using the X11 client libraries with no difference in performance from today because it can be forwarded in exactly the same way. Gnome apps use gtk and cairo, 3d apps will use opengl, and all of that can be intercepted by a networking app, and in most cases it will actually be much more efficient to do that than intercepting at the x11 tcp protocol level.

A bit big for their britches?

[jhigh]

I'm wondering if the Ubuntu crowd isn't letting their success go to their heads just a tad. Just because they're the most popular distribution doesn't mean that they can start changing everything around and have everyone else follow their lead. It's one thing to make some incremental changes that you think are best for the distribution or for Linux as an operating system. But to be making statements like this:

We'll help GNOME and KDE with the transition, there's no reason for them not to be there on day one either.

says to me that Ubuntu wants to make substantive changes to the free desktop environment and have everyone follow their lead. As a long-time Ubuntu user, I wish them well. But with the attitude with which they seem to be approaching things, I suspect that we will start to see Ubuntu's share of the desktop start to decline in future years as some other distribution steps up to the plate.

Re:A bit big for their britches?

[commodore64_love]

>>>doesn't mean everyone else follow their lead

Precisely. I gave Gnome, KDE, and other variants of Ubuntu a try in 2008 and again in 2010, but I think I'm done now. I hate change when the change happens for no good reason (look at my name, or my TV with a VCR still attached to it). Anybody want to buy a shiny Lubuntu 10.0 CD? Hmmm. Maybe I'll put it on ebay.....

Re:A bit big for their britches?

[jgagnon]

Or maybe they were referring to the transition of the Ubuntu distributions that include GNOME and KDE? In other words, changes will need to be made so that all of the technologies could still be used and useful and they will help the GNOME and KDE teams so that things aren't a broken mess on Ubuntu, Kubuntu, etc.

Re:A bit big for their britches?

[somersault]

I remember a discussion a year or two ago here on Slashdot how X was badly in need of replacing. Sounds to me like Canonical have the right idea, and the impetus to make it happen.

Re:A bit big for their britches?

[Hozza]

I'm not sure about Gnome, but KDE already works on non-X11 platforms (e.g. KDE 4.4 is available for Mac), and there's already an ongoing port of Qt to Wayland, so it shouldn't require a big effort from the KDE developers to support this move.

Re:A bit big for their britches?

[Enderandrew]

Ubuntu has repeatedly tried to be as bleeding edge as possible, which is why I don't recommend it to people who want a stable system that just works, and why I never recommend it in the enterprise.

Why are people shocked that Ubuntu is trying to push for change here?

My only concern is that last time I looked Wayland wasn't ready for primetime, and the intent with Wayland wasn't to be a full replacement for X for most users.

Re:A bit big for their britches?

[Anonymous Coward]

Was about to say the exact same thing. Everyone knows X is shit and needs to die, and the primary thing stopping that is momentum. And now someone actually tries to create that momentum and is immediately attacked.

Re:A bit big for their britches?

[Anonymous Coward]

So 'improvement' is not a good reason for change? You honestly think that changing from VCRs to DVDs wasn't for a good reason?

I would really hate to see what your ideal world would look like.

Re:A bit big for their britches?

[cronco]

Now they haven't. Fedora is the "bleeding-edge" distro, which regularly ships with beta software bundled. The only time I remember Ubuntu doing the same was with FF 3 beta, which was included in 8.10 if I'm not mistaken.

Re:A bit big for their britches?

[mcrbids]

But you are missing the point. X is a protocol layed out long, long ago. And there are long standing issues with it that become glaring as we move forward. Long known issues such as 3D support.

Are the advantages of the new server enough to outweigh the costs?

Besides, when did Slashdot become a crowd that believes that there should only be one right way? Forks are GOOD for software evolution - it's how new ideas get tried out!

Go Ubuntu for being brave enough to try to tackle the problems of X!

Re:A bit big for their britches?

[jgagnon]

It's not so much that X is garbage, it just isn't nearly as useful on a thicker client with lots of resources.

Re:A bit big for their britches?

[jhigh]

I'm not attacking them. It's just that coming on the heels of their announcement to move away from Gnome, it seems like Ubuntu is using their large install base to try to steer the free desktop market in their direction. I'm not even necessarily commenting on whether or not this is good or bad...just that I'm not certain that they're doing it with the right attitude to be successful.

I've been an Ubuntu user for years and have a great deal of respect for what they've done for the Linux desktop. I just don't want to see them overreach and fail.

Re:A bit big for their britches?

[DoofusOfDeath]

I'm wondering if the Ubuntu crowd isn't letting their success go to their heads just a tad. Just because they're the most popular distribution doesn't mean that they can start changing everything around and have everyone else follow their lead.

If you're right, it's creepy how much it parallels Obama's and the Democratic party's ambitions over the past two years.

Re:A bit big for their britches?

[MBGMorden]

Actually, I respect this a lot. X11 is an old, OLD and outdated protocol that people have been trying to move past for years. Until now though, people either lacked the resources or the balls to move forward and actually do it though.

Ubuntu has been consistently making great strides because they WILL do stuff like this. I don't necessarily always agree with them (for example, taskbar buttons on the right), but I admire their dedication.

In a lot of ways I see Shuttleworth as a mirror version of Steve Jobs. They both seem to be willing to throw out any ideas on conventional wisdom and what a system "has" to have or do, and do things their own way. Shuttleworth just seems to be using his powers for good. :)

Will I like this? Not sure. Maybe, maybe not. I think though that if they can really get the community to follow them down this road, we'll all benefit.

Re:A bit big for their britches?

[I_R_Che]

But isn't this a good thing? Whether they succeed and advance the state of Linux on the desktop or they fail and other distributions, focusing on other things, retake their current lead because of those 'other things', the Linux desktop will benefit.

Re:A bit big for their britches?

[tjwhaynes]

My only concern is that last time I looked Wayland wasn't ready for primetime, and the intent with Wayland wasn't to be a full replacement for X for most users.

If Mark Shuttleworth was proposing Wayland for prime-time inclusion in Ubuntu 11.04 or even 11.10, I'd be concerned. But if you actually follow this news story to the original source at http://www.markshuttleworth.com/archives/551 you would find this:

Timeframes are difficult. I’m sure we could deliver *something* in six months, but I think a year is more realistic for the first images that will be widely useful in our community. I’d love to be proven conservative on that :-) but I suspect it’s more likely to err the other way. It might take four or more years to really move the ecosystem. Progress on Wayland itself is sufficient for me to be confident that no other initiative could outrun it, especially if we deliver things like Unity and uTouch with it. And also if we make an early public statement in support of the project. Which this is!

So the first likely viewing of this would 11.10 and real integration into the entire stack is more likely in the 14.10/15.04 time frame.

So this is a classic storm in a teacup right now. The reality is "promising project will be supported by major Linux player for future inclusion".

Cheers,
Toby Haynes

Re:A bit big for their britches?

[computational super]

I'm a little shocked at all the negativity. Have you people used X? If Ubuntu can drive a replacement, let them drive a replacement!

Re:A bit big for their britches?

[jhigh]

As I pointed out in a separate comment, I'm not criticizing Ubuntu for what they're doing. Rather, I'm being critical of their approach and frankly for the attitude that they appear to have about it. I'm an avid Ubuntu user and have been for years. However, in terms of Linux development, they're one of the relatively new kids on the block and have achieved success largely on the back of Fedora. To now be trying to steer the entire free desktop market in their direction as they make major changes seems to me a bit presumptive.

Re:A bit big for their britches?

[commodore64_love]

>>>So 'improvement' is not a good reason for change?

Of course it is. But this is not an improvement. This new program is a change simply for change's sake, and is no better than the old program.

As for VCR to DVD, I still consider the VCR better because it's recordable, and the Super VHS can hold upto 12 hours of dvd-quality video for capturing television shows. The DVD-R cannot. The DVD-R also has a bad habit of self-erasing (the dye fades) where a tape can last for decades.
.

>>>would really hate to see what your ideal world would look like.

It would look a lot like this one, but thing's would not be changed simply to change them. For example Microsoft Office would still have its original dropdown menu system, rather than that damnable ribbon shit.

Re:A bit big for their britches?

[Boomerang Fish]

The problem as I see it is not so much how good X is or isn't (and I've seen cases where it's both), but what happens to all of the existing software that is in use, largely in Academia, but also to a significant degree in engineering and other industries using decades old software...

If X compatibility is actually any good, then fine, let change come... OS X was worlds better than the Classic Mac OS, but they did maintain the ability to run a classic emulator (at least until the Intel machines started coming out)... Change can be a fine and necessary thing, but expecting everyone to rebuild and recode before we see how it plays out is just hubris.

--
I drank what?

Re:A bit big for their britches?

[Anonymous Coward]

Ubuntu is the new Microsoft?

Re:A bit big for their britches?

[Enderandrew]

Ubuntu was the first to push PulseAudio, KDE 4, Grub2, and things like that. In each case they were criticized for shipping something that wasn't ready for primetime, but trying to be the first ones to push for the change.

Re:A bit big for their britches?

[0100010001010011]

Debian Sid is "the" bleeding edge. (Or if you're sadistic, experimental). Ubuntu is out of date the day it is released because of the way they insist on 'locking' to a version of software (minus bug fixes).

Ubuntu 10.04 will never get anything other than 2.6.32. (Unless you do extra ppas, etc). All software, alsa, nvidia drivers, everything are locked to what ever shipped with 10.04.

Re:A bit big for their britches?

[Dcnjoe60]

Wouldn't you want one of the big players to be the one putting time and effort into alternatives to old technologies? Most people I know use X, but none of them would say that it is great or state of the art. Who else but Ubuntu, Redhat or Suse, all comemrcial distros would be in a position to look at moving beyond X? While I don't know anything about Wayland, so I can't comment on whether it is a good or bad decision, it seems that the real work falls into two categories: 1) get Wayland up to speed (if it's not already) and 2) make gnome and kde less dependent on the underlying display driver.

It seems that neither of these are bad things and #2 would actually be good. Besides, Ubuntu could always switch back if it didn't work or there were a surge of development for X.

Re:A bit big for their britches?

[Teun]

The inclusion of KDE4.0 was another experiment.

And it backfired on the Kubuntu team.

Re:A bit big for their britches?

[Lennie]

I think the problem with DVD is the business model, not so much the technology.

Re:A bit big for their britches?

[DragonWriter]

I'm wondering if the Ubuntu crowd isn't letting their success go to their heads just a tad. Just because they're the most popular distribution doesn't mean that they can start changing everything around and have everyone else follow their lead.

Are you objecting because Canonical has a vision for the future direction of the desktop, or are you objecting that they are expressing willingness to work with GNOME and KDE to work toward that vision while minimizing fragmentation?

Re:A bit big for their britches?

[Zo0ok]

Exactly... writing a simple application for X was arcane when I did it almost 10 years ago. Nobody really writes applications for X anymore - they write them for GTK, KDE.

Re:A bit big for their britches?

[Lennie]

Only difference is, Steve Jobs can command everyone in his company to work on it. While Shuttleworth needs the community to help him.

Re:A bit big for their britches?

[Dog-Cow]

They aren't trying to steer anyone. You are reading way too much into it. And if I happen to be wrong, so what? If they have a good idea, why not run with it? What the FUCK does their newness in the Linux world matter? They can't force anyone else to play along, so where's the harm in them trying?

You seem like an arrogant prick who hates change just because.

Re:A bit big for their britches?

[jhigh]

They aren't trying to steer anyone. You are reading way too much into it. And if I happen to be wrong, so what? If they have a good idea, why not run with it? What the FUCK does their newness in the Linux world matter? They can't force anyone else to play along, so where's the harm in them trying?

You seem like an arrogant prick who hates change just because.

Actually, I just finished emailing someone telling them I can't wait to try the next version of Ubuntu because I want to see all of the new changes in action. What I'm concerned about is that Ubuntu is letting their popularity get the best of them. What I don't want to see is them make a bunch of changes that results in people dumping them for some other distribution that is more familiar. My fear is that Ubuntu thinks that because they have such a large share of the Linux desktop market, they can make whatever changes they want and their user base will accept it.

Re:A bit big for their britches?

[DragonWriter]

As I pointed out in a separate comment, I'm not criticizing Ubuntu for what they're doing. Rather, I'm being critical of their approach

The only thing you've pointed to about their approach is their statement that they are willing to work with KDE and GNOME. What is the problem with that?

and frankly for the attitude that they appear to have about it.

What attitude?

To now be trying to steer the entire free desktop market in their direction as they make major changes seems to me a bit presumptive.

This kind of thing -- sometimes successful, sometimes not -- is how progress happens. It doesn't come from people who think they have to be the old man of the industry before they can try to promote change.

Re:A bit big for their britches?

[Bill_the_Engineer]

Being OLD is never a good excuse to replace it. There may be other issues that need addressing but OLD is never it. If anything, OLD is a good reason to keep it. It's "survival of the fittest" and X11 apparently outlasted many attempts to replace it.

Lusting after shiny objects and doing something solely because its new or trendy has always been an issue in the software development industry. Sometimes it's a good thing since it introduces new ways of doing things, and more frequently it's not.

Now if you said we need a more efficient method of handling large globs of data between the application and the local display device that would be a good reason, but OLD is never one. BTW, obsolescence due to lack of language support, it's written in a dead programming language or lack of use is not the same as OLD. Sorry about the rant about your using OLD as a reason. When I see OLD I see NIH.

Incidentally I think the timing for Wayland is a little off. Sure there are multimedia applications that will benefit from closer ties to the hardware, but the industry as a whole is trending once again to thin clients. If this continues to be the case, I see Wayland trying to reproduce what X11 already does and for a very long time at that.

Then again I find it somewhat odd that we take a very powerful OS platform and begin to remove its power in order to reduce its utility enough to make it more palatable for the single user desktop use case.

Re:A bit big for their britches?

[Again]

Mod parent up.

Re:A bit big for their britches?

[Per Wigren]

I guess you haven't actually tried KDE for Mac? I wouldn't say "works", more like "good enough for making some carefully crafted screenshots giving the illusion that it works but in reality it's an unusable mess".

I'm pretty sure that KDE will work great on Wayland though, because they will actually have some developers focusing on it. :)

Re:A bit big for their britches?

[Again]

I'm wondering if the Ubuntu crowd isn't letting their success go to their heads just a tad. Just because they're the most popular distribution doesn't mean that they can start changing everything around and have everyone else follow their lead. It's one thing to make some incremental changes that you think are best for the distribution or for Linux as an operating system. [...]

Being the most popular distribution has nothing to do with it. Of course they can do what they want. We've known for a long time now that Canonical isn't a democratic company.

I don't think this should be as shocking to people as it seems to be. Meego is moving to Wayland. We've had numerous articles on how the X.org spec is outdated. And don't forget how Canonical couldn't get their multitouch code in where they wanted it.

Re:A bit big for their britches?

[Anonymous Coward]

And if you're left, it's disturbing how insane the right acts, to the point where they demonize everything you do, even when it's stuff they proposed in the first place.

Ok, so I don't know that there's simply a lot of Oba^H^H^HUbuntu hate driving a frenetic opposition to anything because it is Ubuntu's, but that is my perspective on the political scene, so that's why your analogy doesn't fit quite as well as you think.

Sometimes other people do view things entirely differently from you, and sometimes there's a bit of truth in both sides, and sometimes not.

But really, if you think Ubuntu is behaving arrogantly, or summarily, and you claim the same about Obama, rather than the Republican ideologues I'd say we are seeing completely different realities.

Not that I think the people behind X, Gnome, or KDE are acting like the Right-wing ranters or the Tea Party, not at all. As far as I know they've been relatively sane.

Re:A bit big for their britches?

[geminidomino]

No. They fancy themselves the new Apple.

Re:A bit big for their britches?

[Anonymous Coward]

To be fair, "old" (and outdated) is probably nothing more than a way to describe a problem conversationally without having to get into an argument about what's wrong with X11, so really, do watch your ranting. But anyway, there's nothing odd about reducing excess things people don't really use. Power, in most cases, does have a cost, and simplicity does have a virtue.

Why grab a cordless drill when you only need to turn one screw? Sometimes it's best to just get the screwdriver.

Ok, so a cordless drill and a screwdriver aren't exactly parallel to the case of an OS platform, I'm just trying to explain the value in a limited tool with what I hope is a relatively comprehensible analogy.

Re:A bit big for their britches?

[Anonymous Coward]

>>>So 'improvement' is not a good reason for change?

Of course it is. But this is not an improvement. This new program is a change simply for change's sake, and is no better than the old program.

As for VCR to DVD, I still consider the VCR better because it's recordable, and the Super VHS can hold upto 12 hours of dvd-quality video for capturing television shows. The DVD-R cannot. The DVD-R also has a bad habit of self-erasing (the dye fades) where a tape can last for decades.

Unless you play it, then it has a tendency to wear down and break rather quickly. But I think your confusion is based on a fundamental misunderstanding. DVD was not meant as a replacement for recording, it was meant as a replacement for playing, a function which it serves far better than VHS tapes.

If you want to discuss what replaced VHS recording, then we'll go into things besides recordable DVD's.

Sometimes an improvement is a matter of what you consider the function, not the technology itself.

This is why I don't drive a muscle car. A person who drives a car for "performance" and the ability to drag race probably wouldn't like my little street wagon, but me, I'm happy with it.

If you want to replace your VCR, try looking at HDD recording, with DVD's as player for when appropriate.

In the case of Ubuntu and Wayland, well, I don't know enough about Wayland to say what if offers over X, but I'm sure somebody can offer a robust explanation.

Re:A bit big for their britches?

[JackieBrown]

He also thought that gnome could/should switch to qt4

http://tech.slashdot.org/article.pl?sid=08/07/14/1245204

Re:A bit big for their britches?

[walshy007]

Actually, I respect this a lot. X11 is an old, OLD and outdated protocol that people have been trying to move past for years.

Change for changes sake is not necessarily a good thing. If I were to say "hey lets all switch from a linux kernel to plan9" doesn't necessarily make it smart.

Ubuntu has been consistently making great strides because they WILL do stuff like this.

Aside from including closed source codecs and drivers, upstart seems to be the only thing contributed. Considering the contributions of redhat, ibm etc to the community canonical/ubuntu is just a drop in the bucket in terms of innovation.

What they have succeeded in well is marketing though, I'm still not quite sure how this cult of ubuntu came about, don't get me wrong there's nothing wrong with another distro, but people seem to think they do more than they actually do.

Re:A bit big for their britches?

[swillden]

Everyone knows X is shit and needs to die

I don't know that, and I think most of the people who do "know" that don't really have a clue what they're talking about.

X isn't perfect, but it's darned good. The bottom-level display protocol is weird and quirky, but it's proven to be extremely flexible and extensible. Modern X servers have excellent performance while retaining all of the network transparency and flexibility from the original design.

That said, I think Wayland may actually be a worthy successor. It retains the basic client/server architecture which is really important, but replaces the quirky X protocol with a new (and, hopefully, cleaner) protocol. It provides a way to retain X compatibility by making the X server a Wayland client. It brings the compositor, which is a somewhat-klunky add-on to X into the core of the display server where it belongs, given that compositing is clearly the best way to combine elements into a whole display.

As long as Wayland's protocol is sufficiently extensible to allow future improvements in the way we do graphical environments, the way X's protocol has proven to be, I think it might be an excellent idea. That, really, is my biggest concern. The X protocol has proven to be extremely extensible and I worry that a new technology might not have that same flexibility. We don't know what the future will bring and we don't want to be painted into a corner where massive rework is required to adopt a new approach. Of course, that could happen with X as well -- but so far the X protocol has proven to be remarkably adaptable.

Even that doesn't really worry me, though. I expect that the Wayland team has absorbed the lessons of X's excellence and is endeavoring to retain them in the new design.

Re:A bit big for their britches?

[MostAwesomeDude]

We've advocated not directly programming in Xlib for years now, alongside admonishments to authors of new WMs and toolkits. There are mature, well-tested, usable toolkits out there already, and we encourage people to use them.

Re:A bit big for their britches?

[Bill_the_Engineer]

. But anyway, there's nothing odd about reducing excess things people don't really use. Power, in most cases, does have a cost, and simplicity does have a virtue.

Very true. However, there were other alternative OS offerings that were built to fit the multimedia based single user desktop market like BeOS or now its OSS imitator Haiku.

Re:A bit big for their britches?

[MostAwesomeDude]

What? X's lack of builtin 3D support in the protocol isn't a bad thing. X's core protocol is dead simple because it was designed to be extended. High-quality font rendering isn't in the core either; is that something to be outraged about as well? Imagine for a moment that X *did* have 3D support in the core. It would be PHIGS all the way (http://en.wikipedia.org/wiki/PHIGS) and you'd be complaining about that instead.

Wayland, compared to X, has *no* core rendering protocol. There's no 3D support, and also no 2D support.

This isn't a fork. Wayland shares no code with X, although it uses the same external xkb library.

Re:A bit big for their britches?

[MostAwesomeDude]

The X.Org team has been planning X12 for years. http://www.x.org/wiki/Development/X12 is the roadmap. We will probably start X12 development as soon as all X11 bugs are fixed. (Haha, only serious.)

Also, what do you mean by "resources or balls?" Plenty of challengers have shown up over the years. DirectFB, Fresco, Berlin, Y Windows, etc. None of them displaced X because *X is a hell of a lot better than you give it credit for*. Wayland's developer realizes this; he's not trying to replace X, but to work alongside it.

Re:A bit big for their britches?

[icebraining]

but the industry as a whole is trending once again to thin clients.

Thin clients, as in with networked display? Where? Not in consumer land, that's for sure, which is were Ubuntu lives.

From what I can see, the move it to the Web, with the interface running locally (either using a browser or a light client like an Android app) and the backend on the 'cloud'.

Networked displays were never a good solution for normal users, in my opinion, especially through the Internet. Too much bandwidth even if compressed, too much lag between user input and app response. I think the application specific client-server solutions with local caching and much better fault tolerance will be the future.

Re:A bit big for their britches?

[Anonymous Coward]

Nope, not too big.
Ubuntu just decided to take a certain course for the future, keeping in mind the rest of the world might not follow suit, and saying that that's ok. ubuntu will make sure the products of the people that have not followed suit will stil be useable. I can find no fault in these statements, no arrogance, no "follow us or die". Just plain "this is how we're gonna do it, don't feel forced to follow us, it will be ok for you if you don't". Sounds perfectly reasonable to me.
In the end, I, the archetypical desktop user, have only one interest: that I can turn on my computer and that it then runs the software I tell it to run without any fuss. Everything else is politics, and as we all know, politics are very smelly.

Re:A bit big for their britches?

[Slime-dogg]

they write them for GTK+, QT.

FTFY

Re:A bit big for their britches?

[t2t10]

Have you people used X?

Yup. It's probably the best window system you can get right now. It is a client/server system with an optimized and mature protocol, plus some really good features.

Windows started out as a direct graphics library and then was retrofitted into a client/server model--poorly and inefficiently.

