Shared publicly  - 
151
47
William Dauchy's profile photoShirondale Kelley's profile photopottendo pottendo's profile photoDonald Carr's profile photo
131 comments
 
WAT! Where do they live? I mean in what time zone is it 1 April?
They reinvent xdm/gdm/kdm and now they will reinvent wayland? Have it never struck them that the thing with open source is that they have access to the code so they might fix the problems they see.
I guess we soon see a new kernel project started by +Robert Ancell
Daniel Stone
+
49
50
49
 
The best part is that the input bit of the rationale is totally wrong: there's no way for clients to get another client's input, short of mapping a full-screen transparent window and convincing the compositor to not decorate it.  There also aren't any grabs.  Which I could've told them if anyone involved ever bothered to talk to anyone upstream.
 
+Daniel Stone and the things they claim wayland/weston input can't be extended to support:

"... adjusting and extending X's input model is difficult and supporting features like input event batching and compression, motion event prediction together with associated power-saving strategies or flexible synchronization schemes for aligning input event delivery and rendering operations is (too) complex."

is already implemented and working in weston today...
 
The arguments against Wayland look poor. I suspect a typical "we cannot control that other project so we create our own" thing going on here. Which is ironic given how easy it is to contribute to Wayland, compared to other projects.
 
"A protocol-agnostic inner core that is extremely well-defined, well-tested and portable." [emphasis mine]

What does this even mean in the context of a piece of specification?
 
I am sure "Mir" is going to be a project with a fantastic future, just like bazaar, or Upstart, or Project Harmony before it.
 
They should call the next Ubuntu Jumping Sharks
 
I remember when open source project announcements used to come with source.
 
Especially if I read "as it does not fulfill our requirements completely". Sane people would have sent an email to the wayland list, or would have grabbed +Kristian Høgsberg or any other wayland expert at any of the many conferences, and would have discussed thos issues. Maybe they'd even just had hire one of the many companies around wayland and had them fix the few issues. Well, but not so Canonical... ... ... ... they burn hundred thousands, if not millions of dollars and rewrite.
 
Also, isn't Mir this thing that burnt and crashed into the South Pacific Ocean near Fiji on 23rd March, 2001, after some dudes in Russia flipped a switch after they gave it up?

This must be metaphor for something, haven't figured out for what yet, though, must be something deeper than "This software includes a space toilet and Canonical will give it up one day, when it will burn and crash and then we'll be in a south pacific paradise setting."
 
+Kristian Høgsberg BELIEVE IT! No, not really. I'm really upset about this. I guess I can now forget about any hope of running desktop Linux apps on my phone any time soon. Wayland was supposed to fill the gap of almost-nonexistent X11 binaries for mobile GPUs.
 
Mir is, I believe, the Russian word for peace. The irony.
 
So in the long standing tradition of Canonical naming their projects the exact opposite of what they are actually about (unity > split): Is it safe to assume that "Mir" ("peace") is actually about starting a war? ;-)
 
has someone forked emir yet? its been like a half a day.
 
I'm upset about this because it discards all the effort that has gone into Wayland for reasons that apparently don't hold much water. I just hope it doesn't marginalize all other desktop Linux distros if they decide to go ahead with Wayland.
 
+Lennart Poettering Actually, Mir is one of the great success stories of the space age. It was started in '86, just two years after X11. "Mir" means peace in Russian.

 I think Canonical's naming is all about irony: Unity, Harmony, and whatever is next to come. Nomen est omen kind of works differently...
Philip Witte
+
39
40
39
 
This is ridiculous. Unless Canonical has secretly convinced Valve, NVidia, and AMD developers to contribute it's hard to believe they could match the existing efforts and experience of the Wayland developers. Especially in a single year.

If this gains any momentum at all, it will at very least hinder NVidia and AMD's willingness to support Wayland. Thanks a lot Canonical.
 
A reads of the mir stuff says to me "oh look. we don't understand wayland at all, neither do we grok X11 that much either".

Their reasoning on the input in wayland are just total bunk. It doesn't suffer from the X11 snooping/security issues. It's rather clean and neat and the compositor (be it weston or any other one) can decide where input is routed and when, and i'ts easy to do.