OS X started out with NeXT's client/server DisplayPostscript. It had really bad performance and Apple spent a lot of time trying to fix up its inefficiencies. But they still only get good performance because they use tons of cache.

We'd all be better off if Apple and Microsoft also switched to X11, instead of spreading their marketing b.s.

Re:A bit big for their britches?

[t2t10]

they write them for GTK, KDE

That, and dozens of other toolkits, all of which will break if you don't support the basic X11 protocols and support them well.

Re:A bit big for their britches?

[Slime-dogg]

Being OLD is never a good excuse to replace it. There may be other issues that need addressing but OLD is never it. If anything, OLD is a good reason to keep it. It's "survival of the fittest" and X11 apparently outlasted many attempts to replace it.

How about OLD, BLOATED, and CRAPPY, as well as built upon a foundation that is no longer necessary or desired?

When you install a standard distro as a desktop OS, you get a full blown X server. It's a massive beast, and causes a bit of a bottleneck. There are also design choices that were made which ended up being poor choices, though the reasoning for this was that it was a very new thing to everyone. Now, after many many years of experience and study, we finally have the opportunity to discard the old and outdated system in favor of a modern one.

You talk about power in the OS, but then also talk about X11 as doing what we need. X11 does do what we need, but it also provides more than what we need, to the detriment of what we need. Let it be the client that sits on top of a system that really is what we need. That way, we can actually realize more of the power of the OS platform we run on. mmkay?

Re:A bit big for their britches?

[noidentity]

Being OLD is never a good excuse to replace it. There may be other issues that need addressing but OLD is never it. If anything, OLD is a good reason to keep it. It's "survival of the fittest" and X11 apparently outlasted many attempts to replace it.

And by your logic, if it gets "replaced" with Wayland, then Wayland was obviously better. I mean, how do you think survival of the fittest works, if it doesn't involve competing entities?

Re:A bit big for their britches?

[Bill_the_Engineer]

Also by my logic being OLD was never a factor.

Re:A bit big for their britches?

[Bill_the_Engineer]

How about OLD, BLOATED, and CRAPPY, as well as built upon a foundation that is no longer necessary or desired?

Evidently you have a short attention span because you didn't bother to read the whole comment. Being BLOATED and CRAPPY are good reasons, but being OLD isn't.

Re:A bit big for their britches?

[Bill_the_Engineer]

Thin clients, as in with networked display? Where? Not in consumer land, that's for sure, which is were Ubuntu lives.

Wait for it....

From what I can see, the move it to the Web, with the interface running locally (either using a browser or a light client like an Android app) and the backend on the 'cloud'.

There it is. The protocols are different but it's still thin client. Incidently, X11 seems to handle firefox very well.

There's still demand for remote desktop. Just ask GoToMyPC of course nothing prevents us from using VNC, RDP, etc. Though remote rendering is better than pushing bits from a server frame buffer. Not to mention having remote desktops on netbooks, laptops, or other lower powered consumer devices is actually quite nice.

Networked displays were never a good solution for normal users, in my opinion, especially through the Internet. Too much bandwidth even if compressed, too much lag between user input and app response. I think the application specific client-server solutions with local caching and much better fault tolerance will be the future.

Funny. I'm using it now. It runs as if I'm at the desktop. My server is currently 360 miles from where I am at this very moment. No lag, the mouse works great, running full screen eclipse and everything. Including *shudder* Flash apps when I need to browse using Firefox.

May I suggest FreeNX for the server and OpenNX for the client? Nomachine's NX protocol does a great job at moving X long distances. Oh did I mention I have to go through a VPN gateway too?

Re:A bit big for their britches?

[MBGMorden]

Much less true now than it once was. When I used Linux ~10-12 years ago THEN it really was "dozens of toolkits" which an app might be written in. Nowadays? With the exception of JBidWatcher (which is a Java app using SWING IRRC), and Chrome, I don't use anything that's not GTK+. I don't even use any QT apps, much less any of the more obscure toolkits.

Virtually any major app? QT or GTK+. Get those two natively ported, provide a rootless X11 server that runs ATOP the new system for anything else, and you're in business easily.

Re:A bit big for their britches?

[MBGMorden]

Except that they're not really moving away from *Gnome*. They're no implementing Gnome-shell. You're talking about a tiny bit of the interface that Ubuntu will be using their own version of, not the whole DE. And in that regard, it's GNOME who are goofing this up. Gnome-shell is a radical shift from any current UI, and many, many users (including myself) have been stating their dissatisfaction with that interface since it was announced.

Instead of going that route, they're transitioning to a default UI that is based on an icon-dock - like BOTH the other mainstream desktop OS's use now. Instead of keeping the old X11 stuff, they're looking to transition to a technology that is tied much closer to the hardware and can provide much better performance for all the apps running locally . . . just like OS X and Windows do.

Ubuntu isn't being radical here. They're making a Linux DESKTOP system, and if they have to drag some people along kicking and screaming then so be it. There's a reason why "the year of Linux on the desktop" is always like the fruit trees out of Tantalus's reach - people keep dragging their feet and not making the tough calls. Why should everyone start whining when Ubuntu starts taking the steps needed to possibly make that pipe-dream a reality?

Re:A bit big for their britches?

[djdanlib]

Steampunk, obviously

Re:A bit big for their britches?

[MBGMorden]

Mac OS X is a perfect example of the possible situation here. Apple (wisely) chose not to use X11 is their display technology with OS X. And say what you will about their business practices (I admittedly can't stand them), but the GUI on a Mac feels smooth as butter.

What they DID do however is make available a rootless X11 server that runs ON TOP of their display technology. Any X11 app that launches will run just fine. Windows has them available too. Google Xming for a good free one - I used Hummingbird eXceed and later Xwin32 on my Windows machine back in college to bring up Solaris X11 apps on my system. It worked fine.

The same thing can happen EASILY on a new dislpay technology, and was even stated in the article. Setup your new Weyland display system. Port GTK and QT to it and most apps are native at that point. Then port xorg over to act as a server to X11 apps, but a client to the Weyland display (exactly what xming does with Windows - it's an xorg port). At that point you get compatibility with any existing X11 apps until they get ported over. They won't get the benefit of the new display technology, but they'll still work.

If you do the rollout properly it can be pretty painless, and the gain you get in the end can be HUGE.

Re:A bit big for their britches?

[Compaqt]

I think you're right. There's attitude of "whatever Mark says goes" at Ubuntu. It would be nice if there were (at least some level of) discussion with other projects to get their concerns before making a headlong rush.

Re:A bit big for their britches?

[shutdown -p now]

Porting Qt to Wayland is already in progress. I believe there was something for Gtk as well.

Re:A bit big for their britches?

[MikeBabcock]

Three-phase 60Hz A/C power is an old standard (in North America). We should replace it.

Sure, it'll break everything that uses electricity, but those things can all be replaced.

Re:A bit big for their britches?

[Anonymous Coward]

No they don't. GTK and Qt don't have to use Xlib as their backend. How else would I be able to run gtk and qt apps on Windows where X isn't available?

Re:A bit big for their britches?

[HermMunster]

I use Ubuntu and have for the past few years. Before that I tried various other distributions of Linux. I have no good will toward the RPM based installs. That said, I know that Fedora has about 24 million desktop installs in use on a day to day basis, whereas Ubuntu only has about 12 million. Ubuntu does seem to get the most press because it is very easy to install and there are a number of nice things about it that make working with it and maintaining it a pleasure. There's lots of support and Canonical contributes (though indirectly) to the projects and provides services that help development.

Someone's operating with a heavy hand at Canonical because I doubt that educated minds would conclude that these changes are good (especially since they are both new technologies that are extremely immature and in some ways many steps backwards).

If you want to draw a lot of negative attention to this, keep going in that direction. Most people I know using it will either not upgrade or they will change distributions.

Re:A bit big for their britches?

[ADRA]

Gnome apps run on windows, so they've in one way or another accomplished the same thing.

The question really lies in if they believe that this new server is in fact a step in the right direction. If they do, then it is feasible that it'll get some developer love, but if they're making a bad step then I wouldn't expect support from others. I would say that XFree86 was definitely being held up by the group themselves, but since the X.org fork there has been some good changes to the system that have moved it along. I'd say the worst part of my desktop Linux experience has been related to driver support, and frankly you can run any number of graphics servers on top of something, but if you don't have good drivers then nothing will be solved. Someone thinks that flaky composting support is X's fault? X and Compiz have it done. If a graphics card doesn't support the needed GL extensions sufficiently enough, it looks or performs like garbage.

To all those that truly love and understand Wayland, how does this graphics environment improve the quality, efficiency, and advancement of Linux based graphics drivers? Because frankly, that is really the only problem I ever seem to have with Linux based desktops.

Re:A bit big for their britches?

[ZaphDingbat]

"Survival of the fittest" requires diversity. If you're saying nobody should even attempt to replace X, then you have no diversity, and you have no "survival of the fittest," merely "survival of the incumbent."

Let Ubuntu develop Wayland. If it's a good idea, people will pick it up. If it's not, people will stick to X. THAT is survival of the fittest.

Re:A bit big for their britches?

[HermMunster]

In the past there hasn't been that many changes that had no value. I'm sure Canonical has a timeline set up for what changes they will implement (small and large alike). Though I do believe they have made some changes that were done for change's sake, I don't believe that's their overall plan as there are many changes are actually pretty nice.

One most notable thing about those that claim to have tried Linux is that they have not used Linux seriously for an extended period of time. Linux itself is a great OS and once you understand it you will understand where so many of us are coming from, but if you give it a half hearted effort for a few months and then claim it is too different without putting the effort into it to learn it you'll never be satisfied and you'll always end up giving it bad press, even to some who would seriously benefit from using it over Windows.

A new install of Ubuntu is pretty nice, though I wish they'd pull all the notebooks from their designers and force them to use various desktop and odd displays for extended periods of time. That'd give their designers some perspective on the way a good percentage of people see their product. Wide screens make Ubuntu look very nice. They should put some effort into those older non-wide screen displays. And those purple colors have to go. I think it is universally considered ugly. Is the person that developed that color-blind?

Re:A bit big for their britches?

[ADRA]

Pulse was in Fedora 8 -- November 8, 2007
Pulse was in Ubuntu 8.04 -- April 24, 2008

You may want to fact check before making assumptions.

Re:A bit big for their britches?

[t2t10]

Many Ubuntu users are in academia, research, and industry, and they use tons of other X11 toolkits. Wx and Tcl/Tk are still extremely common, for example. Some apps have their own built-in toolkits that you don't even know about.

Nobody has made running an "X11 server ATOP a new system" work, not Microsoft, not Apple, not anybody else: keymaps don't work properly, drag-and-drop doesn't work properly, notifications don't work properly, etc.

And why even bother? X11 is fast, it has a good wire protocol, and it gets the job done. Why would you possibly want to spend years figuring out how to do all of that from scratch? What's the benefit? What real-world problem is this supposed to fix?

Re:A bit big for their britches?

[teridon]

Nobody really writes applications for X anymore [...]

I must be a "nobody", because I just spent the last few hours adding a "plasma-saver" bar (a black veritical bar marching across the transparent lock) to alock. Of course, starting with someone else's code is way different than writing from scratch.

Re:A bit big for their britches?

[suy]

I remember a discussion a year or two ago here on Slashdot how X was badly in need of replacing. Sounds to me like Canonical have the right idea, and the impetus to make it happen.

You know that the guy developing Wayland works for Intel, and previously worked for Red Hat? I'm puzzled that everyone gives credit to Canonical for everything, when they usually seem to do very few work outside of their distribution, with not much interest in pushing changes upstream.

Re:A bit big for their britches?

[somersault]

I didn't say that Canonical were creating the project. I'm sure there have been many ideas and projects trying to replace X. I'm saying Ubuntu as a distro have the right idea. There may be many smaller distros that use a different windowing system, but the heavyweights tend to be wary of rocking the boat too much. Canonical have the right attitude of not being afraid to make the changes that are necessary, even if some people are going to be frothing at the mouth and screaming "HERETIIIIIIIIICS!!! >F ".

Re:A bit big for their britches?

[Magic5Ball]

What's the business or technical reason for maintaining compatibility between home and technical computer desktop environments? The largest potential for growth would seem not to be in the technical market, where adoption of (other brands of) Linux is already strong and approaching saturation, but in the relatively untapped home or business environment where Ubuntu has some core competencies.

Also, I don't think this move asks "everyone to rebuild and recode" as much as it asks those developers who want to follow Ubuntu into the non-technical market to do the work to understand their users' needs, rather than imposing a one size fits all paradigm. The scientific, mobile, and entertainment appliance Linux distros have been far more successful than most because they understand market segmentation.

Re:A bit big for their britches?

[HermMunster]

I believe X was re-written between that announcement and now. Not long ago they completely redid X.

The thing is that Canonical isn't the impetus to this. The product was developed, and has been used in Meego. Meego's interface is a precursor to Unity. And, I doubt anyone believed that you'd be running too many gnome or KDE apps on Meego.

I don't think many vendors are using Meego now as most have shifted to light-weight Linux installations or to Android.

Dear Mr. Shuttleworth, the PC is not a tablet in disguise. We want powerful desktops. Weak menu emulators are not powerful desktop managers.

Re:A bit big for their britches?

[nedlohs]

So the "new kids" aren't the ones who are supposed to try out the new stuff? The old guys with old set in their way users should do so instead? I'm pretty sure that in almost every human endevour the "new kids" try the new things.

And what does Fedora have to do with Ubuntu's success? They're build on the back of Debian.

Re:A bit big for their britches?

[Anonymous Coward]

Also, the primary reason people advocated not directly programming in Xlib was that it would drive you insane. As in, screaming, gibbering, wild-eyed, tearing-out-chunks-of-your-neckbeard insanity.

Ia! Ia! Codethulhu fhtagn!

Re:A bit big for their britches?

[somersault]

Just had a look at what they're doing with Xorg. That's cool, but rewriting a current system is a lot more work than starting afresh. I'd never heard of Wayland before today, and while I've heard of Meego I haven't really looked into it.

What I'll ask though is: why do you think a windowing system first developed for use with tablets is inappropriate for more powerful machines? Just because it's efficient doesn't meant it won't scale well. In fact, it's much better to be efficient even if you are running a powerful machine. I hate the attitude of "we have powerful computers, so we don't need to write efficient software". Piling more layers on top of X to get desired results is resulting in fragmentation and inconsistencies. Using a system that's designed from the ground up to provide a suitable environment for modern UI rendering is just good sense IMO.

Re:A bit big for their britches?

[maugle]

I don't think Ubuntu is trying for an even larger share of the current Linux desktop market so much as they're trying to grow the Linux desktop market itself. Much like how the Wii was derided by existing gamers, yet made a killing because it drew in people who had never even considered buying a game system before, Ubuntu aims to draw in people who had never even considered running Linux before.

And the only way to do that is to push changes like this. Let's be honest, X is usable and well-established, but it's not exactly desktop-user-friendly. I can't count the number of times I've been stuck on the command line trying to get X back into working condition. Throw average users into that situation, and they'd run screaming back to Windows.

Re:A bit big for their britches?

[Anonymous Coward]

If the Ubuntu crowd wants to spend their time, resources, and energies writing the neccesary code to let GNOME and KDE work with Wayland, why stop them? The attitude I got from reading the article was that Ubuntu is trying this out and willing to invest time and resources into other projects to improve the overall quality of the Ubuntu project. Why is that such a bad thing?

And further, who the fuck are you to tell them how to proceed?

Re:A bit big for their britches?

[Anonymous Coward]

Forks are GOOD for software evolution - it's how new ideas get tried out!

Agreed. Let's leave it for natural selection to decide if this is a good or bad idea

Re:A bit big for their britches?

[elrous0]

When I tell my techie friends that Linux will never, ever, ever be a significant player in the desktop market they often challenge me. My answer to them is thus: Linux is exactly the kind of product whose most ardent supporters actually *resent* success. Anytime a distro like Ubuntu looks like it might go even slightly mainstream, the Linux community viciously turns on it. Linux is like a starving artist whose entire existence is predicated on STARVING. The Linux community will never have success because they don't WANT it, wouldn't know what to do with it if they got it, and resent even the attempt to achieve it. Linux is a born loser, championed by born losers. It can't be anything else.

Re:A bit big for their britches?

[Knuckles]

This new program is a change simply for change's sake, and is no better than the old program.

Can you elaborate how you came to this conclusion?

Re:A bit big for their britches?

[Knuckles]

There was absolutely no announcement to move away from Gnome. Unity is no move away from Gnome.

Re:A bit big for their britches?

[Thumper_SVX]

So your argument is that changes to Ubuntu will cause a drop in it's market share? I highly doubt it, to be honest. The traction is already there and the only people who are going to avoid Ubuntu because of the window compositor they're using are going to be those "in the know". Sorry, those people aren't the target market for Ubuntu anyway.

Seriously, what's going to be the actual cost of this move? Some older application incompatibility? Irrelevant to the target market because generally the target market for Ubuntu likes "new and shiny" and doesn't give a fig for the underlying architecture (except that it's Linux and free).

Me, I'll be one of the many "in the know" who won't really care because I know that all the applications I actually run on Ubuntu are going to be recompiled for this layer anyway. I rarely use software from outside the official repos because there's just nothing I feel like messing with. It works, and so long as it continues to work I will continue to use it. I don't see this changing, and this change that's being proposed will come up slowly and carefully. I have complete faith in that.

I'm not one of the Ubuntu faithful necessarily... but I do like it when my computers just get out of my way and let me get my real work done.

Re:A bit big for their britches?

[kaffiene]

Yeah, I agree. I use remote X sessions from time to time, but there's nothing I do that can't be run in a different way (VNC is fine). And, we've all been aware for a long time that Xwasn't the most efficient beast around.

Let's see what Cannonical comes up with.

Re:A bit big for their britches?

[Anonymous Coward]

However, there were other alternative OS offerings that were built to fit the multimedia based single user desktop market like BeOS or now its OSS imitator Haiku.

Eh, that's a design decision of a different scope, which may be influenced for non-technical reasons such as name recognition. Nobody wants that Haiku OS, they want this thing they keep hearing about, Lean Nux.

Can't fault them for kow-towing to that reality.

It may not be the ideal decision, but sometimes you have to play with the cards on the table.

Me, if I were a magic genie in charge of Linux, MacOS, Windows, or computers in general, I'd probably dump everything and start fresh. Including QWERTY!

Re:A bit big for their britches?

[Anonymous Coward]

Exactly... writing a simple application for X was arcane when I did it almost 10 years ago. Nobody really writes applications for X anymore - they write them for GTK, KDE.

Did you bother to RTFM? You were never supposed to use xlib directly. That's what toolkits are for!

How many people program win32 directly these days?

Re:A bit big for their britches?

[Anonymous Coward]

Work on QT has allready started and people at Intel and nokia are playing around with wayland on MeeGo so that shouldn't be much of a problem. GTK has a wayland branch aswell so things have been moving in that direction for a while allready.

Re:A bit big for their britches?

[Anonymous Coward]

...but I do like it when my computers just get out of my way and let me get my real work done.

Which is exactly why many will jump ship. We don't want to have to wait around until everything works like it does right now. We want to have a stable system with no major, mostly untried, changes. Why can't Ubuntu fork, iron out the problems and then make the switch when their user base can be confident that their systems aren't going to get screwed over as a result of over-eagerness. Canonical have made errors in pushing things out too quickly before and I do not trust them with this change.

Wayland may well be a step forward when it matures. Until then let us have a choice.

Re:A bit big for their britches?

[Anonymous Coward]

> And there are long standing issues with it that become glaring as we move forward.

Such as?

> Long known issues such as 3D support.

IOW, you don't have a clue. OpenGL was specifically designed for X. Ports to other platforms came later.

In fact, X's client-server architecture has had a big influence on graphics on all platforms. As the demands have increased, the need for pipelining has also increased. The end result is that even "direct rendering" APIs have started looking more like X, except with the network connection replaced by the AGP/PCIe bus. This was never a problem for X programmers, who are intimately familiar with the concept of a round trip. For Windows programmers, the concept of structuring code to avoid pipeline stalls took some getting used to.

The problems with 3D support on Linux have nothing to do with the X architecture and everything to do with it being a minority platform, particularly for gaming (until relatively recently, the "serious" 3D world was dominated by SGI systems, which use ... X).

Re:A bit big for their britches?

[epine]

I live on the not-so-Pacific coast. One day the buildings here will be sorted into two rough groups: the OLD buildings from the Port-au-Prince distribution, and the NEW buildings from the Santiago distribution.

Oh, hell. I'll include a link for the geographically impaired. Peter Haas: Haiti's disaster of engineering

It's not the age so much as the ethos of the day. Network transparency in X was never foreseen to handle streaming HDTV, 3D animation frame rates at the limits of human perception, or strange constraints of the pocket toys that are gaining prominence.

What I don't get in this discussion is the fear that native Wayland apps won't play nicely with remote X sessions. How many people do full motion video editing over remote X sessions? This is the kind of app most in need of a tighter and more seamless binding to the local display.

Is someone going to port the 2D Oracle installer to Wayland native? What the hell for?

What I expected to see here was more debate about whether Shuttleworth is a positive driving force in the Linux user experience ecosystem.

I'm inclined to think from the comments I've read here that this is a needed initiative. "We've always done it that way" is hardly a better reason to keep something around, than "old" is a reason to toss it out.

I suspect there's an element to this agenda of improving the working relationship with graphic chip vendors so that Linux does not run so far behind proprietary releases, moving X a little further out of the picture of what open source contributions from Intel and AMD need to validate as part of the Ubuntu release cycle, in performance validation, if nothing else.

Re:A bit big for their britches?

[SanityInAnarchy]

Chrome does use GTK+ on Linux. Not for everything, but WebKit needs some sort of widget toolkit.

Re:A bit big for their britches?

[mdmkolbe]

Have you people used X?

Have you looked at the Wayland design? It doesn't do rendering ... at all. Wayland just does compositing.

Yes, X needs to be replaced (or so the meme goes), but Wayland is only half a replacement and in this case half is worse than none.

Re:A bit big for their britches?

[mdmkolbe]

Do you suppose writing an app directly on Wayland will be any less arcane?

Re:A bit big for their britches?

[Anonymous Coward]

Being OLD is never a good excuse to replace it.

Um... being old is always a good excuse to replace things. That's why they die.

Re:A bit big for their britches?

[npsimons]

I remember a discussion a year or two ago here on Slashdot how X was badly in need of replacing. Sounds to me like Canonical have the right idea, and the impetus to make it happen.

People have been saying X was in need of replacing shortly after it was created (and yes, that was before I was born). Does X have issues? Sure. Are they dealbreakers? Obviously not, otherwise X would have been dropped instead of being forked those many years ago.

I've not had a look at Wayland, but it sounds to me like the same-old, same-old whining by "end users" and "gamers", specifically "waaah! my 3D isn't fast enough!". As for the speed, 90% of users don't care how fast it is (which I might add, X is not that slow; *by far* faster than VNC and RDP over the network, and none to shabby with accelerated drivers on the local machine). As for the complexity, 90% of programmers are using a toolkit that eliminates the difficulties in programming for X. And X has it's benefits (such as network transparency, which *will* become more important in the future; this whole "network isn't important, we should optimize it out" thing is a fad).

The thing that bothers me about Wayland is that it seems to want to enforce policy and have it's own windowing system built in. I *like* the fact that there is no One True Desktop for Linux (or X); I can pick a different desktop any time I please, and still run graphical applications from my headless servers. And no, VNC and RDP don't come close; I don't want to have to dedicate a desktop just to run a single GUI app from my servers, and wait for the horribly slow refresh. I also like the fact that X is lightweight enough to run on my netbook and my smartphone. Which also means I can run graphical apps to or from either of those, and from my aforementioned headless server.

As for those saying that Wayland will probably have support for X11 "much as Windows and Mac do", I say to them, the way Windows and Mac "support" X is one of the reasons I don't run Windows or Mac.

Re:A bit big for their britches?

[Anonymous Coward]

Me, if I were a magic genie in charge of Linux, MacOS, Windows, or computers in general, I'd probably dump everything and start fresh. Including QWERTY!

Meh, you think too small. Revamp the alphabet in use. 16 letters should be enough for everybody, enabling usable functional 3-row keyboard (2 rows of letters, 1 row of numbers). And this will of course require changing language too, so might as well go with a regular, unambiguous one that could be easily understood by computers too. We'd get an easy-to-learn global language *and* accurately speech controlled computers, and avoid development of local dialects because computers would use the official version anyway, so people could not diverge too far from that.

Re:A bit big for their britches?

[Have Brain Will Rent]

Exactly... writing a simple application for X was arcane when I did it almost 10 years ago...

Ahhh that's just how they separated the men from the boys. Once you read the first 4 or 5 volumes of the X books using xlib wasn't that hard. Of course if you were all fancy like or lazy there was Motif.

Re:A bit big for their britches?

[PincushionMan]

I'm still not quite sure how this cult of ubuntu came about, don't get me wrong there's nothing wrong with another distro, but people seem to think they do more than they actually do.

Actually, I think the Cult of Ubuntu came about with their Free CD project. I know one day, after being frustrated with Gentoo, I gave Ubuntu a whirl and liked it. At work, I was allowed to install a major Linux distro. I choose Gentoo at first, couldn't make it work (reliably) with the hardware, and converted back to Ubuntu. Except for a brief stint with Fedora (and OSS) when PulseAudio shipped and I was required to have working audio for voicemails, I've been fairly faithful.

So there's where the CoU comes from - free swag.

gnome developers what?

[Lord Ender]

The move means there is now little reason for GNOME developers to recommend Ubuntu as an operating system.

I don't know a single person, not one, who makes his OS choice based on what "gnome developers" recommend. Why was this bit even added to the summary?

Re:gnome developers what?

[Dog-Cow]

It's a quote from the article, you ignorant and stupid prick.

Re:gnome developers what?

[Anonymous Coward]

Who are you talking to? The /. editors never read the comments silly!

Re:gnome developers what?

[Fnord666]

I don't know a single person, not one, who makes his OS choice based on what "gnome developers" recommend. Why was this bit even added to the summary?

Because it was in the article and summaries have become nothing more than cut and paste jobs. There isn't any actual summarizing any more.

Re:gnome developers what?

[epo001]

I don't know a single person, not one, who makes his OS choice based on what "gnome developers" recommend. Why was this bit even added to the summary?

It wasn't "added" it is a direct quote from TFA, the correct question would be why was this quoted?

Re:gnome developers what?

[kaffiene]

Because it was written by a whiny bitch? I mean, did you read the article? It's basically "they took my favourite toy! wah!"

Re:gnome developers what?

[Requiem18th]

So you don't know Foresight Linux users.

woohoo!

[Anonymous Coward]

about damn time somebody had the balls to drop X. Oh, but i can't run x over ssh over a 300 baud modem!!!!111!!!!eleven!! Well, you can't drive a ferrari through the outback either, but it get more pussy than a jeep.

Re:woohoo!

[AnonymousClown]

You couldn't drive today's Jeeps in the outback either.

Re:woohoo!

[Anonymous Coward]

I beg to differ. I'm not near the outback, but I have damn sure gone places that I had no business going. I got all sorts of looks from "hell yea" to "crazy bastard". If you think the new Jeeps are pussified, then you are only judging the exterior. Under the beauty, it is definitely a Jeep. /me has a 2010 Unlimited untouched (for now). :)

Re:woohoo!

[Anonymous Coward]

about damn time somebody had the balls to drop X

Android doesn't use X.

Re:woohoo!

[Anonymous Coward]

I like your car analogy. However, I fail to see how Wayland will give me any pussy. Will you care to explain to a fellow lonely slashdotter?

we'll be able to retain the ability to run X apps

[commodore64_love]

He may be "confident" we'll be able to retain the ability to run X applications in a compatibility mode, but I'm not. MY PS3 doesn't run a whole host of PS1 or PS2 games even though Sony claimed it would. (So I bought a space PS2 instead.) Windows Vista and Seven doesn't run old 3.1 or 95/98 apps. Mac OS X doesn't do well with classic PPC or 68000 apps.

Nope. Not confident at all.

Wayland...

[DaPhil]

For anyone else wondering what Wayland is: "Wayland is a lightweight display server for the GNU/Linux desktop. Started by Kristian Høgsberg [...] the software's stated goal is "every frame is perfect, by which I mean that applications will be able to control the rendering enough that we'll never see tearing, lag, redrawing or flicker"" (Wikipedia)

Here is the website and the wikipedia entry.

Re:Wayland...

[zpeidar]

"Development status: Alpha" says the Wikipage,
"Interesting", says Mark ShuttleWorth
"going into production Ubuntu" says Canonical,
"going away from Ubuntu" says I.

Re:Wayland...

[maugle]

"Interesting", thinks Shuttleworth,
"Will devote resources to the project, hopefully shift to using it in a year, but will focus on maintaining compatibility with X applications", says Shuttleworth on his blog,
"OMG, UBUNTU DUMPS X!!!", reports Slashdot.

Re:Wayland...

[Knuckles]

"going into production Ubuntu" says Canonical, .

"+3 insightful" /rolleyes

What Shuttleworth actually wrote: "Timeframes are difficult. I’m sure we could deliver *something* in six months, but I think a year is more realistic for the first images that will be widely useful in our community. I’d love to be proven conservative on that :-) but I suspect it’s more likely to err the other way. It might take four or more years to really move the ecosystem. Progress on Wayland itself is sufficient for me to be confident that no other initiative could outrun it, especially if we deliver things like Unity and uTouch with it. And also if we make an early public statement in support of the project. Which this is!"

      -- http://www.markshuttleworth.com/archives/551

Re:Wayland...

[Anonymous Coward]

> Wayland is a lightweight...

Let ten years pass.

Re:Wayland...

[p0larity]

I love you Kristian Høgsberg. That's one of my main gripes with desktop Linux in general these days.

Summary's BOGUS...

[Anonymous Coward]

Uh... Guys... Wayland doesn't preclude X11. Think of X11 as a two part system. One's the rendering and compositing layer and the other is the network transport layer that makes it network transparent. Wayland's the driver backend guts. They've shown MULTIPLE X11 desktops being ran on top of Wayland.

This isn't the thing that many make it out to be. SERIOUSLY.

Re:Summary's BOGUS...

[Waffle Iron]

That's good to hear. All I care about is one thing: does "ssh -X" work correctly and transparently out of the box with all included apps. If so, no problem. If not, I'll switch distros.

Re:Summary's BOGUS...

[Nimey]

But then people wouldn't be angry or afraid.

Re:Summary's BOGUS...

[mdmkolbe]

Wayland is a protocol for a compositor to talk to its clients

That seems to contradict you. It is from the Wayland homepage. As best I can understand Wayland moves the compositor into the server and delegates as much functionality to standard libraries (e.g. OpenGL ES).

I've never heard of Wayland before so if I've wrong please correct me. I just want to understand.

Re:Summary's BOGUS...

[interval1066]

So, can one ask; WHY THE MOVE? I see random comments here ("X sucks") and the like, but what are the specific technical reasons for this? I don't know that I really care one way or the other, I think the only thing I care about is compatibility. But I am curious to know Canonical's decision for this.

Re:Summary's BOGUS...

[Patrice]

I guess this is the case for many people: as soon as you have a setup of several computers at home and organised them into a local network (and not just a bunch of individual machines that connect to the Internet independently), you start to rely on network management (not even speaking about enterprise setups). I use daily "ssh -X" to run remote GUI apps on my netbook.

I also happen to maintain my parents' computers (located 25 km away), and once in a while I pop up one of their applications on my screen, to reconfigure it when something's broken or not working as wished, at least when there are no good CLI way of doing it, which is the case with most GUI apps today (it's quite slow over the internet with different providers, but it just works, which is enough for my needs).

Being a long-time Ubuntu user, I'd hate it to have to switch away from it (it takes time to reinstall all those machines - and I can't upgrade them remotely as I'm doing today with Ubuntu), but remotely running GUI apps is a must.

If in two years (or whenever the switch is done), I can log in to any computer running Ubuntu 12.04 or whatever and run any regular app forwarded to my own netbook/laptop transparently, then no problem... otherwise...

I'm OK with Ubuntu switching away from X and all, but I'm wary of the apparent lack of concern for network transparency support for regular apps. But they still have time to understand that their user base does not only consist of single computers connected individually to the internet.

Re:Summary's BOGUS...

[Anonymous Coward]

Yes, but can you run wayland apps on a remote host but have the graphics shown on your local desktop?

The X network transparency for me is one of the biggest killer apps on Linux, and now we want to abandon that so we have the same outdated, pre-internet conceived local only GUI model that is used by Apple and Windows? Seriously? We want to take a step backwards in development? Just for extra eye candy?

I hope they make wayland network transparent. And I'm sorry, but saying "use VNC" like some have said is not an option. Only those who have not used remote X can imagine that VNC is a good replacement. VNC sucks, RDP less so, but remote X is better than both, despite being a 20 year old protocol that has been hacked and kludged to death.

I'm happy that people have the option to push with new technologies, but I think this is a bad direction. Either way I'm willing to give them the benefit of the doubt. If Ubuntu stops being good for me and those I support I'll just switch to another distro. I'm sure there will be people willing to maintain X if/after Ubuntu switches.

Re:Summary's BOGUS...

[Svartalf]

I'd have a problem if you couldn't do that or have NX be busted completely- neither are acceptable and shouldn't be considered so unless you've got comparable answers in hand. Unless you can do something identical with Wayland, you're going to have fewer takers than they think on this.

Re:Summary's BOGUS...

[Svartalf]

Because you can do touch displays and similar things VASTLY easier with the move than staying with the stock xorg stuff. The upshot is that they've already demonstrated compatibility in the large with the current ecosystem and if they come up with comparable answers while they're moving things to the better solution, so much the better, really.

Re:Summary's BOGUS...

[Anonymous Coward]

You do know that when doing that 'ssh -X' thing that anyone who has compromised the remote host can sniff all keystrokes on your local X server (regardless of focus), take screenshots of the entire display, and other such things, right?

Re:Summary's BOGUS...

[Anonymous Coward]

VNC sucks, RDP less so, but remote X is better than both

Aside from being a huge security risk since remote hosts can sniff all keystrokes and take screenshots of your local X display...

Re:Summary's BOGUS...

[electrosoccertux]

at least when there are no good CLI way of doing it,

shh quiet you.

Re:Summary's BOGUS...

[Abcd1234]

Someone's never heard of "ssh -X"...

Re:Summary's BOGUS...

[markatto]

`ssh -X` doesn't usually work "correctly and transparently out of the box," at least for most people's definition. You might want to look at `ssh -Y`.

Re:Summary's BOGUS...

[Dr_Barnowl]

I'd suggest it's mostly to do with the move to compositing window managers, which X was not designed for.

The most specifically obvious example - video.

Video is these days presented by slapping a video texture on a window. If the render of the video frames to the texture isn't done simultaneously with the render of the video frames to the screen, you see "tearing", or a line where one frame ends and another frame begins, something you shouldn't see at all.

The same phenomena happens with composited windows, which are also rendered textures ; drag a window around at speed on a poorly configured X / Compiz desktop and you will see the tearing where two frames of the desktop texture are being rendered out of sync with the screen.

Many users now have more than one screen. X cannot render a single desktop to two screens and keep both synced. While it's possible to get one screen synced, it's not always easy, and getting both screens synced seems to be impossible.

Now, this clearly isn't a hardware limitation, because Windows can and does do this right - I have never seen application windows or video tear on the Aero desktop. This is on the same nVidia card, both operating systems running the official nVidia drivers.

This is one of the few things about Linux that annoys me when I compare it to Windows. The other is PulseAudio - but I have workarounds for my PulseAudio problems. It just looks sloppy to have great big tears in your otherwise very pretty composite display, and if you want to enjoy a movie, you are either stuck with using only one of your screens.

Re:Summary's BOGUS...

[Anonymous Coward]

Someone doesn't know that that issue is due to X itself and isn't affected by tunneling...

Re:Summary's BOGUS...

[Abcd1234]

Well, then please explain. How can a *remote* host sniff my keystrokes and snapshot my local X display if the protocol itself is encrypted end-to-end?

Re:Summary's BOGUS...

[Anonymous Coward]

I'm talking about the remote host you sshed into, not any remote host.

Re:Summary's BOGUS...

[Abcd1234]

You still haven't explained how that's possible. You would have to intercept the traffic between the X client and ssh. How, precisely, would you do that, short of having root access on the box and some serious kernel-level hackery?

Re:Summary's BOGUS...

[ADRA]

Is the fact that window tearing, or only a single display handling poor video interaction a failure in X, or a failure in the driver developers? I really really want to know what AMD, NVidia, and Intel think about the change. Because frankly, if they aren't behind the shift, does anyone find this solution making it in the long term?

Re:Summary's BOGUS...

[Anonymous Coward]

-1 Flamebait for pointing out a clear security problem, HAHAHAHA.

Re:Summary's BOGUS...

[Abcd1234]

This is not the case for the other remote display systems that you were originally comparing X to. You obviously need to have a far higher degree of trust in a host that you 'ssh -X' to than VNC or RDP to.

Bullshit! If someone has root access on, say, a Windows box, all they need to do is install a hacked VNC or RDP server to proxy all traffic to them.

In short: if your box is rooted, you're fucked, end of story.

Re:Summary's BOGUS...

[Anonymous Coward]

All what traffic? The VNC or RDP client isn't going to be sending it to them in the first place unless the client is broken as well.

Re:Summary's BOGUS...

[Knuckles]

I really really want to know what AMD, NVidia, and Intel think about the change.

As for Intel: Wayland was started by Kristian Høgsberg, of Intel's Open Source Technology Center (OSTC).

Re:Summary's BOGUS...

[Abcd1234]

You have no idea what you're talking about, do you?

In the world of SSH-tunneled X, you have:

(Local Machine) X Server <- loopback traffic -> SSH tunnel <- loopback traffic -> (Remote Machine) X Client

The X Server forwards keyboard/mouse activity to the X Client, and the X Client responds with graphics data for display on the server. Now, your claim is that if the *remote machine* is compromised, then they can sniff traffic from the X Client, and you're hosed. Fair enough, that's true.

But let's compare to VNC/RDP. In that case, we have:

(Local Machine) RDP/VNC Client <- loopback traffic -> SSH Tunnel <- loopback traffic -> (Remote Machine) RDP/VNC Server <----> Underlying OS

The client forwards mouse/keyboard activity to the server, and the server sends that data to the underlying OS. The server then takes the results and sends back display information to the client.

But in this case, if the *remote machine* is rooted, it's trivial to create a hacked RDP/VNC server that would forward all keyboard/mouse events, and all display activity, to an eavesdropper. You're *no better or worse off* than with X.

Once again, if your remote box is rooted, remote or local, *you're fucked*.

Re:Summary's BOGUS...

[Anonymous Coward]

it's trivial to create a hacked RDP/VNC server that would forward all keyboard/mouse events, and all display activity, to an eavesdropper.

Really? ALL keyboard & display activity? Completely *outside* of the VNC client window, like I said? Ok, what's the IP address of your VNC server?

Re:Summary's BOGUS...

[Anonymous Coward]

so you're saying if I connect to your hypothetical trivially hacked VNC server, minimize my VNC client, open firefox on my local machine, you'll be able to see what I'm doing in firefox on my local machine at your hacked VNC server?

Re:Summary's BOGUS...

[Abcd1234]

Since when do X servers forward *all* display activity to an X client?

"But the X client could be hacked!" You say.

Yes, well guess what? The SSH server could be hacked. At which point I think you have far bigger problems, as your credentials are now likely compromised.

Besides, if you're *that* paranoid, just run an embedded X server, and display the X clients there.

Re:Summary's BOGUS...

[Anonymous Coward]

Since when do X servers forward *all* display activity to an X client?

uh, since X was created?

localhost% ssh -X rhost
rhost% import -window root shot.png
localhost% scp rhost:shot.png .
localhost% display shot.png
# observe that rhost can see your ENTIRE LOCAL X display

for even more fun:

localhost% xterm &
localhost% ssh -X Rhost
rhost% xev -id `xwininfo -root -children -tree -int | grep 'xterm' | head -n 1 | awk -F ' ' '{print $1}'` | grep XLookupString
# type something in the xterm started above
# observe that rhost can see keystrokes typed into windows that have nothing
# to do with the remote host
# this assumes you have only 1 thing named "xterm" running on your local machine.

"But the X client could be hacked!" You say.

Yes, well guess what? The SSH server could be hacked. At which point I think you have far bigger problems, as your credentials are now likely compromised.

Unlike the X issue, the SSH server being hacked doesn't automatically result in your local client machine being hacked unless you have a buggy SSH client or did something stupid like use the same password in both places or agent forwarding with a key that's also in your local authorized_keys.

Re:Summary's BOGUS...

[metrix007]

With the X server the machine does not have to be rooted, you have to trust who owns the machine. Not the case with RDP.

Re:Summary's BOGUS...

[Anonymous Coward]

You may want to try ssh -Y these days.

Re:Summary's BOGUS...

[Abcd1234]

localhost% ssh -X rhost
rhost% import -window root shot.png
localhost% scp rhost:shot.png .
localhost% display shot.png
# observe that rhost can see your ENTIRE LOCAL X display

That's not at all what you're claiming. Yes, run an Xclient whose entire purpose is to read the root window, and yes, they'll get it. But *your* claim is that if I run a simple xterm on the remote machine, the attacker will be able to view my root window. Please, explain how that's possible

# type something in the xterm started above

Once again, you're deliberately running a tool designed to attack your X server.

Once again, that's not the same as what you're claiming. Your claim is that by *passively inspecting* traffic on the remote machine, my entire desktop can be viewed.

Again, a hacked Xclient? Yup, that's a problem. Again, run in Xnest if you're that worried.

Re:Summary's BOGUS...

[Anonymous Coward]

Ah, but what of running multiple Wayland desktops being run on top of a single X-server?

Re:Summary's BOGUS...

[mdmkolbe]

Wayland does not do tunnelling (see [1]).

You can layer X on top of Wayland but then you may as well use X without Wayland.

The Sep 18 2010 7:31 pm post to [1] suggests designing a common network protocol on top of Wayland, but it doesn't look like the Wayland devs are going to implementing that.

In theory implementing such a protocol is easy. In practice it doesn't matter how easy it is if no one implements it or few applications use it.

Such a protocol might simply take X and subtract out the parts that Wayland takes over, but that is a good idea only if both the X protocol and the X server significantly simplify in the process. I don't know enough about either X or Wayland to know whether this is the case.

[1] http://groups.google.com/group/wayland-display-server/browse_thread/thread/e7ed0c0118fb31b4

Re:Summary's BOGUS...

[mdmkolbe]

Think of X11 as a two part system. One's the rendering and compositing layer and the other is the network transport layer that makes it network transparent.

But can I run X with just the network half and without the code/install/memory/cpu bloat of the other half? If so then it sounds like X-network-only would be a good complement to Wayland (though so far I haven't seen anyone implement it). If not then Wayland is just increasing the complexity of the system because I'll have to run both a full X server and Wayland server.

Re:Summary's BOGUS...

[Anonymous Coward]

Holy fuck. I never told you to run an xterm on rhost. re-read those directions. Those aren't tools designed to "deliberately attack X", they're bog standard X utilities that come with X.

I'm claiming the same thing I was claiming the very first time I replied to you - THERE IS NO equivalent to 'import -window root', or "tool designed to attack your X server" (by the way, those tools are standard X tools, they come with X) or whatever that for secure VNC or RDP clients. There is no way that a VNC or RDP server can read ALL keystrokes typed on the computer you're sitting at or get a screen shot of the ENTIRE screen you're sitting at like you can with remote X, which puts just an itty bitty ding in your "VNC sucks, RDP less so, but remote X is better than both" claim. There are things to like about X, but the security is a huge problem.

You don't need to run an xterm on the remote host. All you need to do is type 'ssh -[XY] rhost' and anyone with root or your acct on rhost can do all the above. You VNC into rhost and there's no way people on rhost can keysniff your local machine. I just told you exactly how to do it.

A bit sensationalist...

[Scyth3]

They're slowing transitioning away from X to Wayland. They're not straight up "dumping" X. It'll be there for quite a few releases. http://arstechnica.com/open-source/news/2010/11/linux-beyond-x-shuttleworth-contemplates-wayland.ars

Re:A bit sensationalist...

[Enderandrew]

We don't allow sensible comments here. You need to change your post to "Shuttleworth kicks Gnome developers square in the nuts".

Re:A bit sensationalist...

[Shotgun]

Dude, "kick in the nuts" is SOOO 2009. You need to change it to "Shuttleworth kicks Gnome developers in the face".

http://www.dvorak.org/blog/2010/10/25/dirty-political-ad-du-jour-mayor-kicks-kids-in-the-face/

Re:A bit sensationalist...

[urulokion]

It's also wrong. X will always be present. If they don't have X, you are shooting themselves in the foot with a bazooka. Without X, they loose a vast amount of software. And they loose all of the advantages of X.

Also I think that people still have an outdated view of X being a bloated monolithic mess. That's no longer the case. Since the migration from XFree code base to the X.org code base, X has been slimming down and improving drastically but slowly. They aren't making huge sweeping changes so they don't break things for a lot of people. For example, all of the bloat of xlib is being examined and excised a bit at a time. If they remove something and people complain about it, it gets put back in.

Re:A bit sensationalist...

[Kjella]

I figured it was something like that, because Wayland is far from a finished state. Also it depends extremely heavily on KMS, that Intel and AMD is working on but won't be widespread until at least another few years. And nVidia is not doing KMS at all, and probably has no plans to. In short, this is way, way out there.

Re:A bit sensationalist...