Their reasoning with respect to the Shell just smells of hand-wavy bunk to me. There is nothing privileged about it. It is part of the wayland protocol spec that a compositor may implement any way it likes as long as the results "work" given the shell style it is implementing. You need a shell because things like focus policies, other display policies etc. are implemented outside of clients and there has to be enough control and data sharing to enable that to work.

I would just say "ignore mir/canonical and just keep plodding on with wayland".
 
+Philip Witte Oh yeah, Valve is probably really happy about this. Along the the rolling release thing this should really make a great base to build a console on in the future ;-)
 
Their points about Wayland's shorcomings don't stand as there's nothing from Canonical on the Wayland mail list to address them before 1.0. As with upstart/systemd, they suffer from NIH and want control over the whole stack. What's next, their own kernel?
 
+Ignacio Martín they barely have anyone competent enough to write a display server, the fact that they are actually quite ignorant of how wayland works makes it even more apparent.
 
"Mir/Мир" means "peace" in slavic languages. In Russian it also means "world".
 
(+Ignacio Martín yeah, stupid Canonical, how could they start working on upstart when systemd was already working so finely and was included in so many distributions.)




— also: see my comment below.
 
+Pavol Klačanský why do you think wayland will be dead? seems everytime Canonical announce something non-ultras jump on the bandwagon
 
@ the people talking about upstart as if they started working on it after systemd, upstart has been around for years and before systemd every existed...
 
+Pavol Klačanský I'm not worried about Wayland's future at all.  I'm just irritated that this means more work for us, more work for upstream developers, more work for toolkits, more work for hardware vendors, and years of explaining that most of the page explaining that Mir had to be created because Wayland was a) hugely deficient, and b) unfixable, was total bullshit.
 
The one thing I'm a bit interested in is their "focus on CPU cycles, GPU cycles, memory and power consumption", along with benchmarks to make their case.
Frankly I'd be surprised if the most efficient protocol/implementation in the world burned more than 80mW, on average, than X, and, I'd imagine, Wayland is somewhat more frugal than that:)
 
+Pavol Klačanský
 Yeah, well, um, you see, creating a display server from scratch is tough. Even when you have all the experts in the FOSS community working on it. How Canonical plans on doing this themselves...
 
+Pavol Klačanský Wayland never relied on Ubuntu to use it and Canonical has never contributed anything to it, other distros will ship Wayland when it is ready which takes engineering and time
 
+Pavol Klačanský but wayland was a necessary evil :) something had to trim down the current x11+drm+dri... monster. it works, but is a beast. we COULD have implemented everything via x11 - or almost everything except input security and transforms. wayland solves it. mir just re-solves what wayland already solved, and THAT is why it is "just more work". and +Daniel Stone is right - it's EVEN more work that now all these devs have to spend large blocks of time fighting the mis-infomration canonical is spreading about.
 
Hopefully we can get C to update the page of ignorant comparisons with some real info
 
but who wants to bet that they're referencing the KHR_% gl extensions, much like Android, hmmm?

Oh well, I guess the reason that a spec is a spec is to allow for implementation...
 
+Pavol Klačanský No, Wayland won't go away. It got a significantly better community model than Mir and already is proving itself in real products. In contrast to that Mir just nothing but propritary hot air so far. Edit: As you ask for major players: Linux isn't just desktop distros. As you know although planed as desktop OS, Linux actually is much more successful in embedded space. So just check the wayland mailing list to learn about major players betting on Wayland. ;-)
 
+Daniel Stone Well, regarding more work for toolkit developer: Toolkit developers could just adopt Canonical's childish behavior and refuse cooperation. Let the trouble makers deal with the trouble they cause. I am kidding.
 
+Dave Airlie , might this at least encourage ATI and nVidia to produce KMS drivers that will help Wayland/Weston?
 
+Geoffrey Pursell no it probably won't, the thing is ATI/nvidia are much more catering to the workstation market with their Linux desktop work, I suspect they'd invest in this only if they could see an upside to that business, most of which doesn't run on Ubuntu. Their game supporting desktop work is ran as an offshoot of that area. They won't have to produce KMS drivers, just drivers that can do EGL and some extensions.
 