[urulokion]

Something I forgot. There is also a lot of work being done on X for the future, like being able to seemly move an application windows from one device to another. Imagine you have a Photo Album or Media application running on you smart phone. With a gesture or couple of touch, you are able to seamlessly move that application window to your large screen TV in your living room. That's the kind of future that I want to see. Discarding X and programs that use X keeps us stuck in the cable connected quagmire we are in now.

Oh no

[Artem S. Tashkinov]

Let me be the first to say that the current X.org drivers for DAAMIT, NVIDIA and Intel are all incomplete and buggy and what Mark offers will hit him and all Ubuntu users very hard.

And I'm quite sure ATI and NVIDIA won't bother releasing their binary drivers for this thingy in the foreseeable future (or maybe who cares where there are no AAA native games for Linux). With thousands of unresolved bugs in KDE/Gnome/X.org server itself, with many devices still unsupported or barely supported, I don't think it's the best endeavour in the immature Linux world.

Re:Oh no

[nixkuroi]

There weren't a lot of touch screen phone options either before the iPhone. Sometimes you have to lead.

Re:Oh no

[muridae]

Since it looks like it is aimed more squarely at the embedded device market, as a way to provide a full desktop on your phone, I think they are a little less worried about the ATI/Nvidia/Intel binary drivers. I suspect they are more concerned about companies like PowerVR and everyone who uses their designs.

Re:Oh no

[jedidiah]

I'm cool with this being intended as a strategic direction for mobile devices.

For the desktop, it's just plain retarded.

They seem to be fixating on the usual lame trolling directed at X rather than thinking about what this will actually mean once it is imposed on the users.

Re:we'll be able to retain the ability to run X ap

[Anpheus]

So run an older version of X in a Linux VM? All the current kernels run pretty well on all the hypervisors, so just download whatever current distro you like, put it in a VM, and store it for "legacy programs".

Likewise DOS, early Windows and Win9x all run in VMs fine.

Breathe Deep...

[gti_guy]

Calm down people. This isn't any different than Mac OS X using Cocoa for the desktop display and still having X11 available to run as another app. And yes (if you've never tried it), X tunneled through ssh works just fine on Mac OS X. It will be the same thing with the next release of Ubuntu. The sky is NOT falling.

Re:Breathe Deep...

[makomk]

Calm down people. This isn't any different than Mac OS X using Cocoa for the desktop display and still having X11 available to run as another app. And yes (if you've never tried it), X tunneled through ssh works just fine on Mac OS X. It will be the same thing with the next release of Ubuntu. The sky is NOT falling.

That's exactly why people are so worried, though. Like on Mac OS X, all the major applications will be non-X11 and will not be able to be tunneled over SSH. We're talking all applications that use GTK+ or KDE for a start, followed by other applications as soon as the manpower is available to port them. Currently on Mac OS X you need to use some horrid remote-desktop hack like VNC that essentially forwards the entire desktop over the network very, very slowly and it looks like Ubuntu is going to end up in the same situation.

Re:Breathe Deep...

[jedidiah]

No. X11 being on the Mac is like X11 being on Windows.

It is something that was bolted onto an already functioning system with all of the relevant bits including DEVICE DRIVERS.

Re:Breathe Deep...

[Jonner]

If this ends up like OSX, it will be a giant mistake. I am using OSX right now on my work-provide machine and running X11 apps. Though the X11 server is included and starts automatically, X11 apps are second class citizens. They don't interact with the task manager or menu bar the same as native Aqua apps. Every X11 window is treated as part of the X11 app, regardless of how many distinct programs are running.

If, on the other hand, Canonical is smart and doesn't hate their users, they will put a high priority on X11 clients continuing to function the same as they do now, which should be possible when Xorg runs as a Wayland client AFAICT.

Re:Breathe Deep...

[Anonymous Coward]

Yeah, I can tunnel X11 software from Mac OS X over shh, but the problem there is that there are no X11 applications for Mac OS X*

The end-game for Ubuntu using Wayland must be that all applications be Wayland not X11, at which point network transparency is lost forever.

Re:Breathe Deep...

[Jonner]

I also don't think OSX is a good example to follow as I'm using it right now and even on the same machine, X11 clients are second-class citizens. The way to avoid killing remote capability is to leave the X11 client in the widget libraries like Gtk+ and Qt and add the Wayland client, so that apps continue to work with either protocol. As much as Canonical may want it to happen, Gtk+ and Qt (and various other X11 widget libraries) will not suddenly transition to an entirely different protocol.

Apple, starting with the NextStep/OpenStep GUI libraries, had the choice to using an existing windowing system underneath (such as X11) but chose to write a brand new one and only later added X11 as an Aqua client. Canonical has a system based on X11 and wants to transition to a different protocol, but needs to support existing clients, so they shouldn't have an Apple-style attitude that every user and developer is captive to their whims.

Re:Breathe Deep...

[Richard W.M. Jones]

If it's like Mac OS X then it's still a step backwards because the native OS X apps cannot be run remotely (only X11 ones). That's removing a useful feature for no benefit (despite what people who don't know what they are talking about say, there is no performance penalty to network transparency, because locally it's all using shared memory).

Rich.

Re:Breathe Deep...

[WindBourne]

In general, most apps that are X11 tunneled are simple desktop apps. Few are things like Doom. It would not make sense. Now, once Qt/gtk are fully ported to Wayland, it does not mean that the X11 port will die. At least not for another 5-10 years. As such, the core apps will likely be available in both wayland and X11 versions. The issue is, what will speeds for the network look like in 10 years? It is obvious to me that VNC will be a speed burner as the norm is 1G connections and thin clients have fast GPU for processing VNC quickly. IOW, the move to Wayland/VNC is not a big deal.

Re:Breathe Deep...

[repetty]

If it's like Mac OS X then it's still a step backwards because the native OS X apps cannot be run remotely (only X11 ones). That's removing a useful feature for no benefit (despite what people who don't know what they are talking about say, there is no performance penalty to network transparency, because locally it's all using shared memory).

A useful feature that no one uses. Hell, most people forgot X11 support for remote graphical sessions even exists. That feature made a big difference in the 80's but not now.

VNC, anyone?

Re:Breathe Deep...

[EyelessFade]

That's exactly why people are so worried, though. Like on Mac OS X, all the major applications will be non-X11 and will not be able to be tunneled over SSH. We're talking all applications that use GTK+ or KDE for a start, followed by other applications as soon as the manpower is available to port them. Currently on Mac OS X you need to use some horrid remote-desktop hack like VNC that essentially forwards the entire desktop over the network very, very slowly and it looks like Ubuntu is going to end up in the same situation.

If you read their mailinglist you will see that they want to have some kind of ssh -X possibility.

At last.

[Anonymous Coward]

X11 needs to die.

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

No kidding. That was a big 'use a different distro' to me. How many devs are going to bother tweaking for 1 distro? Esp for something they wrote say 5 years ago and they have lost interest in? Oh thats right very few. Newer applications you may get people to bother to go fix it. Older ones? Not so much.

I read that as 'hey we would like to kill ubuntu pretty quick and we found a way to do it'.

Good On Them

[Anonymous Coward]

I think a big player should do something innovative like improve the user experience. If you don't like it then fork their distro or choose another one this is what linux is all about. So stop your bitchin until you give it a go.

Will it be faster and more responsive?

[anandrajan]

Now this is interesting. Currently, on my work PC, linux/X11 seems to be a bit slower (KDE 4 especially, GNOME a bit less and definitely not Enlightenment) than Windows XP. Will this move bring the response *feel* of the linux desktop (in Unity on Ubuntu) to be on par with XP? While there are many anecdotal complaints all over the web regarding the intrinsic slowness of X, this seems to be disproven by my Enlightenment 0.16 experience.

Re:Will it be faster and more responsive?

[3.1415926535]

No it won't. Contrary to popular belief, Wayland is not a panacea for poorly-written window managers or compositors (except that you won't be able to run kwin4 on it, so it'll force you to use something different and -- since kwin4 is so terrible -- likely better).

Re:Will it be faster and more responsive?

[Jonner]

Were there any compositing effects on the GNOME and KDE desktops you ran? If so, it doesn't make sense to compare directly to XP on the same hardware, since XP has no compositing. Comparing current GNOME, KDE, Win7 and Aqua on the same hardware makes more sense.

Since the primary motivation for Wayland is better compositor management on modern hardware, I'm sure it won't be a reasonable replacement for either Xorg or XP on older hardware. Therefore, it's important that default Ubuntu continue to be able to run with just Xorg on hardware that Wayland doesn't handle well.

Re:Will it be faster and more responsive?

[Hatta]

The lack of responsiveness is due to the fat toolkits people use. Moving to Wayland won't fix GTK or QT.

Re:Will it be faster and more responsive?

[spitzak]

It looks to me like the one thing Wayland replaces is the window manager and the compositor. Therefore it IS a "panacea for poorly-written window managers or compositors" (on the assumption that it itself is not poorly-written which I know may be a problem).

Like the X11 rootless server on OS/X, only client applications will work. X window managers and compositors are not supported, and they are replaced completely.

Re:we'll be able to retain the ability to run X ap

[LWATCDR]

Actually I don't see why this would be an issue.
You can run X on OS/X and WIndows so running X on top of Unity is probably going to be okay as well.
The reality is that very few programs use X directly. Most everything goes through GTK or QT. There is a version of GTK that runs on just the framebuffer already for embedded development.

In theory you could pick GTK, QT, or some other frame work to replace X and then run X on that. You could also have QT for GTK if you really wanted too.
Just to be clear I am not sure that GTK or QT have all the feature you want in a moder graphics system.
A lot of people still have issues with X as display tech. A big one I hear about all the time is that it doesn't unify video display with printing.
It may be time to let X go. At least someone is going to try. Who know it may be the best thing to happen to Linux in a long time.

X11

[muridae]

People griped about moving from X11 to X.org, they will whine and moan about this too. It's only taken X.org 6 years to go from "that thin fast X11 replacement" to "too bloated for new computers", that is slightly impressive.

</sarcasm> It will be a good thing to have a separate thin X server for netbooks and phone-like devices. Leave all the super fancy graphics available on the bigger graphics cards, and pare down what is overkill on palm sized screens with tighter power requirements.

Re:X11

[Anonymous Coward]

I'm not quite sure what your point was, but mine is this: X is a POS.

It never was very good, and as graphical displays have moved forward X has remained stuck in the past. It can't be fixed and it's time to start over. A lot of smart people have poured their lives into it, and it's still the same dog it was twenty years ago.

Re:X11

[jedidiah]

X may be a piece of sh*t. However, X on Linux got hardware accelerated 3D before BeOS did. This was all despite the fact that BeOS was supposed to be "better". All of that didn't matter in the end because modern hardware hooks weren't there. So the niftiness of the OS ended up being nearly entirely irrelevant.

Wayland sounds a lot like BeOS.

What else are you going to have to give up for Wayland? How far back will the Linux Desktop be set back because someone wants to recreate the BeOS experience?

Some of us actually use modern GPU features. So this isn't just an academic issue.

Re:X11

[3.1415926535]

X.org *is* X11. I think you meant XFree86 (which is also X11, by the way). That issue was political squabbling and an argument over whether or not forks are bad for open source projects. This is something entirely different.

Re:X11

[Microlith]

Leave all the super fancy graphics available on the bigger graphics cards

In case you missed what was going on in the mobile space, they absolutely can't leave support for the "super fancy graphics" out of any display server.

Re:X11

[Jonner]

X.org was never intended to be a "thin fast X11 replacement", but a fork of XFree which was more modular and easier to add new features to. Both X.org and XFree always have been and always will be X11 servers.

There are several small X11 servers, but Wayland isn't one of them (it is not an X11 server). In contrast with X.org, it actually requires the "super fancy graphics" features, which are present on most new graphics chips, whether on the motherboard or on a big card. X.org never has required anything fancy from the graphics hardware. It can still run on any device that provides VGA or VESA modes.

So, if you want as small of a system as possible using as little power as possible, you're almost certainly better off using X.org built in a minimal fashion or one of the other small X11 servers. However, the size of the screen is irrelevant, as neither X11 nor Wayland cares about policy (how widgets and text and graphics are drawn). The look and feel of the GUI is entirely up to the clients of either X.org or Wayland, which are mostly GNOME ones in the current default Ubuntu desktop, but will be Unity in some later release.

Re:X11

[Anonymous Coward]

BeOS? It died right before hardware accelerated 3d became mainstream. And not because of anything wrong, per se, with BeOS in a technical sense. It died because of business issues, nothing more.

I'd actually say it was dead before it came out, but I suppose there was a chance of some miracle happening.

Wasn't going to be the 3d graphics though.

Re:X11

[ADRA]

All true, but being somewhat interested in the X server around the time of the fork, very few people found the change to be a negative. XFree86 guys that didn't move with the change seemed to be against adding changes and extensions into the server, and only after the fork could things like RENDER and RANDR become first class citizens in the distro instead of being relegated to the experimental branches like KDrive. The Since it was a fork, its not like things were instantly broken. I think the big change was the executable was named Xorg instead of XFree86 which I assume broke some custom scripts.

That's an EXTREMELY bold move...

[starseeker]

There have been other projects over the years that have tried to improve on X (Fresco/Berlin and picogui readily come to mind) but I don't believe any of them have demonstrated results that seriously threatened the revitalized Xorg project.

I hadn't heard of Wayland, but I must admit since Xorg got going I haven't kept a close eye on that level of the graphics stack. Mark's blog post makes it sound like they're willing to ditch network transparency for better graphics effects, which makes me a little leery. Undoubtedly for most users that's the "right" approach, but if they do lose network transparency it's going to make Ubuntu an impossible choice in a lot of business environments where running apps from a server is part of day-to-day business.

Also, the amount of work to port all the requisite software/toolkits to a non-X platform is going to be... impressive. Haiku faces this problem, as do a fair number of older applications when looking at running native on Windows and OSX - it ain't easy. Plus, we're talking an entirely new backend in Wayland, one that's going to require (from the sound of things) rock solid OpenGL support.

Ubuntu has shown they can deliver in the past, and perhaps they can do it now, but I can't help but wonder if they realize the magnitude of what they're undertaking here.

Re:That's an EXTREMELY bold move...

[makomk]

Plus, we're talking an entirely new backend in Wayland, one that's going to require (from the sound of things) rock solid OpenGL support.

Not only that. It requires rock solid OpenGL support from open source drivers that implement both DRI2 and KMS. Closed source drivers cannot play at all; they're effectively locked out of implementing KMS support.

Re:That's an EXTREMELY bold move...

[Jonner]

I have also played around with Fresco and noticed that while there have been many claims that X11 needs to be replaced and a number of attempts to do so, it has not proven to be necessary and the attempts to replace it have fizzled. I don't know enough about the X11 protocol, the Wayland protocol, or the low level details of managing composition to evaluate how urgently X11 needs to be replaced, but it seems plausible that doing so could allow optimizations that are currently difficult to impossible. Smooth desktop effects on small and/or inexpensive hardware and simultaneous low power usage are great goals, but the need for a change as big as this needs to be well-justified.

In contrast to Fresco, Wayland is nowhere near as ambitious and therefore has a much greater chance of success IMHO. Like X11 and unlike Fresco, it doesn't attempt anything related to GUI policy, but concerns itself only with the low level display manipulation.

I'm sure that to provide adequate backward compatibility, Canonical will need to make sure X.org is available whenever a non-Wayland client needs it and those clients aren't second-class citizens. If they expect all currently used Widget libraries to be adapted to be Wayland clients, that's a huge mistake. It's similar to the transition to using Pulseaudio by default: when libasound clients worked as well as they did talking directly to hardware (which sadly was long after Pulseaudio was run by default), that took care of the majority of transition problems. Hopefully they've learned from their mistakes in that transition and will be more cautious and/or diligent next time.

Also, I'm pretty sure they'll need to allow selection of X.org instead of Wayland to talk directly to hardware on systems that don't have suitable graphics hardware.

Re:That's an EXTREMELY bold move...

[Anonymous Coward]

Qt already runs on Wayland and you better believe Ubuntu is headed towards Qt. I haven't heard any mention of it but I know the idea must exist. That idea being to come out with a GNOME-like desktop (or completely new desktop) that uses Qt instead of Gtk (KDE sucks so I know they won't go that way).

Re:That's an EXTREMELY bold move...

[Homburg]

Not only that. It requires rock solid OpenGL support from open source drivers that implement both DRI2 and KMS. Closed source drivers cannot play at all; they're effectively locked out of implementing KMS support.

Huh. I don't really understand how that would work. What's to stop NVIDIA from implementing the same APIs in their closed-source kernel module that are implemented in open source kernel modules?

Re:That's an EXTREMELY bold move...

[Anonymous Coward]

GTK+ version 3 and Qt already have Wayland backends, so getting applications that use those toolkits to run on Wayland shouldn't be a big deal. Also, I believe it was a recent Google Summer of Code project to implement the KMS and DRM from Linux in Haiku, so that shouldn't be an issue in Haiku.

Re:That's an EXTREMELY bold move...

[makomk]

What's to stop NVIDIA from implementing the same APIs in their closed-source kernel module that are implemented in open source kernel modules?

IIRC, kernel mode setting is tightly intertwined with core kernel stuff like virtual terminal and text console support, and the APIs used to do this aren't available to closed source modules. So they might be able to implement the appropriate user-facing APIs, but there's no way to get VT switching or text mode to work

Correction

[Enderandrew]

Ubuntu will still ship X. Unity will run on X. No definitive decisions have been made. Shuttleworth is considering a transition to Wayland, which he estimates will be 4 years down the road. He assumes at that time that KDE and Gnome apps should be able to run natively on Wayland at that time, but you can run a rootless X server alongside Wayland either way.

But it really is more fun to make non-sensical statements, such as suggesting that Gnome and X are intrinsically tied, and that wanting to replace X four years in the future is some massive insult to Gnome.

Pulseaudio again.

[Forty Two Tenfold]

I have a feeling this is going to be the same SNAFU.
And why does Ubuntu NOT support alsa environment (panel volume applet, system sound theme config etc.)?
This is fucking stupid.
I'm going back to Slackware. Or Gentoo.

Re:Pulseaudio again.

[wagnerrp]

Don't forget that Pulseaudio merely exists as a transport layer on top of ALSA.

Re:Pulseaudio again.

[McKing]

Why don't you read the article yourself, instead of relying on /.'s sensationalist headline. Ubuntu is not "dropping" X, Shuttleworth basically says that they are going to look into using Wayland at some point in the future, and even says that it may be years before Wayland is ready for that.

BTW, I like PulseAudio, now that the kinks have been worked out. Did those kinks take a release or two to get worked out? Yep. Did PulseAudio suck at first and was it a bad decision to jump on it so soon? Yep. Is it a better system now than it would have been without all that attention (read: Shuttleworth-funded developer time) put into it? Most definitely.

Re:Pulseaudio again.

[walshy007]

Your best bet if you don't like pulseaudio is probably to use jack instead, you can make it so that pulseaudio feeds it's streams to jack and by changing your asound.conf get alsa programs to use it too.

As a bonus you get high quality low latency audio, that doesn't die all the time like pulse and has more functionality... still think pulse should have never existed.

Re:Pulseaudio again.

[Forty Two Tenfold]

Well, PA still stutters on my Lucid. Sound system (fresh alsa or otherwise) is similarly dysfunctional in Maverick on my system (ALC268). Tried everything. The kinks still exist. Too bad I can't get PA smoke weed or the stutter would go away.

Re:Pulseaudio again.

[Jonner]

Choice of distribution is one of the great strengths of the GNU/Linux community and I wouldn't recommend Ubuntu for everyone, but it is still entirely possible to turn off Pulseaudio and configure libasound to use hardware devices directly, just as it's possible to use Pulseaudio on Slackware or Gentoo. Also, though Canonical did not transition to Pulseaudio carefully enough, I do think it was a good move overall. Hopefully if they decide to move to using Wayland by default, they will have learned from their mistakes in the Pulseaudio transition.

Re:Pulseaudio again.

[Forty Two Tenfold]

Yeah, by getting rid of PA you make several crucial features disappear unless you make some extra (IMO unnecessary) effort.

Re:Pulseaudio again.

[Forty Two Tenfold]

Still talking about Ubuntu.

Re:Pulseaudio again.

[Jonner]

So, what are these crucial features, and which distribution provides them while not running Pulseaudio?

Re:Pulseaudio again.

[Forty Two Tenfold]

As I mentioned above - gnome volume control and applet and gnome sound theme settings.
Dists - Ubuntu Intrepid and earlier, Slackware, Gentoo, didn't try others.

Re:Pulseaudio again.

[Rakarra]

Your best bet if you don't like pulseaudio is probably to use jack instead, you can make it so that pulseaudio feeds it's streams to jack and by changing your asound.conf get alsa programs to use it too.

As a bonus you get high quality low latency audio, that doesn't die all the time like pulse and has more functionality... still think pulse should have never existed.

And... none of the above beats having a sound card with a hardware mixer to avoid all that software mixing BS that the long parade of Linux sound servers never solved with any measure of stability.

EMU10k1 + ALSA for the win.

Re:Pulseaudio again.

[walshy007]

And... none of the above beats having a sound card with a hardware mixer

Actually, it can, the cpu can beat pci latency.

long parade of Linux sound servers never solved with any measure of stability.

Jack did, but people bitched about the extra 1% cpu usage, thus pulseaudios invention.

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

X isn't a completely different machine code architecture, so it's not nearly as bad as PS2 / PS3 compatibility or PPC / x86.
And Windows has poor compatibility because of either stupid things the old Windows did or because of stupid things the programs did that newer Windows can't support (glitch exploits)

It's just a windowing protocol, I'm sure the X compatibility mode for Wayland will be fine.

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

Correct me if I'm wrong, but most of your examples pertain to an hardware architectural change, for which a compatibility mode is (at least theoretically) much harder to solve than a software change.

And thus the curse of Open Source manifests itself

[mrpacmanjel]

And thus the curse of Open Source manifests itself - Develop something to to the virge of usability and robustness...then BAM! "Fuggit" let's start again in a new direction and it will be "better" and spend more years in development wilderness.

The Ubuntu distro is showing great maturity and evolving nicely....then "Fuggit!" let's go a different way and start the integration again!

The n900 Nokia computer, no it's a phone, no it's a computer (ad nauseum). The OS (Maemo) reached good stability and then BAM! Nokia said "Fuggit! We can do this better - let's start an OS called Meego(still not running properly)"

KDE 3 to 4, Gnome 2.x to 3....

"Welcome to the world of Linux and Open Source where everything is in a perpetual state of development and a finished release is just a pipe-dream"

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

..........and what's the big draw with ubuntu and this wayland thing then? It's ability to run Unity and a virtual machine where you have to have a real linux distribution where you can do your actual work? *Snort*

Like Mac OS X

[Balthisar]

So, it'll be kind of like running X on my Mac OS X machines. A modern display server, with the ability to run a non-root X on top of it.

Re:Like Mac OS X

[jedidiah]

> So, it'll be kind of like running X on my Mac OS X machines. A modern
> display server, with the ability to run a non-root X on top of it.

Yes. In other words it will be something that makes you want to delete the offending OS and run a proper Unix/Linux.

MacOS fits into the "nifty like BeOS but less useful in practice" category here.

Re:Like Mac OS X

[Dog-Cow]

Mac OS X is a "proper" Unix. X is crap. It was crap when you had to configure it on Linux by playing with modelines, and it's crap now.

Re:Like Mac OS X

[Anonymous Coward]

I think you can already run a non root X for the last year or so.

Re:Like Mac OS X

[Anonymous Coward]

Geez, you really, REALLY hate Apple. And shut up with this "proper" stuff - you honestly and truly are clueless. Thank god you don't work for me, or I'd have you fired. Bitter losers make for terrible workmates.

Re:Like Mac OS X

[Anonymous Coward]

Yes, exactly. Good news all round.

Re:Like Mac OS X

[jedidiah]

Oh don't kid us.

You don't have the authority to fire anyone.

You're just another bitter troll in his mother's basement.

My attitude about "MacOS as Unix" comes from decades of using real Unixen. Being able to employ tricks and tweaks that predate Linux is what makes Linux more of a Unix than MacOS.

"Forking the display driver" will similarly create a great deal of distance between Linux and every other real Unix just like the same thing on MacOS does.

Re:Like Mac OS X

[onefriedrice]

So, it'll be kind of like running X on my Mac OS X machines. A modern display server, with the ability to run a non-root X on top of it.

About time.

Re:Like Mac OS X

[DdJ]

Heh, my Unix experience predates X11.

Remember how SunView used to work back in the day? A whole family of virtual framebuffer devices, just like the pseudoterminal devices that are still in use. It was glorious. Then X11 came along and killed it (largely because of cross-vendor compatibility).

As a heavy duty Unix user since, oh, about 1985 or so, I find modern MacOS to be more Unix-y than Linux. (Going into precisely why would take pages.)

Re:we'll be able to retain the ability to run X ap

[MBGMorden]

Not too bright are you? X11 is just a display protocol. While all those things you list might be true, I think it's SPECIFICALLY telling that both Windows (choose your version) AND Mac OS X BOTH run X11 applications just fine using compatibility layers.

Setting up a compatbility layer for people to be able to use X11 apps on a non-X11 display is utterly trivial. I was doing it with Hummingbird eXceed on Windows over 10 years ago (and I'm sure people have been doing it a lot longer than that).

Wayland can host X

[sd.fhasldff]

Then it's a good thing Wayland can host X. It would require some (reportedly) minor adjustments to X, but it would be transparent to individual applications.

Re:Wayland can host X

[mR.bRiGhTsId3]

However, if there are programs on a Wayland machine that others want to run remotely in an X-server they will be unable to. All machines on the Wayland machine will expect a Wayland server. So, while Wayland users can access X programs remotely, X users will be unable to use Wayland programs remotely in the same way.

Re:Wayland can host X

[AmaDaden]

Is there any reason Wayland can't make an X-server look alike layer for that as well? It would be extra over head for the PC hosting a wayland server but it would a low things to work and ease a transition to Wayland.

Re:Wayland can host X

[Kjella]

It should not be too hard to make a mode where Wayland renders it, then sends it over X as a dumb pixmap. It would not be pretty nor fast, but it would function for occasional use.

Re:Wayland can host X

[shutdown -p now]

So, while Wayland users can access X programs remotely, X users will be unable to use Wayland programs remotely in the same way.

... which is not the scenario GP described as something that may be realistically needed, and which should be considered in any migration from X.

The problem is with legacy apps written against X11. Updating the clients is much easier.

Re:Wayland can host X

[Zo0ok]

Wayland programs will not be able to run remotely. X programs should still be possible to forward between two computers using Wayland, as long as the client runs X-server on top of Wayland. The appliation server (X client) should only need a few X libraries.

Re:Wayland can host X

[Anonymous Coward]

(can't post on my username because I'm moderating)

The difference is which machine runs the code that draws on the windows. On X, it will be the remote machine (ssh target). O Wayland the client would render the windows, making it less thin

Re:Wayland can host X

[mR.bRiGhTsId3]

My point is, most of the programs on a Wayland running machine will expect the Wayland backend. Thus, legacy terminals will be unable to use them. Wayland is backwards compatible on the display server side, but not on the display client side. This means that upgrading is not completely easy. X servers are available for most platforms, but what is the point in installing programs using X display backends on a Wayland machine?

Re:Wayland can host X

[badkarmadayaccount]

Or st up wayland as an X compositing manager? Via RPC?

Re:Wayland can host X

[jbolden]

Well yeah that's the way upgrades / backward compatibility works. Wayland is acting like X12, you can run X11 apps on Wayland but you can't go backwards.

binary video drivers and wine

[atfrase]

Can anybody more knowledgeable than I comment on how this will affect binary (nVidia/AMD) video driver performance or compatibility, and thus Wine gaming?

Re:binary video drivers and wine

[makomk]

Can anybody more knowledgeable than I comment on how this will affect binary (nVidia/AMD) video driver performance or compatibility, and thus Wine gaming?

As far as I can tell, it'll mean the end of them for good. Wayland fundamentally relies on kernel modesetting support by the drivers to work, and closed source drivers are locked out of the kernel APIs required to implement this.

Re:binary video drivers and wine

[Narishma]

What's to stop them from just implementing kernel mode setting as an open source module and keep the rest of the code (the vast majority) closed?

Re:binary video drivers and wine

[Jonner]

Assuming Wayland and the hardware get along, which requires modern Linux interfaces like KMS, and X.org as a Wayland client has full GLX functionality, it should have little effect. That would currently (probably indefinitely) exclude Nvidia proprietary drivers: just one more reason to dump them.

Re:binary video drivers and wine

[Rakarra]

Assuming Wayland and the hardware get along, which requires modern Linux interfaces like KMS, and X.org as a Wayland client has full GLX functionality, it should have little effect. That would currently (probably indefinitely) exclude Nvidia proprietary drivers: just one more reason to dump them.

Actually it sounds like a great reason not to run Wayland.

A very large number of Linux users rely on the proprietary drivers, and the open source drivers will never be as good.

Re:binary video drivers and wine

[Jonner]

A very large number of Linux users rely on the proprietary drivers, and the open source drivers will never be as good.

That is a depressing attitude which could become a self-fulfilling prophecy. It's also irrational, as many Linux drivers are superior to their proprietary counterparts. If I believed it, I'd just settle for using OSX (or maybe even Windows) rather than trying to use a Free system. Why would you want to use a system that's always behind?

Re:binary video drivers and wine

[makomk]

What's to stop them from just implementing kernel mode setting as an open source module and keep the rest of the code (the vast majority) closed?

IIRC, NVidia don't want to release full source code for modesetting and frame buffer setup on their newer hardware, let alone more interesting stuff like memory management that they'd probably also have to open source at the same time.

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

You will be able to run X application on Wayland in a similar fashion that you run them on OSX. See "X as a Wayland Client" in http://wayland.freedesktop.org/architecture.html

OSX, Windows 95, Vista, Windows 7

[goombah99]

apple, Microsoft and Sun all have radically changed their widowing systems on many occasions while maintaining continuity for their developers. It did not mean no work, it just meant that recompiles could produce a functional product in most cases, albeit one that might look like poo and not have any of the new capabilities of the windowing system.

I find it somewhat hard to believe that the original design of X was so perfectly extendible that after decades of use it is not straining its seems.

So a change may be good.

However, i do see a downside. The nice thing about X unlike Windows and Macs main display interface is that it is more easily separated from the desktop. If you want to use a mac or windows system remotely you have to use something like VNC or a remote desktop app. In both cases you are getting the whole desktop not a display window. You can't run multiple instances of it. That's the main thing I like about X.

Re:OSX, Windows 95, Vista, Windows 7

[Lennie]

Actually I know for a fact RDP can do just apps*, just is Microsoft does not want you to that because of their licensing/business model.

http://rdesktop.sf.net/ and seamless rdp

Re:OSX, Windows 95, Vista, Windows 7

[sarhjinian]

X isn't actually that good at remote desktop, at least not by comparison. You have to install a fairly thick local client, and if you lose connection your apps die.

Holistically speaking, Citrix et al completely eclipsed X in terms of network-retargetable display a long time ago and for those times when you want to run an app remotely but don't want to lose the app if your connection dies (which is pretty much all the time) you end up running X over VNC anyway.

Me, though, I'd like to know if this change will finally allow me to have use a compositing window manager without tearing (you know, like MacOS and Windows have been doing for years now) and without having to restrict myself to an ancient or gutless graphics card.

Re:OSX, Windows 95, Vista, Windows 7

[Cajun Hell]

apple, Microsoft and Sun all have radically changed their widowing systems on many occasions

Sure, they changed certain aspects of the system, but they never strayed from the core design assumption: the husband must die.

Re:OSX, Windows 95, Vista, Windows 7

[Heddahenrik]

ThinLinc can handle advanced graphics where the graphics are generated on the server's graphic card, so it can be used on a client with a slow graphics card. It isn't (only) open source though, but there is a free version for up to 10 clients.

Re:OSX, Windows 95, Vista, Windows 7

[Abcd1234]

Holistically speaking, Citrix et al completely eclipsed X in terms of network-retargetable display a long time ago and for those times when you want to run an app remotely but don't want to lose the app if your connection dies (which is pretty much all the time) you end up running X over VNC anyway.

Well, smart people run X over NX, which provides wicked performance far eclipsing VNC, integrated encryption (it uses SSH, but it's part of the standard tool), detachable sessions, etc, etc. And that performance is specifically a consequence of NX proxying the X protocol. You'd never be able to achieve that kind of performance using a simple framebuffer-based remote display technology.

Re:OSX, Windows 95, Vista, Windows 7

[Knuckles]

Me, though, I'd like to know if this change will finally allow me to have use a compositing window manager without tearing (you know, like MacOS and Windows have been doing for years now) and without having to restrict myself to an ancient or gutless graphics card.

'the software's stated goal is "every frame is perfect, by which I mean that applications will be able to control the rendering enough that we'll never see tearing, lag, redrawing or flicker".'
    -- http://en.wikipedia.org/wiki/Wayland_(display_server)

Re:OSX, Windows 95, Vista, Windows 7

[Anonymous Coward]

Parent comment has not had enough love. Seriously.

Re:OSX, Windows 95, Vista, Windows 7

[Prien715]

...without having to restrict myself to an ancient or gutless graphics card.

The binary NVidia drivers work great for me on my new cards. Oh, you wanted Open Source drivers? Like Windows/Mac have had for years?;)

Re:OSX, Windows 95, Vista, Windows 7

[sarhjinian]

I'm actually using the binary nVidia driver and a (fairly) new card. If I use Compiz or KWin, I get tearing. It's gotten better, but it's still there. Intel cards work flawlessly.

Re:OSX, Windows 95, Vista, Windows 7

[staalmannen]

apple, Microsoft and Sun all have radically changed their widowing systems on many occasions while maintaining continuity for their developers. It did not mean no work, it just meant that recompiles could produce a functional product in most cases, albeit one that might look like poo and not have any of the new capabilities of the windowing system.

I find it somewhat hard to believe that the original design of X was so perfectly extendible that after decades of use it is not straining its seems.

So a change may be good.

However, i do see a downside. The nice thing about X unlike Windows and Macs main display interface is that it is more easily separated from the desktop. If you want to use a mac or windows system remotely you have to use something like VNC or a remote desktop app. In both cases you are getting the whole desktop not a display window. You can't run multiple instances of it. That's the main thing I like about X.

I would like a 9P2000 file system interface similar to plan9 /dev/draw, which could be used for totally network transparent sharing standard of Wayland and X11 or whatever windows over the network. http://doc.cat-v.org/plan_9/IWP9/2008/v9fb.pdf

Re:OSX, Windows 95, Vista, Windows 7

[Anonymous Coward]

The competition doesn't have to use a simple framebuffer-based remote display technology either - RDP can do that, Microsoft just hasn't figured out how they are going to monetize it.

Re:OSX, Windows 95, Vista, Windows 7

[Anonymous Coward]

And smarter people run XPRA, or X Persistent Remote Application. It lets you only forward the only applications you are running (not an entire 1920x1600 desktop), reconnect to lost or disconnected sessions, can still forward the desktop (using NX) if you want, lets you eavesdrop on other X clients (after you've authenticated), forwards audio if you want, allows you to send messages between X sessions, automatically discovers other XPRA servers on the network, and presents you with a menu of applications you can run on remote machines (and forward the display to your local desktop) once you have authenticated to other servers.

The catch? It needs a bit of work (integrating everything is done in Python, could use a bit of interface work, etc.) and FreeNX is based on Tcl/Tk, which has caused many beating of breasts and gnashing of teeth. I have found it useful though.

Re:OSX, Windows 95, Vista, Windows 7

[Anonymous Coward]

apple, Microsoft and Sun all have radically changed their widowing systems

Yeah .. they're just like .. killing them off .. for the .. pensions ..

Re:OSX, Windows 95, Vista, Windows 7

[daboochmeister]

> You have to install a fairly thick local client

Of course, that only applies if your client is not also running X (e.g., if your coming from a Windows PC). Otherwise, it's completely native.

Just clarifying for newbies, who will see your low id# and assume omniscience :-) ...

And, of course, downchain, there's mention of NX. (hmm ... pondering how NX would adapt to support Wayland)

Re:OSX, Windows 95, Vista, Windows 7

[Anonymous Coward]

^ This... just this.

I mean I've had years of wonderful desktop use from X but try to watch a video or do anything 3D and the lack of proper v-sync is horrific.

Most people don't seem to notice/care, but I do.

Typical Ubuntu

[Murdoch5]

Lets change what works and make it harder and more awkward to use.

Guess I'll keep on Loving Gentoo.

Re:Typical Ubuntu

[Murdoch5]

Got to love when people don't understand the rating systems. A troll isn't someone tell the truth, it's someone saying shit for the sake or shit.

Drivers?

[Anonymous Coward]

Has wayland re-implemented all the user-space portions of the graphics drivers? What about nvidia binary drivers? Unless wayland has some kind of compatibility layer for loading X drivers I don't see this transition happening anytime soon. And if it did have a compatibility layer for X drivers underneath in addition to the X protocol compatibility on top, then how much is really different than X?

Very skeptical.

[seeker_1us]

I looked at the description of the Wayland project. Granted, it's a couple years old, but it looks like it only works on a limited set of hardware.

They talk about using nouveau for NVIDIA cards... well that is really not an option for some people.

Ubuntu should NOT limit their hardware choices.

Re:Very skeptical.

[Narishma]

Don't trust anything you read on Phoronix. Go read directly from the source: http://wayland.freedesktop.org/architecture.html

Please retain network transparency

[Urban Garlic]

I am all in favor of a new and better graphical system, but for the love of God, PLEASE keep network transparency. I want to forward my graphical session to other hosts, and have windows from remote systems show up on (and be managed by) my local display. This is *essential* for some sysadmin tasks I have to do, on a remote system that *has* *no* *graphical* *console*, but for which some of the tools *require* a GUI. At the moment, the saving grace of this system is system is that I can ssh in, forward my X connection, and run GUI software remotely.

On a related note, I wish to inform the community at large and Ubuntu in particular that not everyone is using a personally-administered workstation with a local file system. Some of us NFS-mount our home directories from a central server, and some of us install software on application servers which are also NFS-mounted. Please take care that "new improved" installers and desktop systems do not break in this environment.

Thank you.

Re:Please retain network transparency

[Late Adopter]

On a related note, I wish to inform the community at large and Ubuntu in particular that not everyone is using a personally-administered workstation with a local file system.

Point taken, but Ubuntu is not for you then. They trade assumptions for simplicity, and their target market, like Apple's, is specifically people for whom those assumptions hold.

Re:Please retain network transparency

[Anonymous Coward]

you don't use ubuntu for those tasks

Re:Please retain network transparency

[MostAwesomeDude]

It was gone from day one. Just keep using X.

Re:Please retain network transparency

[Beelzebud]

So then use a distro well suited for what you want to use it for. Why try to impose X on the majority of Ubuntu users who will never use the network capabilities and don't really care about them?

Re:Please retain network transparency

[Anonymous Coward]

Why don't you just use a different distribution?

Re:Please retain network transparency

[Tailhook]

keep network transparency

GUI systems that were not designed for network transparency work very well on remote terminals. See RDP. I have used both extensively and Terminal Services is simply better, without qualification, for the administration use case. It also works extremely well for a large subset of desktop applications.

Welding network transparency into the heart of the GUI system imposes serialization complexity and overhead. The recent history of X development is a series of workarounds to overcome this. When the overhead, complexity and workarounds can't be suffered then X is abandoned; see Android, arguably the most successful Linux platform ever.

The Linux desktop would work better with a thin, efficient, reliable, kernel controlled display server for local GUI applications. An optimized screen scraper (ah la RDP) could provide more than sufficient GUI fidelity for administration. If the two were developed together such that the display server provided hints to optimize the screen scraper the result would be superior.

Re:Please retain network transparency

[Khashishi]

Why would you run Ubuntu on a headless remote system? I'm not sure why Ubuntu needs to cater to everyone. That's what other distributions are for.

Re:Please retain network transparency

[MikeBabcock]

Because their OS of choice is run by people with a lot of money who will invest in technologies that may take over the desktop on us and make our lives more difficult for how we use computers.

Re:Please retain network transparency

[couchslug]

"but for the love of God, PLEASE keep network transparency. "

If you know what that is, you don't particularly need Ubuntu.

There is a very nice, but nearly forgotten distro for power users:

http://www.debian.org/

Re:Please retain network transparency

[spitzak]

My understanding is that network transparency will continue to work in various ways. As accurately as I can see here is what will happen:

1. X11 applications will work just like now, they can directly talk to a remote X server, and a remote X11 application can talk to the local X server that is running atop Wayland.

2. Although it is likely that Qt and KDE, etc, will be rewritten to talk to Wayland directly, they will retain the ability to talk to X11 in their code. In the worst case this means you have to run a version of the program linked with this alternative version. Much more likely the toolkit will automatically use the X11 backend if it detects a remote X11 server is in $DISPLAY, similar to how OpenGL right now uses XGL only for remote displays and direct hardware manipulation for local. Though I would worry about bitrot making this ability fail after a while if nobody uses it. Another possibility is that changing LD_LIBRARY_PATH will cause an X11 version of the toolkits to be used instead of the Wayland version.

3. Wayland itself does not have to "composite" all the windows onto one local screen. I suspect there will be a protocol to send the window changes to a remote Wayland server and thus the window appears on a remote screen and input comes from the remote. This is pretty much what RDP does with official Microsoft clients.

4. Arbitrary programs can pretend to be a Wayland server and thus receive the above window changes. This is pretty much what all non-Windows RDP clients do, and it works well.

5. The composited result can be screen-scraped and sent over the network to a client that displays it.

Re:Please retain network transparency

[mebrahim]

Because X11 can be run over Wayland, ssh -X should continue working. So I'm not so worried about that.

Re:Please retain network transparency

[dch24]

But because Wayland is an API, it can be added as a plugin. It may not be fast or work well (at first) but it is still possible.

Re:Please retain network transparency

[evilviper]

Terminal Services is simply better, without qualification, for the administration use case.

Absolutely not. A desktop inside a window has always been bad design. You know what happens when you load up your desktop at 640x480? Icons scattered across your desktop. No sane way around that. And how about when you want to click on the very corner pixel on your screen, do you go with jailing the input and the associated hassle, or do you make accessing the edge of the virtual screen practically impossible?

"Rootless" is the only way to go. An application just pops up like magic and integrates with your current desktop. Wonderful if you frequently need to run GUI programs remotely. Hell, I can have an app from one server side by side with an app from another server... no bounding boxes to worry about. And the few hacks that can make something resembling this happen on Windows are a clumsy mess.

But perhaps more importantly is the design model the two methods impose.

When a Windows user sees some antivirus program in their system tray, they've actually got a system-level privileged application right there, sharing user-level memory space, exposed to their abuse, and easy to exploit to get root level access. With X11, that GUI interface to whatever is just an unprivileged module passing data back and forth between the system and the user level, it doesn't essentially need to operate with system rights in the user's memory space to work, as Windows' model basically imposes. This is just one more place where Windows is inherently inferior and insecure.

The recent history of X development is a series of workarounds to overcome this.

The development history of any platform is a series of changes to cut down on the impact of inherent design limitations.

When the overhead, complexity and workarounds can't be suffered then X is abandoned;

X11 worked well on devices less powerful than the cheapest phone you can find today. It runs fine on 26MHz ARM CPUs with severely constrained RAM when xterm is considered a memory hog (~2MB): (http://www.openpsion.org/howtos/series5mx_new/x628.htm)

see Android, arguably the most successful Linux platform ever.

It's a somewhat unfortunate side effect of the openness of open source, that everyone feels compelled to "roll (their) own" at the drop of a hat. Back when the first PDAs were coming out, there were a dozen contenders to replace X11. Damn near all failed quickly. It seems everyone is doomed to repeat the same mistakes, and misunderstanding that most of the overhead of X11 is necessary overhead, in the same way that while a multi-user, multi-tasking operating system has overhead, it is very much NECESSARY overhead.

Sure, invent your new display technology. It will either perform terribly or be terribly hardware-specific and fall apart when the next generation of video chips come out that deprecate some feature you used. It will either be extremely good at a specific use case, and extremely poor at all others, or will simply be mediocre all-around.

Would you care to guess why the ubiquitous framebuffer hasn't killed off X11?

It's not to say X11 is perfect, but it's pretty damn good and shows just how incredibly flexible it is to be a top competitor after all these years, and there simply isn't enough room for performance improvement to make a change worthwhile for anybody.

Re:Please retain network transparency

[Professional Slacker]

I have looked forward with dread to the day I'm going to have to fall back to Debian proper because of some stunt pulled by Ubuntu. I've run Debian Stable on a number of servers, and like Stable, that said on the desktop I can't see Stable working to well on the desktop. How does Sid stack up to Gentoo in terms of breaking all the damn time and requiring maintenance, hacks, and demonic sacrifices to keep working day to day? Would Testing with an absurd amount of apt pinning be better?

Also is there some obvious replacement to the PPA system/plethora of third party repositories for running XYZ random application or Foo version 0.99-svn-rightnow-so_new_it_hurts?

Re:we'll be able to retain the ability to run X ap

[commodore64_love]

(Joe Q Public) - "Not a clue what that buddy's talking about. WM? I'll just windows like everyone else."

The trademark owner defines it

[tepples]

OSX isn't Unix. "UNIX certified" is not UNIX.

UNIX is a registered trademark, and under United States law, the owner of the trademark gets the first crack at establishing a definition. How were you defining UNIX? Derivative work of the Bell Labs source code?

Re:The trademark owner defines it

[Anonymous Coward]

If you define unix as derivative work from Bell Labs unix then almost all BSDs get included by definition.

By that same definition linux is not included.

If you look at any unix family tree you will see this is a very common criteria and the BSDs(including OSX) get a direct line descendency from Bell UNIX, while linux does not.

Re:And thus the curse of Open Source manifests its

[russotto]

And thus the curse of Open Source manifests itself - Develop something to to the virge of usability and robustness...then BAM! "Fuggit" let's start again in a new direction and it will be "better" and spend more years in development wilderness.

In the time X11 has been around, Apple has switched processor families twice and gone through two rewrites of the operating system (System 7 and OS X). Microsoft has gone from Windows 2 to Windows 3.1 (16 bit!) to Windows 95 to the NT based versions. Sun has gone from SunOS 4 to Solaris 2 and ceased to exist.

And you think starting in a new direction is an Open Source curse?

Re:And thus the curse of Open Source manifests its

[am 2k]

"Welcome to the world of Linux and Open Source where everything is in a perpetual state of development and a finished release is just a pipe-dream"

That's what you get when you let programmers to the managing and leave out the money motivation. Why should they ever release a finished product? That's just boring grindwork, better add features, or scrap the whole thing and start again in a better way.

Finally

[Zo0ok]

Finally this happens... just a few years late. Apple got it right with Mac OS X - what has the rest of the *NIX community been doing all years when the solution was right in front of their nose? And since Mac OS X runs X applications properly, there is no worry this can not be done. It works today and it has been done.

I admit I have not contributed at all to this move either - but I welcome it now when I hear about it!

Application forwarding? This is not about killing the possibility to write applications for X, and forward them over X, if needed. This is about all graphical applications should not have to depend on TCP/IP.

Re:Finally

[GioMac]

>Mac OS X runs X applications properly
Funny, isn't it? No, it doesn't.

Re:Finally

[Patrice]

This is not about killing the possibility to write applications for X, and forward them over X, if needed. This is about all graphical applications should not have to depend on TCP/IP.

Well, what about running the regular apps installed on the computer I'm remotely connected to ? What if I want to run just any application remotely from my server to my netbook, will I need to keep two different versions of my programs just to be able to run them remotely ? I hope not.

What needs to be provided is, given computer A & computer B both installed with a plain Ubuntu distribution (desktop and/or server), I can do "ssh -X" (or "ssh --Wayland", I don't really care), and get any app running on A being displayed on B. If GTK/Qt makes it transparently using X instead of Wayland direct rendering, fine for me. If I need to install special packages (e.g. "firefox-x11") and run special versions of apps, then it'll be a serious regression, and I'd be looking seriously at alternatives (probably a X11-buntu sub-distribution ;) ).

Re:Finally

[Jonner]

Apple has not gotten it right. I am running several X11 apps on OSX right now and they are second class citizens. Apple has made it very clear that there's only one right way to do things (use their proprietary protocol and libraries) and they don't let you forget it when you run X11 apps. Also, various X11-based desktops have very similar capabilities to Aqua in desktop effects, so dumping X11 is clearly not necessary.

X11 never has and never will depend on TCP/IP. It only depends on some kind of reliable bi-directional byte stream. The common case is a local (Unix domain) socket in the file system, which is very efficient. In that common local case, further optimizations are routinely used via the shared memory X11 extension and graphically intense apps that do things such as play video or display OpenGL graphics usually use direct rendering, which doesn't go through the X11 protocol stream and therefore wouldn't be improved by a switch to Wayland. The proponents of Wayland contend that certain optimizations related to desktop compositing are difficult or impossible when using X11, which seems plausible, but I don't think it's clear that it offers a giant improvement in performance.

Re:Finally

[Rakarra]

And since Mac OS X runs X applications properly, there is no worry this can not be done. It works today and it has been done.

Err, no it doesn't. It doesn't even come close.

Some servers need X; others don't

[tepples]

I was under the impression that having X on some servers is a waste of resources.

I'll grant that this is true of some servers, but X11 is handy on servers in a lot of other cases. Servers that host applications used remotely over X11 over SSH need the X11 client library. This could include servers with a tool to edit the configuration in a structured manner. Still other servers need an X11 server because they are also occasionally clients: either the administrator can configure certain services using the structured tool before connecting the machine to the network, or it's a development workstation hosting a copy of the server environment.

Re:Some servers need X; others don't

[underqualified]

something you wouldn't want your sysad to say... "where's the GUI on this thing?"

Ubuntu is systematically ruining their distrobutio

[Anonymous Coward]

come back to earth. You guys have been making some absolutely STUPID decisions as to the direction you are trying to go. I will be moving away from ubuntu because it is getting worse and worse. Wake up, you can't even call your crap linux anymore

Toolkits that wrap X

[tepples]

I see plenty of people who depend on X being X, and plenty of people who are being advised to depend on X being X. A move to Wayland will create all kind of confusion for those people

Applications are typically not coded directly to X11; they're coded to toolkits that wrap X11. GTK+, Qt, and GNUstep could easily be ported to wrap Wayland, just as GTK+ and Qt have been ported to wrap GDI on Windows. In addition, X11 can run on top of Wayland, as one of the articles points out, much like X11 on Mac OS X runs on top of Quartz.

Re:And thus the curse of Open Source manifests its

[Anonymous Coward]

"Welcome to the world of Linux and Open Source where everything is in a perpetual state of development and a finished release is just a pipe-dream"

Actually this problem exists in all software. No piece of software can ever be completely "done" and meet everyone's needs. The nice thing about Open Source software is that it is possible to modify the program to fit an individual's/group's/company's needs, which is not possible with proprietary software under normal conditions.

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

MY PS3 doesn't run a whole host of PS1 or PS2 games even though Sony claimed it would.

My PS3 plays ever PS1/PS2 game I've ever thrown at it. The only one I couldn't get to work was Guitar Hero, and that was only because I couldn't get the controller hooked up properly. It sounds like you bought one of the later models using software emulation.

Windows Vista and Seven doesn't run old 3.1 or 95/98 apps.

The 32-bit versions of those operating systems will run old 16-bit applications just fine in compatibility mode.

Mac OS X doesn't do well with classic PPC or 68000 apps.

OS X RUNS on PPC systems, and that was the ONLY thing it ran on for the first several years of its existence. Did you actually mean pre-OSX apps?

Of course on the other hand, Ubuntu already has a poor track record of this, having switched from ALSA to Pulseaudio. Pulse was supposed to have an ALSA compatibility layer, but its full of no-ops that cause applications using it all sorts of problems. On that note, they switched from an audio interface that required direct access to an audio server that allows networked playback. Now they're switching from a display server that allowed networked graphics to an interface that requires direct access. Go figure...

Re:And thus the curse of Open Source manifests its

[devent]

And what that have to do with Open Source? You must be new to I.T. because this happens all the time. Recently Microsoft stated that Silverlight will be more for the Windows Phone and developers should focusing more on HTML5. With Open Source it's more transparent so maybe you have the impression that it's happening more often.

If Wayland will be good, then more distributions will use it but someone have to start to test it otherwise we never know if Wayland is better or not.

"Welcome to the world of Software and I.T. where everything is in a perpetual state of development and a finished release is just a pipe-dream"

Re:And thus the curse of Open Source manifests its

[MonsterTrimble]

Um, what? How is this insightful?

How much do we bitch about Debian's "Not released until it's ready" credo? How many distros didn't jump ship to KDE4 until 4.2 came out? Gnome has been at 2.x for years and it doesn't get much more stable then that. Also, closed source is no different - witness Microsoft's complete jump in GUI from 2000 to XP to Vista. The only difference is that with open source we get to see the new stuff go from Alpha to Beta to released. We only see closed source when it's released.

Windows XP and 3.1 run inside Windows 7

[tepples]

Windows Vista and Seven doesn't run old 3.1 or 95/98 apps.

Windows 7 Pro comes with an included copy of Windows XP to run in a virtual machine, and an old copy of Windows 3.1 from an eBay or CL seller will run in the DOSBox emulator. Yes, you lose some performance to virtual machine overhead, but how fast was a typical Windows 3.1 or Windows 98 box anyway? And just as you can install these into Windows 7, and just as you can install X11-on-Quartz into Mac OS X or X11-on-GDI into Windows, you'll still be able to install X11-on-Wayland into Ubuntu.

Re:Windows XP and 3.1 run inside Windows 7

[jabelli]

Actually you have to download it from MS. Also you can use Virtual PC for free on Win 7 Home, you would just need an extra XP license and to install your own VM; then you can install the integration components and have the same experience. The problem is you have to dig to find the WVPC-only download.

Dos 6.22/Win 3.11 will also work on Windows Virtual PC in Win 7. I just copied over my VPC 2007 VM and WVPC upgraded it as soon as I opened it.

what problem is this solving?

[t2t10]

The X server is an efficient, widely used, and standard graphics platform. And the X server is where all the drivers are being developed. Finally, the X server actually gets things like multi-monitor support, resolution switching, etc. right.

Why go now with something totally new and largely unproven? What problem is this solving?

Re:what problem is this solving?

[Tetsujin]

The X server is an efficient, widely used, and standard graphics platform. And the X server is where all the drivers are being developed. Finally, the X server actually gets things like multi-monitor support, resolution switching, etc. right.

Why go now with something totally new and largely unproven? What problem is this solving?

The basic "problem" that's existed in X forever is that it's network-transparent first and local-optimized second, via extensions.

Whether this is a "problem" is largely a matter of perspective. Obviously many people, myself included, consider it an asset.

However, a lot of Linux usage is moving away from this scenario. Or perhaps it's better to say, this scenario is no longer considered the primary use case. Even software implemented on top of X is often written without remote operation in mind: the result is software that could be run remotely via X, but which in practice is too network-intensive to be useful in this mode. And, of course, there are some programs that simply don't make sense to run via remote display.

The problem this solves is X, basically. Most of what X has to offer has become extraneous for many users. They don't use remote display, very little of the software they use in practice is really implemented in a way that makes it viable over a remote display anyway, and they want to run a lot of software that really does need a short path to the display hardware in order to run properly. So the idea is to "cut out the middleman", trim a couple decades of legacy cruft.

We've heard this story before, of course. It's not the first time people have tried to unseat X as the dominant display system on the Linux desktop. Still, despite my attachment to X, I'm not convinced that getting rid of it would be a bad thing. It seems to me that most software written to run on it today isn't really written to work well on remote display.

Re:what problem is this solving?

[t2t10]

The basic "problem" that's existed in X forever is that it's network-transparent first and local-optimized second, via extensions. Whether this is a "problem" is largely a matter of perspective. Obviously many people, myself included, consider it an asset. [...] So the idea is to "cut out the middleman", trim a couple decades of legacy cruft.

Windows has a lot of legacy cruft in its graphics system, because it moved from a frame buffer model with synchronous calls to a client/server model. OS X has a lot of legacy cruft in its graphics system because it moved from a DisplayPostscript model to its current model.

X11, instead, uses an efficient binary encoding and an asynchronous communications protocol, plus optional shared memory optimizations. It is what a modern desktop window system should be and how it should communicate. X11 happens to also support remote displaying of windows, but you pretty much get that for free.

There is some legacy cruft in X11 related to fonts, 2D graphics, and resources, but that code is pretty harmless and it's actually still used by many apps.

"X11 is inefficient because it has a lot of legacy cruft due to having to support network transparency" sounds plausible, but it is just totally wrong.

Re:what problem is this solving?

[Tetsujin]

Well, admittedly there are bits of the whole situation I don't understand fully enough. But I am trying to make sense of this situation with Wayland... And in the process, learn a thing or three.

I know there's kind of a knee-jerk reaction to assume X11 should be done away with simply because it's old. (Or more precisely, that it has a long history...) I try not to make this mistake myself - but with the things people are trying to accomplish on Linux desktops these days (i.e. slick effects, and with no particular demand for network transparency as far as most users are concerned) I can see why people might feel X isn't necessarily the right choice. If the code you're writing is fundamentally not well-suited to working over the network (for instance, animated transitions) - then X has little to offer over a simplified local-display system.

I also think the way they're expecting Wayland to be used in Ubuntu - as a layer under X or other remote display technologies - could make a lot of sense. It seems more sensible to start with a layer that emphasizes performance and implement network transparency on top of that rather than the other way around.

What do you know about the vblank syncing issue that the Wayland FAQ cites? Is there really no way to sync display changes to vblank on X? Or is something like that offered in, say, the compositing extension, or GL?

Re:what problem is this solving?

[t2t10]

It seems more sensible to start with a layer that emphasizes performance and implement network transparency on top of that rather than the other way around.

All major window systems these days are architected the same: they are all user-mode client/server systems. X11 was designed that way from the ground up, while Windows and OS X were retrofitted (in different ways) and don't do it as well. X11 would be designed largely the same way if you assumed no network transparency; all the hard work is in the client/server aspect.

The features that make X11 network transparent have nothing to do with its graphics protocols, but with features such as window management and resources. They don't cost performance, but they have proven extremely useful even for local apps. For example, window decorations are consistent and functional no matter what toolkit you use. Wayland is trying to push window decorations into toolkit libraries, which is the same stupid crap Windows users have had to suffer under for years.

What do you know about the vblank syncing issue that the Wayland FAQ cites? Is there really no way to sync display changes to vblank on X? Or is something like that offered in, say, the compositing extension, or GL?

There's the SYNC extension (pretty standard). What else do you need?

Re:what problem is this solving?

[Tetsujin]

What do you know about the vblank syncing issue that the Wayland FAQ cites? Is there really no way to sync display changes to vblank on X? Or is something like that offered in, say, the compositing extension, or GL?

There's the SYNC extension (pretty standard). What else do you need?

I don't know. That's why I asked! :)

Re:what problem is this solving?

[t2t10]

Well, SYNC has been a pretty standard component of X.org for a long time, so I wasn't sure whether you didn't know about it or whether you wanted something else.

Re:what problem is this solving?

[Tetsujin]

Well, SYNC has been a pretty standard component of X.org for a long time, so I wasn't sure whether you didn't know about it or whether you wanted something else.

Didn't know about it - and from searching online I saw no mention of it being used to sync to vblank... Though I suppose if that's one of the things the extension is for then that must be one of the events you can sync to.

Re:what problem is this solving?

[t2t10]

Note that there's also the double buffer extension, which used to be frequently used with the synchronization extension. The shared memory extension was also around very early. X11 has pretty much everything a modern window system needs for more than two decades.

(I think modern monitor technologies make this stuff increasingly irrelevant.)

Re:And thus the curse of Open Source manifests its

[Alkonaut]

There is another problem: one is creating a free platform to solve the problems that proprietary software does at a cost, or with a lock in. The other problem is making this platform interesting enough to use, and interesting enough to develop (and develop FOR) so that it keeps its momentum (or gains a critical mass, as is the case with desktop linux). Understandably, these targets are somewhat in conflict.

Wine is just as native

[tepples]

(or maybe who cares where there are no AAA native games for Linux)

As long as your PC has an x86 CPU, and as long as the developer tested the application on Wine, Wine can be just as "native" as toolkits more traditionally associated with GNU/Linux environments. An app that runs on top of the Wine toolkit is just as native as an app that runs on top of the wxWidgets or Qt toolkit. A game that runs on top of the Wine toolkit is just as native as a game that runs on top of the SDL or Allegro toolkit. Is your rant directed at the different executable format that apps made for Wine use (PE instead of ELF) or at developers of apps for Windows who refuse to test them on Wine?

What He actually said:

[Dcnjoe60]

For anyone who is interested, here is what Mark Shuttleworth actually said: http://www.markshuttleworth.com/archives/551 . In his post he gives his reasoning and alternatives they looked at. Seems pretty well thought out. Ubuntu always gets slapped about not giving back to the community. Well, here they are announcing they are giving back and they still get slapped. It seems as if they are damned if they do and damned if they don't.

Re:What He actually said:

[jav1231]

They're not damned until RMS damns them. :p

Re:And thus the curse of Open Source manifests its

[Anonymous Coward]

Totally agree. That is why I use Windows, errr, I mean Vista, wait, no I went to that newer version. Fuck it. I am getting a mac. They would never switch anything like that.

Why does a user care what GNOME is running over

[DrXym]

GTK is cross platform and I assume GNOME could be too with the proper abstractions. At the end of the day why should GNOME app developers or users really care what windowing technology is underneath powering their desktop? Apps just call the APIs and provided the port maps them onto the appropriate native equivalent, things just work.

Obviously not every GNOME app is "pure" and I'm sure there are plenty of hacks to work out. But at the end of the day QT devs don't really have to care too much which OS or platform they're targetting and I don't see the situation being that different for GNOME either. We already have apps like GIMP ported to Windows so clearly the concept is already much the way there.

The question is whether Linux would be better off doing away with X11 in the long run. Fundamentally why does Linux need to stick with using X11? After all, if you need X11 you could always run it as a server over wayland (this is how X11 runs on OS X or Windows). And if you don't need X11, then the system benefits from a more streamlined graphical layer with all the compositing done in the display driver next to the kernel rather than being farmed out by X to an extension and back and forth several times with various hacks.

Re:Why does a user care what GNOME is running over

[0123456]

Fundamentally why does Linux need to stick with using X11? After all, if you need X11 you could always run it as a server over wayland (this is how X11 runs on OS X or Windows).

Because many of us routinely run X11 programs on one system with the display on another? Switch to Wayland and you lose that capability, which is a massive benefit to professional Linux users.

And if you don't need X11, then the system benefits from a more streamlined graphical layer with all the compositing done in the display driver next to the kernel rather than being farmed out by X to an extension and back and forth several times with various hacks.

X11 is probably more efficient on a modern multi-core system since the rendering and display logic run in different processes.

Re:Why does a user care what GNOME is running over

[DrXym]

Because many of us routinely run X11 programs on one system with the display on another? Switch to Wayland and you lose that capability, which is a massive benefit to professional Linux users.

A X server can run over wayland. It's not hard either to envisage that GNOME dynamically chooses the right backend to use dependent on the circumstance.

X11 is probably more efficient on a modern multi-core system since the rendering and display logic run in different processes.

Excessive context switches and an event driven GUI don't necessarily mix. Sometimes you want less processes, and if concurrency is desirable in cases then they could be done with threads in the same process. Anecdotally with similar drivers (e.g. commercial NVidia) I've never thought to myself, damn Linux is faster than Windows. Usually its the exact opposite and that's when Windows is running Aero compared to a vanilla GNOME with minimal effects. The only time when the systems get closer parity is for games and such like which cut a hole through X with DRI and more or less hit the hardware directly.

Re:Why does a user care what GNOME is running over

[ADRA]

When running and debugging my code, my Linux PC (same hardware) as my Windows XP based coworkers would always be faster to use. I can't say why, but that's the way it was. It was a painful experience to sit at a windows dev's desk to wait a minute for things to compile when it took me, oh 10 sec max.

Re:Why does a user care what GNOME is running over

[Anonymous Coward]

Maybe the program you used to compile with on Windows was just poorly done?

Re:Why does a user care what GNOME is running over

[ADRA]

Oh, forgot to mention it was Eclipse, running JDT's built-in compiler. Nobody (myself included) made any gigantic changes to the eclipse setup from stock that would affect performance dramatically. When I was developing, it was always on the latest Fedora.

you don't know what you're talking about

[t2t10]

This is not about killing the possibility to write applications for X, and forward them over X, if needed. This is about all graphical applications should not have to depend on TCP/IP.

X11 uses shared memory and local IPC for local applications, same as OS X.

Apple got it right with Mac OS X - what has the rest of the *NIX community been doing all years when the solution was right in front of their nose?

Apple's graphics subsystem is a derivative of DisplayPostscript from NeXT, an old, slow, and bloated technology. Other UNIX vendors tried shipping it and it was a complete failure. Apple managed to hack it so that it performs halfway decently in OS X, but it is still worse than X11.

And since Mac OS X runs X applications properly, there is no worry this can not be done. It works today and it has been done.

Desktop integration of X11 apps on OS X is essentially non-existent; not even key maps work properly, and X11 on OS X is dog slow.

What users want is irrelevant.

[Anonymous Coward]

This was why I left Ubuntu. I didn't want PulseAudio by default. I wanted to be able to uninstall it to get better performance in my games with lower CPU usage. They built the system so the volume control depends on PA, so in order to purge it, you have to use a custom PPA to get re-compiled volume controls. Then there was the whole button placement thing.

Now we have this. I don't think users will be happy running their traditional X apps in a compatibility layer, unless it is 100% perfect (which I doubt). We've already got WINE, which is a compatibility layer in its self. Now I know anyone can fork Ubuntu and restore the previous choices in software, but that isn't the point. Ubuntu is the distro with the most well-known name. Many people who don't know about anything other than Windows know about Ubuntu, but no other distros. This means they should be doing their very best to provide a stable system with as few hassles as possible for the users. Nobody wants to add pasuspender to individual programs to get them to work correctly, or have to adjust the mind-numming window controls because they must use multiple OSs, or deal with whatever compatibility issues will surely result from the drop of X. We want things to perform well and in a consistant manor. I've switched to Debian because of all of this.

Re:And thus the curse of Open Source manifests its

[Bill_the_Engineer]

The Ubuntu distro is showing great maturity and evolving nicely....then "Fuggit!" let's go a different way and start the integration again!

Some people will question the maturity of Ubuntu. I like it's much larger support lifespan when compared to Fedora, but Ubuntu has been making some architectural changes between releases that has broken some of my packages.

Anyway, I think this has more to do with appearances than anything else. Ubuntu has been getting a lot of heat lately about their lack of participation in maintaining the Linux architecture outside their own distributions. So Mark Shuttleworth's ego is a little bruised. Therefore they are picking pet projects to promote in order to give the appearance that Ubuntu is giving something back. I don't think the fact that both Unity and Wayland is immediately apparent to the end user is just coincidental.

XMonad

[Anonymous Coward]

the more important question: how long will it take to port a window manager to this.

I am thinking about xmonad or similar.
since the "window manager" will then be part of the whole binary.

Good bye effective multi-user graphical computing

[PotatoHead]

X has a LOT of power, but remains poorly understood.

Now we are going to get applications targeted to a single user GUI, and with that all the problems the other OSes have with their single user GUI. :(

Blah

[likuidkewl]

And to think it all started with moving the buttons to the left....

Headline wrong?

[FunkyELF]

Shouldn't the headline read "Ubuntu Dumps X For Wayland On Unity"?

Re:Headline wrong?

[jav1231]

Which is ironic, when you think about it.

Re:we'll be able to retain the ability to run X ap

[JackieBrown]

I don't know. Ubuntu has managed to get KDE and Gnome to adjust their release cycles to every 6 months. I am guessing that Ubuntu has a lot of sway

Linux Mint ...

[scharkalvin]

This kind of stuff will probably accelerate Linux Mint toward their Debian based distro. Right now I'm running Linux Mint Debian on my thinkpad, if they come out with a 64 bit version I'll migrate my desktop to that unless I install Debian Testing first.

Re:And thus the curse of Open Source manifests its

[Anonymous Coward]

perfection is a moving target and good enough does'nt last.

Won't matter.

[MarcQuadra]

X programs on Ubuntu clients and servers will still have to run in X, but X will be running inside Wayland. Think of it like running Xming inside Windows. I can still 'host' and 'join' X sessions to and from it.

I think this might actually be a good move, X is overkill for 98% of users. A smaller, slimmer, faster-moving display foundation that still allows the 'legacy' X stuff to run in a runtime would be great.

X is lightweight, that's the problem

[Burz]

A desktop system needs something more robust, with internal protocols that anticipate the core use cases for a consumer / end user. X is and always was for use by technical staff on managed workstations. It 'beaks' easily, by which I mean that configurations are very fragile and often result in non-working displays because something is a little-bit off. The Unix model of vi-editable ASCII config file (with N possible layouts) creates some of the fragility for X, when something like XML is more appropriate. Further, a subsystem like X should have an API not only for changing res and other params during display operation, but also for SAVING those settings: A big problem with X is that is never managed its own da^# config files.

X is also missing some modern features that would allow easy sharing of apps and desktops *efficiently* over the Internet. I'm tired of seeing add-ons line NX popping into slow bitmap-tossing (VNC) mode whenever more than one instance of a window is involved. X's architecture prevents some very important end-user needs from being effectively fulfilled. Windows and Mac currently have X beaten to a pulp and please don't go into how X inventing remote displays makes it superior because its inferior and insular. If the X part of the ecosystem were healthy someone prominent would have already found a way to make NX functionality standard and given people a way to graphically start remote apps too.

Re:X is lightweight, that's the problem

[aardvarkjoe]

The Unix model of vi-editable ASCII config file (with N possible layouts) creates some of the fragility for X, when something like XML is more appropriate.

What do you mean by this? Do you mean that X is fragile because it's hard to parse the ASCII config file or correctly handle the configuration options? (If so, I'd like to see some evidence, since I don't think I've ever seen a problem with X caused by that.) Or do you mean that it can be hard for a user to create a correct config file? I can see that, but it's only a factor if your users are actually creating a config file by hand. If it's not necessary to have a configuration built by hand, then you might as well just create a decent GUI for configuring X. (I think they exist, but honestly I've never had any trouble just editing the config file.)

Re:X is lightweight, that's the problem

[Anonymous Coward]

[citation needed]

I've been using X11 with a window manager as my desktop for decades. I've hardly ever had a "vi-editable ASCII config file" break by itself. Usually there was a person involved. Oh, and XML blows, and is not any less fragile when attacked by vi.

Re:X is lightweight, that's the problem

[Burz]

What do you mean by this? Do you mean that X is fragile because it's hard to parse the ASCII config file or correctly handle the configuration options?

It should be obvious what I mean. There is no cardinal way to lay out the ASCII text of xorg.conf. If a sysadmin or a 3rd party utility needs to make changes, the distro-supplied settings tool will often not know how to deal with it and try to replace the file with default settings. Further, updates that require tweaks to xorg.conf will tend to leave a custom-edited config in a mangled state.

If everything goes through X's (non-existent) API for changing and saving, then these conflicts don't arise. Also, if the file is stored as an XML schema, even manual changes should avoid conflicts.

There shouldn't be a million ways to create a correct X config file, and a utility shouldn't have to include a whole copy of X's own parser just to avoid conflicts.

Re:X is lightweight, that's the problem

[Burz]

[anecdote is not data]

And I suggest you consider finding a good XML editor.

Re:X is lightweight, that's the problem

[AJWM]

Who the hell hand edits X config files any more? I've used X since X-10 and I sure as hell don't. Modern systems -- at least, the commerical Linux distros -- do a fine job of autodetecting the hardware and providing a GUI for configuration.

Changing and saving configs shouldn't be part of the X protocol any more than tweaking your browser settings is part of HTTP or HTML. That's up to the particular X server implementation, of which there are more than just x.org's and xfree86's.

Re:X is lightweight, that's the problem

[Burz]

Who the hell hand edits X config files any more?

Its all over the help forums of distros like Ubuntu and others.

As for your confused bit about X protocol vs various implementations, TFA *is* about Ubuntu you know... which is a specific OS that's trying to keep people's screens from going dark.

Where other capabilities like screen/window sharing are concerned, I do take issue with the X protocol... which doesn't allow this very common use case (though may be rare with old people named 'Alastair'). And I do not consider VNC (a 1980s bitmap-only protocol) a competitive means to accomplish sharing.

Re:X is lightweight, that's the problem

[jbolden]

No they don't. I had a with Puppy Linux on a Macbook, wouldn't detect the mouse properly.

As for config files the X Server should include an X configuration tool that can work with local hardware, and make adjustments and...

Re:And thus the curse of Open Source manifests its

[Anonymous Coward]

And thus the curse of Open Source manifests itself - Develop something to to the virge of usability and robustness...then BAM! "Fuggit" let's start again in a new direction and it will be "better" and spend more years in development wilderness.

In the time X11 has been around, Apple has switched processor families twice and gone through two rewrites of the operating system (System 7 and OS X). Microsoft has gone from Windows 2 to Windows 3.1 (16 bit!) to Windows 95 to the NT based versions. Sun has gone from SunOS 4 to Solaris 2 and ceased to exist.

And you think starting in a new direction is an Open Source curse?

Um, for Windows and OS X, the changes have been an improvement. Windows 7 is finally nearly usable, and OS X can run nifty Mac apps as well as many of the open source apps that Linux has. Switching from X willl just be adding a layer, as every major app will stay running under X or have at most a half-working port, while developers work frantically on... terminals and soundcard controls.

(I swear I didn't google to see what is natively available for Wayland, so if I'm way off I will be pleasantly surprised)

LOL

[Anonymous Coward]

I've used Linux off and on before it was even a 1.0 program in the 90s (when you had to compile it yourself!). I even emailed Linus support questions (which he responded to almost immediately). I'd love for it to become a mainstream consumer OS (not just something hidden in a set top box or hidden deep under covers in a cell phone) but it will never get there as long as these stupid fights between UI developers (KDE, GNOME, etc.) continue.

And yet you wonder why people choose Macs?! Even more funny is when you call Mac users 'smug' but can't see past your own bullshit.

X is Ok for beginners?

[Burz]

Hey you BUTU, did you ever ask yourself why a *nix user has to go to a CLI to start graphical apps? Or why those remote apps are slower than their Mac and Windows counterparts? Why those users can share multiple copies of their windows while you can't do the same without something like VNC (which is primitive and slow)?

Good!

[Beelzebud]

It's time to retire the dinosaur. X was designed in the late 70's. I KNOW we could do better with something designed around modern technology.

PulseAudio and Volume Control

[viralMeme]

"This was why I left Ubuntu. I didn't want PulseAudio by default. I wanted to be able to uninstall it to get better performance in my games with lower CPU usage. They built the system so the volume control depends on PA, so in order to purge it, you have to use a custom PPA to get re-compiled volume controls. Then there was the whole button placement thing"

I'm using Lubuntu and the Volume Control applet in LXPanel seems to do the job without PulseAudio.

Who has graphics problems, anyway?

[Anonymous Coward]

"Graphics problems"? I haven't had those since I switched to a GPU company who is actually competent (NVIDIA). I'm not a paid shill or anything, I've just been around the block with the competitors' products and come to the realization that they are all garbage. One company drops driver support in under three years, forcing you to use the "open" and pathetically slow driver, and the other offers 2001-era hardware with equally slow drivers under Linux. The netbook struggles to play Quake 3 in Linux but it runs fast and smooth in Windows. Inexcusable.

So, what would happen if they go through with this switch? Will my NVIDIA drivers quit working?

This is so 2001...

[Anne Honime]

All this fuss reminds me of how GtkFB was supposed to storm the handheld realm circa 2001. Never happened, never will.

Truth is, for all its assumed weaknesses, X is still the more polished, thought-out graphical protocol in widespread use. It's there, it works, it boasts features (network transparency for instance) no single other GUI system can provide. There may be more eyecandy in windows©® or MacOS©® (and then, it remains to be proven, when I can run 'aero like' effects under X on Intel hardware not cutting the set minimum for windows), but when it comes to sheer possibilities, X beats everything else hands down.

And X has an enormous advantage : it works NOW, not when pigs will fly.

Re:This is so 2001...

[Beelzebud]

Good thing they aren't replacing it NOW... They're talking a long term strategy here, and if you RTFA he says this won't happen for maybe 4 years or so.

Re:This is so 2001...

[Anne Honime]

You're mistakenly assuming X will stay the same during that time frame while Wayland will be evolving. But this is wrong : X evolves at quick pace too, and Wayland will certainly play catch up and fail. The basis of X are well known, well understood, and quite solid. There is already an ungodly investment of man-hours in it, and many people pushing the project forward. I can't see Wayland gathering enough followers in 4 years to make up for the head start of X.

Re:This is so 2001...

[Beelzebud]

It just depends on who is working on it. Apple replaced X is less time than 4 years.

Re:This is so 2001...

[Tetsujin]

All this fuss reminds me of how GtkFB was supposed to storm the handheld realm circa 2001. Never happened, never will.

Truth is, for all its assumed weaknesses, X is still the more polished, thought-out graphical protocol in widespread use. It's there, it works, it boasts features (network transparency for instance) no single other GUI system can provide. There may be more eyecandy in windows©® or MacOS©® (and then, it remains to be proven, when I can run 'aero like' effects under X on Intel hardware not cutting the set minimum for windows), but when it comes to sheer possibilities, X beats everything else hands down.

And X has an enormous advantage : it works NOW, not when pigs will fly.

Some of this I heartily agree with - and some of it feels like the kind of attitude that gets people locked into bad design decisions with decades of backward-compatibility issues.

It's hard for me to contemplate Linux without X. The network transparency is something I will miss if we get rid of it. At the same time, though, I feel like it's become a fairly minor piece of how I use the system. A lot of the apps I run just aren't worth running over a network. I doubt I'd ever run Firefox or Blender remotely, for instance. If I did, my usual "remote" display server (my cheap-ass netbook) has a lot less graphical rendering power than my application server (my desktop PC) - so X doesn't bring a lot to the table. GLX probably isn't going to serve me as well in that case as a network transparency layer like VNC. I believe a lot of the modern software written to run in X isn't really written (or tested) to work well over the network - if people aren't writing the software itself to work well over a network, then what use is a network-transparent display system?

Re:This is so 2001...

[Anonymous Coward]

X needs replacement, and Wayland is its successor. Accept it, you can't stop progress.

Re:This is so 2001...

[Anne Honime]

I'm all for progress. Today's X is vastly more usable and user friendly than XFree86 I used back in 1998 ; but it does it without giving up what was good then.

Progress is not fashion. You don't have to trash last week's trend because it's old.

Progress can be measured in usability and functionality. Until Wayland proves itself worthy of taking over X, I see no need to fuss about it, and I'm betting X can improve (and therefore progress) faster than wayland in the same amount of time.

Re:we'll be able to retain the ability to run X ap

[icebraining]

Support for the PowerPC platform was dropped after Mac OS X 10.5. 10.6 and above require an Intel CPU.

... fuck it.

[Arancaytar]

I've used Ubuntu 8.04, 8.10, 9.04, 9.10, 10.04 and now 10.10. Go ahead with this and I'll be installing SuSE.

Re:we'll be able to retain the ability to run X ap

[Anonymous Coward]

Windows Vista and Seven doesn't run old 3.1 or 95/98 apps.

Aside from this blurb, I mostly agree. I've had no issues running Win98 and Win95 apps in Win7... Can you give me an example Win3.1 app?

fuuu

[kallisti5]

The first time I ' ssh -X ' and have it not work I will be seeking new desktop choices. Pretty much all Linux apps are written with X11 or Xorg in mind... since 99% of the apps will be running in *compatibility mode* I tend to wonder if GUI performance will drop like a rock. Go Go Ubuntu, your beginning to u-turn from my favorite os... Debian anyone?

Re:fuuu

[Ash-Fox]

The first time I ' ssh -X ' and have it not work I will be seeking new desktop choices.

I'm pretty sure you're supposed to use ssh -Y on Ubuntu, not -X.

So what?

[WindBourne]

This will still have X and the ability to forward X over SSH. Normally, the X-server is the base app and then has drivers to interact with kernels/Drivers. NOW, what will happen is that Wayland will be the 'server' and X-server will simply run as an app talking to it. On the local application, you WILL see a slowdown. The reason is that you just added another layer. BUT, when X is going over the network, the slowness is the network. As such, you should not notice ANY difference at all in terms of speed.

Finally, there is VNC, though I myself prefer X/ssh.

Translation Please?

[assertation]

I've been using Ubuntu for eons, but I do not follow the OSS world as closely as I used to.

Is "Wayland" a replacement for X Windows? If so, does that also make it a replacement for the KDE and GNOME or do those two things sit on top of X windows?

What is Unity and how does it relate to GNOME or the KDE?

Is Ubuntu moving to these technologies because they use less resources are faster and will allow Ubuntu to work better on devices other than PCs?

I did read the article and google on these terms

Thanks

Re:Translation Please?

[Tetsujin]

Is "Wayland" a replacement for X Windows?

Basically, yes. The X Window System is a very old architecture by computing standards. It has served us well and I am not entirely comfortable with dismissing it... but the basic difference in approach here is that "Wayland" is strictly a local-display implementation, where X is built from the ground up to be network-transparent, with local-display enhancements added via extensions. The idea behind ditching network-transparency is to optimize graphics performance, making applications running on the local machine perform and behave better (i.e. giving video players the control they need to be able to synchronize video frames with the monitor's vertical blanking).

If so, does that also make it a replacement for the KDE and GNOME or do those two things sit on top of X windows?

The latter. Both KDE and GNOME can work directly on Wayland, I believe, if compiled for that.

What is Unity and how does it relate to GNOME or the KDE?

As I understand it, Unity is a window manager/desktop environment on which other applications can run. So it's like GNOME or KDE in that regard, but the GNOME and KDE projects themselves include a bunch of applications which could be run within the Unity environment... So they're not mutually exclusive things.

Is Ubuntu moving to these technologies because they use less resources are faster and will allow Ubuntu to work better on devices other than PCs?

Well, I think "less resources and faster" is basically what they're after with Wayland. A lot of what's included in X just isn't necessary or useful for modern applications. A lot of the things people are doing with slick GUI transitions and the like really aren't compatible with X's network transparency anyway.

With Unity I guess they're just trying to build a better GUI. Among other things it's supposed to be good for use on touchscreens (meaning, for instance, items on-screen have to be big enough to tap accurately with a finger...)

Re:Translation Please?

[assertation]

Thanks a lot of the info!

It's all part of the plan

[Minwee]

Once the switch to Wayland is made, Canonical will change its name to the Wayland-Ubuntu Corporation,

Fix the bugs first

[jirka]

The main problem with linux is how terribly quirky it is. this will not be solved by moving to a new graphical platform which will yet again not work on many machines simply because it is not getting tested enough.

If instead of focusing on blue sky nonsense, they focused on making things work. Making sure the software works out of the box FOR EVERYONE. Right now Ubuntu installs and works for many people but it does not handle corner cases. Lots of things remain difficult or broken. Try plugging in a new display (on my computer the xrandr stuff tries to be really smart and generally screws up). About 1 out of 5 times I boot up I get an interrupt storm that leaves the display unsable (bugs have been files by many people in all kinds of places, this seems to be a bug that is at least a year out there ... no fix coming). About 1 out of 20 times I boot up I get no mouse. Running windows software is very quirky, and wine adds weird items to my menus. Most of the games I installed for my 4 year old either crash at various times or have really bad usability problems. Starting abiword takes about as long as starting openoffice nowdays. Gedit is also not instantaeous. Damnit, a text editor started up IMMEDIATELY on an XT 20 years ago. The screensaver keeps starting up at random times (sometimes even while I am typing). Not all movie players disable the screensaver. Pretty much every music player I tried chokes at some point on my mp3 collection. Either it crashes, stops playing, or does something odd such as the UI freezes and Chuck berry keeps playing (my wife now hates chuck berry because of this). The list just goes on and on.

I know windows has many problems, so does the mac. But if linux is to become maistream, first it needs to work 100%, not 95% or 90%. I don't actually see much difference in usability since about the year 2001 or so. Some things have more graphical effects now, take longer to load and have more bugs. But that's it.

Moving to Wayland (or even working on Unity) will not solve ANY these problems.

I have recently moved back to Fedora. Fedora is still as user unfriendly as it was before and has all these same problems, but at least it's OK for a geeky me. But I assume Fedora has been focusing on the server, and presumably they might have gotten ahead in that department. On the desktop though ... things suck (they would suck for me a lot more on windows or mac though ... so ...)

Jiri

An Ignorant End User's Opinion

[assertation]

I've been using Linux as a distro since about 1999. I've been using Ubuntu exclusively since the first release.

Ubuntu made GNOME palatable to me, but the few complaints I have had I have been told to take it up with GNOME because it was in their bailiwick. I have found a "have it OUR way" and a "if you don't like it vote with your feet" attitude prevalent in that org. I am told that it is their decision and that mentality why I live with these issues as an Ubuntu user:

1. A GUI interface that limits the number of hot keys _easily_ made
2. Why bringing up the calculator via a keyboard command takes several times longer than using the GUI
3. Why bringing up the System | About Ubuntu window takes a minute and a half while similar shell commands are instant
4. Why I can't pick "Monday" as the start of my week in the calendar applet or have it stay that way once I hacked it

I realize there a lot of technical and political issues of which I am ignorant. I'm just saying as an ignorant desktop end user and Ubuntu fan, that if Unbuntu can make all of these technologies work well I don't have a reason to miss GNOME.

"Compatibility Layer" and performance

[assertation]

Will the "compatibility layer" cause a visible performance hit?

See also Ubuntu brainstorm #18253

[tepples]

something you wouldn't want your sysad to say... "where's the GUI on this thing?"

To which one would ideally say: "let's apt-get install one". There's no reason one can't make or use a tool under *Linux or *BSD that performs a similar function to MMC under Windows if it helps an admin set up a one-off server or the master image for a workstation farm, server farm, or cloud. I'm not the only one to have thought of this.

Stay away from Mac and Microsoft!

[formfeed]

apple, Microsoft and Sun all have radically changed their widowing systems on many occasions

I knew it! Better stay with Linux then.

FreeNX is great for low latency connections.

[yossarianuk]

FreeNX / NXserver is great for low latency links. After moving house recently I was without broadband for a week or so and had to end up using a 56k modem to connect to my companies gateway - it was actually usable. With FreeNX you can have an unlimited amount of connections (will limited by machine ram,etc only) - works fine with gnome/kde/lxde

Re:we'll be able to retain the ability to run X ap

[jo_ham]

Yes, but Intel Macs feature Rosetta, which enables you to run PPC-only binaries in 10.6 if you have some legacy software.

It no longer supports Classic (for OS 9 apps) but it will run PPC binaries still.

first screenshout

[Anonymous Coward]

First screenshot from inside Wayland/Unity

http://img716.imageshack.us/img716/8281/unitywaylad.png

OMG OMG OMG

Re:And thus the curse of Open Source manifests its

[forkazoo]

And you can plausibly argue three processor family switches if you count ARM. iOS is basically derived from OS-X. The sales model for iOS devices is different from the Mac, but the underlying technology is really not that different from what an ARM based OS-X Mac would be.

Also, X11 has been around since the Amiga was relevant.

rootless

[Junta]

Only X has a pretty solid seamless story. NX added better network performance and connection loss tolerance. I would say NX is the optimal approach. It is, however, not without it's warts (the one I can think of is the inability for remote apps to get into the systray when using NX as opposed to X.

Re:rootless

[sarhjinian]

I had forgotten about NX, thanks!

GNOME Developers

[sarhjinian]

I'd feel more amenable to the GNOME developers plight if they'd give some attention to some of the three-plus year old bugs that don't seem to get any attention. Some of the bugs aren't serious, but they're so simple and yet so annoying you'd figure someone would have fixed them years ago.

I actually prefer GNOME to KDE because it seems cleaner, simpler and less cluttered, but the KDE people seem much more willing to fix bugs. Mind you, KDE 4 gave them ample opportunity to fix lots.

Re:GNOME Developers

[ebassi]

I'd feel more amenable to the GNOME developers plight if they'd give some attention to some of the three-plus year old bugs that don't seem to get any attention. Some of the bugs aren't serious, but they're so simple and yet so annoying you'd figure someone would have fixed them years ago.

and I'd feel more amenable about closing "simple" bugs if you provided patches for them instead of moaning and bitching about them on /. if you provided a patch and didn't stick around to actually have it committed then you're not doing open source development right either, by the way.

Re:GNOME Developers

[sarhjinian]

And this attitude is precisely why people get frustrated about reporting bugs.

Not all of us can or have the time to code, but we do have the time to fill out a bug report, thinking we're helping you find an issue. It seems a great system, and the various projects are generally pretty open about taking bug reports.

And then the bug sits there, gets reassigned, gets flagged as WORKSFORME, gets pushed back or obviated for the next version, etc, etc. I've caught bugs, gone to report them, noted that there's a similar bug that's been open since 2007 and, at that point just given up.

And you know, it's made all the worse by developers who either brush it off or, worse, make comments like the above. Experienced users who can't or don't code will work around it, but new users will just go away and never come back.

Not that closed source is necessarily much better, but at least it's more professional and less egotistical. Heck, I won't even say that this is the defining characteristic of all OSS projects (Zimbra, for example, does a pretty good job at this) but it's too common, and both Ubuntu's LaunchPad and Gnome's Bug Tracker are prime examples.

Is that the "open source way"? Or are you not fixing a bug that scores of users have taken the trouble to report over three years out of laziness, or is it some kind of nerd pride? What must we go, oh master?

Re:GNOME Developers

[ebassi]

tl;dr. whining never gets you anywhere.

Is that the "open source way"? Or are you not fixing a bug that scores of users have taken the trouble to report over three years out of laziness, or is it some kind of nerd pride? What must we go, oh master?

nobody pays anybody to work full time on the whole of GNOME; it's a volunteer effort, through and through. this is also true of a lot of other open source projects.

if no developer is working on your pet bug you have two choices: a) learn the ropes and fix it yourself or b) convince a developer (with money, if necessary) that your pet bug is important and that he or she should work on it for you.

otherwise you have misunderstood how open source works, and you don't get to complain.

Re:Good bye effective multi-user graphical computi

[TeknoHog]

Those who do not understand (uni)X are condemned to reinvent it, poorly.

Re:we'll be able to retain the ability to run X ap

[Cajun Hell]

You can run X on OS/X and WIndows so running X on top of Unity is probably going to be okay as well.

I think that's half of what people want.

You can run an X server on those OSes, so that a OS/X or Windows user can see the output of some X client that is running on a Linux box. But what it's can't do, is make the lOS/X or Windows apps be X clients. That is, you can't go to your Linux box that is running an X server, ssh to the Mac, launch a Mac app, and see that window on your Linux box's screen. You've gotta do something totally different (e.g. VNC) to do that.

How will one machine running Wayland show a window for an app that is running on another machine running Wayland? We can currently do that easily with X, in a way that is less convenient for Mac or Windows users, with their VNC or Citrix addons.

...not to mention

[toby]

OS X has Quartz... *runz*

Comment removed

[account_deleted]

Comment removed based on user account deletion

How effecient is x really?

[Anonymous Coward]

Over a network x is extremely chatty with no effective round trip minimization optimizations to the point of being unusable over anything but a LAN given latency and bandwidth requirements. When compared with modern remote access technologies from Citrix et al it is evident there are major issues in the existing x technology stack for REMOTE displays.

For local applications I don't know that any of this matters? I loved the replacable window manager concept that allows a great deal of customization of the environment and linux x UI's in general look pretty cool although font rendering/aliasing is much worse than windows/mac.

It would seem to me as long as you could compile an application to target the same windowing API used (Qt or whtaever) by most applications you care about then who cares about the underlying display architecture if the apps still run? If there is some performance or reliability benefits to switching to a new system then why not?

Firefly quote

[TangoMargarine]

My food is problematic.

Headline wrong

[fnj]

Headline should say "Shuttleworth marginalizes ubuntu with stupid decisions."

Thank god ubuntu isn't all there is to linux, and linux isn't all there is to posix operating systems.

Embed/Mobile world

[DrYak]

The difference is that speeds, reliability, driver devl, security, etc will improve.

Also, it's important for the embed world, as it's a huge improvement over a pure frame buffer, without the additional load of an X server.
So it's really interesting for platform running without a full X server, like some embed/mobile platforms.

Before Wayland :
- either you use a framebuffer dev, and then one and only one toolkit, which is able to handle several concurrent application on 1 buffer (GTK or QT).
- or you have to run a full blown Xserver (which is much heavy from the constricted point of view of an embed device) but any tool-kit you want can collaborate.

Now with Wayland :
- Wayland provides an accelerated local compositing manager. You can use it to run several apps on several tool-kit, as long as the tool kit are adapted to run on this backend (GTK and QT are already getting adapter).

Thus it enables future Nokia Nxx devices to free some resources by dropping the full blown X server.
Or it enables framebuffer-only phones and PDA to share several type of UI widgets.

X in Wayland

[Tetsujin]

Backwards compatibility is achieved by (optionally) running X(server) as a sub-process of Wayland.

Isn't this inherently inefficient?

X is designed as a network-transparent windowing protocol first, with optimizations to improve performance on a local display added-on.

If you start with a display system that's optimized for local display, and then implement, on top of that, a network-transparent display system, there's no reason the implementation of the latter should be inherently less efficient than its direct implementation - unless the display server or the compatibility layer are implemented badly, or there's some level of incompatibility in the basic concepts of the two systems that makes a compatibility layer difficult to achieve.

I know very little about Wayland - but if it's largely based on DRI and OpenGL, then implementing X on top of that shouldn't have a significant negative impact on X performance.

Personally I'm not sure how I feel about moving away from a network-transparent rendering system. It's something I've grown used to in my years using X (about 1996-present). That alone is enough to make me uncomfortable with the change. I don't relish the idea of moving to a system where some apps will support remote display via X and others won't - or where I might have to choose between an X version of an app and a Wayland version... It reminds me of the situation on my Windows machine at work: choosing between Win32-native, and Cygwin/X versions of packages...

Though, on the other hand, how frequently do I actually use this feature? I use it for Emacs and a few other things, and that's about it. I never attempt running Firefox or Blender or GIMP or VLC remotely via X, I always just run those on the local machine. If my experience really is typical, then the network-transparency feature of X is being underused, typically, to the extent that it's not worth making it a design priority. (And, actually, I think people tend not to design Linux GUI apps with remote display in mind. I think they're more commonly developed for a local display, with the result that their behavior might be a bit too network-intensive or latency-sensitive to work with a remote display...) It might really be better to optimize for local display and then have remote display via a special layer: VLC or whatever else.

The Wayland FAQ was kind of interesting to read. It's interesting what they have to say about X's legacy baggage, for instance. Of course, I've heard a lot of this stuff before... I remember "Berlin" and GGI as a previous attempt at roughly the same thing. Maybe Wayland will yield a better result in the end? I don't know.

Re:X in Wayland

[t2t10]

If you start with a display system that's optimized for local display, and then implement, on top of that, a network-transparent display system, there's no reason the implementation of the latter should be inherently less efficient than its direct implementation

Geez, you think the X11 designers were that dumb? That's of course what the X11 designers did.

X11 is, however, unapologetically a client/server window system. And you know what? That was the right choice, because 25 years later, Windows and Macintosh have also moved to client/server window systems. Unlike WIndows and Macintosh, X11 was designed to do the right thing from the ground up, while those systems had to be retrofitted.

The Wayland FAQ [google.com] was kind of interesting to read. It's interesting what they have to say about X's legacy baggage, for instance

Yeah, there is some legacy baggage, like old font models and old graphics models. Any UNIX/Linux desktop needs the libraries to implement them anyway. It makes no difference whether those libraries happen to be in the window system server (X11) or two separate processes (X11+Wayland); either way, they are always going to be loaded. But that code doesn't affect performance, and it is pretty much self-contained. It's also not that much work to implement from scratch if you really have to.

Re:And thus the curse of Open Source manifests its

[AdamWill]

Thing is, Wayland isn't a Canonical pet project; it's the pet project of Kristian Hogsborg, who used to work for Red Hat and now works for Intel. As far as anyone can tell, Canonical doesn't contribute anything to Wayland at present. This is just Mark saying 'hey, we like Wayland, we want to use it'.

GNOME?

[supersloshy]

The move means there is now little reason for GNOME developers to recommend Ubuntu as an operating system.

Um... exactly what does a display server have to do with GNOME or any other desktop environment? Last I checked, GTK was already ported/is being ported to Wayland anyways. What's the deal here? I call bogus summary/article.

C64 love

[Anonymous Coward]

(Joe Q Public) - "Not a clue what that buddy's talking about. WM? I'll just windows like everyone else."

You tend to assume everyone else is a fucking stupid as you are. That is incorrect; you are a special kind of retarded.

How about FreeNX

[GPLHost-Thomas]

Great, a new GUI stack... But how am-I going to run FreeNX then? Instead of working on silly stuff and REMOVING networking, they'd better work on ADDING a nice FreeNX stack to begin with.

You can still run X programs

[ceswiedler]

For all of you complaining about how Shuttleworth is trying to kill the network transparency of X... This doesn't affect your X programs, which are always going to be able to run over the network due to the design of X. There's no reason why a desktop machine running Wayland wouldn't be able to run X programs. The only effect of this is to allow building GUI programs specifically for Wayland.

And seeing as those apps are specifically designed to use advanced features like 3D and compositing--why would you expect them to run reasonably over the network? Do you tunnel glxgears or TuxRacer over a WAN?

If a developer is writing an app which would usefully run over a network, they can write it using X and everybody is happy. If they need the more advanced stuff of Wayland, then network transparency probably doesn't make sense anyway

Moving to Wayland does not prevent fixing bugs

[Tetsujin]

The main problem with linux is how terribly quirky it is. this will not be solved by moving to a new graphical platform which will yet again not work on many machines simply because it is not getting tested enough.

...

Moving to Wayland (or even working on Unity) will not solve ANY these problems.

Improving infrastructure does address these sorts of problems. If your infrastructure is better-suited to what you're trying to accomplish, some of the complications introduced by the old infrastructure decisions will disappear and it'll be easier to accomplish your goals.

As a specific example: a goal of Wayland is to provide compositing and window management on top of a rendering layer that's close enough to the hardware that applications can easily do things like synchronize video playback to vertical refresh. Hence, one "quirk" in Linux video playback goes away.

This isn't going to solve everything, of course, but what they're going for is a step in the right direction. Since Wayland is a relatively new design (i.e. built with present-day assumptions about hardware expectations) it very well may help with the issues you mentioned - especially things like adding a display. (X handles that via the XRandR extension - which is a relatively recent development.) Similarly, for developers who are looking to add visual flair to their apps, it's probably going to be easier to implement that correctly with Wayland than X, and thus they'll have more time to address the bugs in their work.

Big changes like this can introduce a lot of pain up front, in the transitional period. It's not trivial to move away from a display technology that's been the basis of most Unix GUI work for the last 20+ years. But the reason they're doing this is because they hope it will pay off in the long run, in the form of simpler, more reliable, better-performing, and better-looking application software. If this really is the right direction to move - still it's not an easy choice, but an important one nonetheless. If one considers basing Linux GUI on X to be a mistake, then the apparent choice here is to go on building and fixing software that was based on that poor decision, or face the short-term transition pain and go for something better.

Personally I'm not sold on the change. I am fond of the way X apps work, and to some extent I fear losing its capabilities, and the underlying assumption that every GUI app is network-transparent (even if there are numerous problems with this in practice...) But, personal attachments aside, still I think this move could be a very important and valuable improvement to the way GUI is handled on Linux.

Oh thank heavens.

[DdJ]

My first reaction was "oh thank heavens, maybe there will be a desktop Linux I can stand to use some day soon". My second reaction was "hey, maybe it will even happen before Apple makes MacOS unusable for me". (I do Java development for a living right now; that day may be coming.)

Re:we'll be able to retain the ability to run X ap

[LWATCDR]

Funny but VNC works great for me. One solution for that would be to include a remote desktop client into Wayland from the start. That could give you the same functionality of redirecting X. And lets be honest X redirect X also has it's limitations. For instance it doesn't redirect sound or give you the option for easy file transfers.
Yes their are ways to do it but they are not simple for the average end user.
BTW I am the type of guy the always loads joe on his linux boxes because it is a light fast editor that works well over ssh!
But I also realize that I am becoming a member of shrinking minority.

Re:Good bye effective multi-user graphical computi

[Tetsujin]

X has a LOT of power, but remains poorly understood.

Now we are going to get applications targeted to a single user GUI, and with that all the problems the other OSes have with their single user GUI. :(

I would argue that we're already getting this: applications targeted to a single user local-display GUI, that just happen to be implemented on top of a protocol capable of remote display.

When someone tried to implement effects like animated sliding menus or whatever on top of X, what you get is a program that can technically operate network-transparent - but doesn't necessarily perform well.

Fantastic.

[JonJ]

Ubuntu announces it will incorporate another Red Hat technology, news at 11. I'm just surprised they haven't already forked RHDS and just renamed it Canonical Directory Server.

Re:Ubuntu is systematically ruining their distrobu

[Tetsujin]

come back to earth. You guys have been making some absolutely STUPID decisions as to the direction you are trying to go. I will be moving away from ubuntu because it is getting worse and worse. Wake up, you can't even call your crap linux anymore

PulseAudio does come to mind, doesn't it?

Still, they certainly can call it "Linux" as long as Linux is the kernel they use. Although I do like a lot about what has defined "The Linux Platform" so far, I prefer to think that the definition of the "platform" is subject to interpretation and change.

Wayland on how Wayland and X differ

[John Whitley]

The Wayland site has a good exposition of how Wayland and X differ at the architectural level. This also clearly explains (and diagrams) what happens when X runs as a Wayland client.

Looks like a breath of fresh air in the Linux rendering space, and someone with enough momentum behind them to drive it.

Re:we'll be able to retain the ability to run X ap

[Knuckles]

You missed the part where Wayland is MIT licensed, quite unlike your PS3?

Re:And thus the curse of Open Source manifests its

[tenchikaibyaku]

When you think about it, it's pretty amazing that we can sit here with our fancy 3D-accelerated desktops, wobbly windows and all, using software designed in the 80's.

Personally I think that tells us something about how extensible X is, and I get a bit nervous when people talk about throwing it out.

(And then there's the lovely network transparency of course, is it really worth throwing that away..)

Re:And thus the curse of Open Source manifests its

[Anonymous Coward]

Are you seriously claiming that proprietary software ever gets finished ?

The only real difference is that every once in a while you get to pay for the updates and proprietary software tends to use much higher version numbers.

liushengquan

[Anonymous Coward]

I am happy to find this very useful for me, as it contains lot of information.ugg shoes I always prefer to read the quality content and this thing I found in you post.Thanks for sharing.

I am happy to find this very useful for me, as it contains lot of information. I always prefer to read the quality content . Thanks for sharing.

Ubuntu Server

[fatalGlory]

For all the panic about losing some of the networked functionality and Canonical doesn't care about non-traditional desktop setups (with mounted networked home dirs, etc), no one seems to have brought up Ubuntu Server. Ubuntu is not just a desktop OS, the server variant is looking more formidable all the time. Do we really expect Canonical to take its actively maintained server version (which they're pushing for cloud applications) and neuter the ability to deploy it in big server environments where home dirs are NFS mounted and people want to use GUI config tools on headless servers?

I have just finished several years of distro hopping because I felt Ubuntu (my first full-time Linux) wasn't hardcore enough for my taste. After moving to Debian, then Arch, then Fedora for 6-12 months each, I'm now back at Ubuntu with a renewed appreciation for how good a distro it is.

I give Canonical more credit than to screw it up that badly after getting so many things right over the last few years.

No pain no gain.

[Anonymous Coward]

IMHO, Canonical is making big strides for Ubuntu towards trumping Windows, ahead of other Linux distributions. No pain, no gain. Sacrifices have to be made.

Re:And thus the curse of Open Source manifests its

[multi io]

(And then there's the lovely network transparency of course, is it really worth throwing that away..)

Reading the project page, it seems Wayland defines a network protocol with a display server and client libraries too. So you should still have network transparency.

How does it harm? fragmentation

[lpq]

So there's no potential problem with people developing apps that only work on 'wayland' that won't be accessible on non-Ubuntu platforms?

How is this different than when MS, creates a new feature set above the standard. The problem is that if people code to the extensions, then the resulting programs won't work on the standard -- anywhere... Isn't that called "Embrace, Extend, Extinguish?" Initially X is embraced by Ubuntu for remote display of apps. Now they are extending it through Wayland -- of course if developers code only to the 'X' standard, they'll be fine, but if devs write code to take advantage of the new features, then it won't run anywhere else.

How is this harmful? How was IE6 harmful to the internet, or MS's now defunct JSCRIPT (?) that extended Java?

Working to extend 'X' and OpenGL -- something that all platforms could partake of is difficult -- and sure it's easier to go off and do your own thing. Might be the best thing for Ubuntu, but it IS a compatibility problem and it does have the effect of segmenting the market.

That said -- if Ubuntu was my world, and I just wanted the best functionality in that world, trying to update the rest of the world to accommodate new features might be something that could delay new stuff for years, so it might be the best way to achieve an outstanding product. But no one should be deluded that it won't have an effect on app development in general, since there will be those who will develop for Wayland who would have developed generally useful apps that run under 'X'. That will have some negative (maybe small) impact on the 'application market' as a whole (primarily due to fragmentation).

RELEASE: Wayland-Yutani

[Anonymous Coward]

Hello,

just want to let know I release my Mod which runs on top of Wayland.

- Yutani.

Re:How does it harm? fragmentation

[jbolden]

It will lead to fragmentation you'll have:

1) Generic X apps
2) Wayland only apps
3) Apps with an X version and a Wayland version.

Or Wayland becomes super popular, essentially X12, and then only legacy stuff in on X11 and everything is for Wayland. That's forking.

Re:How does it harm? fragmentation

[lpq]

So you agree that it will cause fragmentation...only problem is the forking issue ... will it become X12? If it does, then that's great, but how long before it's supported on Windows -- or more specifically, 'cygwin on windows'. From what I've read it's heavily dependent on linux-only kernel features, so that would require a huge rewrite to emulate on windows with a huge question mark for performance. Right now Cygwin on Windows is maintained by about maybe 2 people at most. You think they can do that port in their spare time when it's stuff that might take someone who's fluent in both Windows and Linux 'internals' to write?

How about the Mac platform? Same problem. Right now, X works because it runs everywhere. Any solution that's considered for X12, also has to meet the "runs everywhere" requirement to be a true replacement. And how long before that happens? I would *love* to see it happen -- and see 3D app acceleration over the net, but right now, OpenGL over a 1Gbit network sucks for performance, Maybe this will optimize remote traffic and that would help, but more than likely it will result in new apps that only run on Ubuntu-enabled platforms. If those apps don't run on windows, it will shut those apps out of the workplace and severely limit their adoption. It may harm Ubuntu's adoption if those apps become important in the management of an Ubuntu server.

Something to consider...

Re:How does it harm? fragmentation

[jbolden]

Lets separate the two.

I'd assume something likes this
1) Wayland is standard on Ubuntu
2) Wayland is standard for most desktop Linuxes
as this stage
3) BSD people create an emulation layer in particular a Wayland for FreeBSD and OpenBSD
4) That creates Wayland for Darwin (i.e. its part of Macports)
5) Apple takes this and puts in into the Quartz window manager (i.e. Wayland for Aqua)

I'd day steps 3 to 5 can happen 3-6 yrs after 2.

I'd say around the time 5 is happening other X platforms like QNX start working on the problem. So say something like a decade after step 4 you get a Cygwin port.

Re:How does it harm? fragmentation

[lpq]

Joyful! :-)
I agree...

But the effect on Ubuntu's spread into the business server marketplace might be stunted, if key-apps to control Ubuntu server's are written using Wayland since most business's will be using Windows desktops. They might have some other 'X' solution than cygwin, but whatever solution they have it will be an impediment to acceptance if they can't access server-control apps on Ubuntu, but they can, on Redhat or SuSE...

That was my main point -- fragmentation often causes interoperability problems even though it may benefit the Ubuntu platform. ***Hopefully*** application writers will be smart enough to offer "slick" functionality under Ubuntu, but just as functional (though maybe not as 'pretty'?) operation under 'X'. Too often app writers don't like the excess baggage of writing for the older standards...(understandably)...

Re:How does it harm? fragmentation

[jbolden]

I doubt there will be many applications that create two layers. I'd say more in keeping with Linux would be the core applications is written for X and then reskinned using Wayland. Any application for Wayland is likely not going to support X.

And yes that means Ubuntu server will require VNC to be usable for Windows. There won't be any cygwin / network transparency. But that isn't any different than the situation with native Windows applications so I don't see it as very much of a big deal. Network transparency doesn't work well enough at the X level anymore. Rich web interfaces like flash offer the same functionality for most apps that need to operate remotely. Wayland isn't designed to be seamless remote at all so remote administration in a Wayland based system won't be X or Wayland apps but rather web apps.

also: DBE

[t2t10]

Ask anybody who programmed using Xlib and you would

Where did I say that you should program in Xlib? Xlib is totally irrelevant. Xlib is merely a crappy and obsolete C library that gives you access to the protocol. Many toolkits don't use it.

What matters is the protocol itself and the common standard by which different toolkits interact with the server, the window manager, and each other.

be immediately advised to use GTK/Qt/wxWidgets/etc instead.

Yes, and if we keep going down this path, the only toolkits you will be able to use are Gtk+ and Qt and things built on them, and that's bad.

Re:also: DBE

[ThePhilips]

Xlib is totally irrelevant. Xlib is merely a crappy and obsolete C library that gives you access to the protocol. Many toolkits don't use it.

You do not know what you are talking about. ALL toolkits use it because interface to X is de-facto standardized on level of Xlib (and de jure via LSB on Linux), not the network protocol. The network protocol is also standardized (and that's why X clients running different systems can connect to alien X servers) but it is pretty much never used by applications or toolkits directly.

Run "ldd" on any GUI application - and watch the libX11 being always there.

Re:also: DBE

[t2t10]

You do not know what you are talking about. ALL toolkits use it because interface to X is de-facto standardized on level of Xlib (and de jure via LSB on Linux), not the network protocol.

Sorry, but you're wrong, and you don't understand what that Wikipedia page is saying.

There have been X client libraries not based on xlib, for example for Java, ML, Lisp, Python, Perl, and many other languages. A lot of embedded software stacks use their own X11 protocol implementations. There is even a new protocol implementation for C called XCB that will hopefully replace xlib (xlib will run on top of it for backwards compatibility). In addition to all that, there are many tools that speak the X protocol directly. Heck, I've written my own X client library; it's not rocket science.

Ubuntu may be able to move a large number of desktop users to Wayland, but at the cost of splitting the UNIX/Linux community by making Ubuntu's X11 implementation effectively as crappy as running X11 on Windows or OS X. That might be a reasonable thing to do if there were significant gains in performance or usability to be expected, but there won't be. Wayland won't be any faster, any easier to program, or any easier to use than the current X11-based systems; to the contrary, a lot of things will end up being harder to do and less consistent than on current X11-based systems.

Re:also: DBE

[ThePhilips]

I'm not sure what you are smoking.

Embedded X server is nonsense: embedded systems do not even implement X server/client per se, they simply wrap graphics right into Xlib-like interface. They are embedded, they do not need the network-transparency. They advertise "We support X!" while in reality it means "stuff which uses Xlib compiles." I worked with bunch of them, and it is considered to be a high-end feature to have the X server at all. In the end, they compile XFree/X.org with special driver which simply strips/skips/bypasses all the network stuff completely, plugging Xlib/Xt directly into the core of X.

Likewise I know for a fact that Sun's Java uses the Xlib. Perl and Python rely on toolkits and are oblivious to the OS or graphics system one runs (== use Xlib).

Heck, I've written my own X client library; it's not rocket science.

Writing Xlib which is a thin wrapper for X protocol isn't rocket science - building applications right on top of it is. Because it is not far from programming in assembler: very efficient but you wont get too much of functionality in any measurable time frame.

That might be a reasonable thing to do if there were significant gains in performance or usability to be expected, but there won't be.

Read up e.g. how FreeType or GDK libraries works to understand why your expertise is probably out-dated. 99.9% of rendering of modern Qt and GTK applications happens inside the client, not on server, only ready-to-display images are transferred to the X server. Logical question arises: why we need to transfer that stuff to server at all?

Re:also: DBE

[t2t10]

I'm not sure what you are smoking.

You're still trying to defend your idiotic, indefensible statement that "ALL toolkits use it because interface to X is de-facto standardized on level of Xlib"? What are you smoking? The Wikipedia entry you yourself pointed to proves you wrong, since there is already another C library, XCB.

Likewise I know for a fact that Sun's Java uses the Xlib. Perl and Python rely on toolkits and are oblivious to the OS or graphics system one runs (== use Xlib).

The fact that you can use Xlib from some of those languages doesn't mean that there aren't other bindings as well. (And you totally misunderstood the embedded comment; not much experience there, eh?)

99.9% of rendering of modern Qt and GTK applications happens inside the client, not on server,

You're stating the obvious.

only ready-to-display images are transferred to the X server. Logical question arises: why we need to transfer that stuff to server at all?

However you want to get those pixels onto the screen, X11 supports it: you can use shared memory, IPC, direct graphics. Wayland doesn't improve on any of that.

I ask you again: what specific problem is Wayland solving? How does it make the user interface meaningfully faster, more usable, or easier to program? Simplification of an architectural diagram is not enough reason to break compatibility and start over from scratch.

What Wayland will do is push window management into the toolkits, make the server less modular, and (apparently) push more functionality into the kernel, all of which are big negatives.