+pavolzetor Wayland 1,0 was only finalized months ago. Moreover, the situation is no different for Mir, very less so in fact (even if Ubuntu's ego and marketing presents it as otherwise). We're talking about THE FUTURE of Display Servers on Linux, and those of us who are out-crying against Canonical's choice to go their own way is precisely because it will hurt Linux development in general (not just Wayland) for virtual no reason (due to fallacious arguments and little to no attempt to communicate their concerns during the many months Mir was secretly under development).

Mir is obviously born out of Canonical's need to control the project, not out of any real technical concerns. It doesn't help anyone, it only makes major preparatory driver developers (needed for Linux-Gaming momentum) less inclined to support either system (unless Canonical has already secured their support, and are just lying about it. a concern all of it's own).
 
I'm still staring at "protocol-agnostic", wondering what the hell it means.
Translate
 
Canonical fork Debian, then they fork sysvinit, then they fork gnome, now they fork X. I guess they will fork the kernel soon.
 
Thanks for finding that. Last time I looked they seem to talk about the "Ubuntu Kernel", but it seems the website's been redesigned.

Nice job on the new site, by the way. It's very nice to read. Kudos. Much better than our http:///fedoraproject.org/ which I'm still unsatisfied with.
 
I'll jump aboard the train here. Why must Canonical want to make everything? Why must they waste time on something that already exists?
 
how does wayland hook into embeded gpu drivers like surface flinger?
 
+Earl Cameron Wayland's just a protocol.  How it hooks into the GPU drivers is up to the implementation, but we have a standard for supporting EGL-based drivers.  It's entirely possible to implement this for Android, just that no-one's done it yet, since integrating Android and generic Linux is extremely painful anyway.
 
The only reason why canonical is coming out with Mir is because they want to be the "google" of desktop linux. Just like google is the shot caller of android and they can direct the core project which ever way they want, canonical wants to be able to control the de facto desktop linux display server so they can tailor it to unity.

There is nothing wrong with wayland. 
 
While Microsoft is spending billions on advertising to break into the mobile market, and failing dismally. Canonical has set its ship on making Ubuntu work on all openned devices.  They need Mir to do it.  Wayland doesn't cut the mustard.

Tell me how Wayland could have been a solution for the great direction of Canonical?  Can Wayland easily be made to support Android drivers?  Is Wayland bogged down in compatibility games? Is Mir a clean approach with parallel compaitiblity using a X11 server ensuring a fast easy goal? Is Wayland top heavy?  Will Wayland be stage 1 finshed anytime soon?

Wake up and start supporting Canonical as it is leading the pack for Linux, so that Linux will be the dominate operating system.
Daniel Stone
+
29
30
29
 
+Jimmy Call Funny that you should heckle Wayland for being unfinished when Mir is a few linked lists and one glDrawArrays() call, nothing more.  As for the rest of your questions: would take too long; yes; no; don't know as it mostly doesn't exist; no; already is (1.0 released, protocol stable).  Thanks for your constructive input though.
Ben B
+
9
10
9
 
+Jimmy Call He seemed pretty straight to me. Anyway, Mir is in a direct opposition to the conclusion Mark Shuttleworth came up with in late 2010; that creating another display server (instead of supporting Wayland) would only split the community. I would agree with his assessment, honestly.

Edit: sorry Daniel, wrong person. ;)
 
Anybody else laugh at the fact this is built by "Compiz" developers.
 
All that is needed now is to accelerate the development of wayland putting it on production lines and make it the best of all the display managers. And this advance will make them reverse their silly decision perhaps...(it will not be the first time that they will reverse a basic decision. Remember how they treated UEFI before changing their stance)
 
Mark in 2010: "we evaluated the cost of building a new display manager, informed by the lessons learned in Wayland. We came to the conclusion that any such effort would only create a hard split in the world which wasn’t worth the cost of having done it. There are issues with Wayland, but they seem to be solvable, we’d rather be part of solving them than chasing a better alternative. So Wayland it is."
 
why does conoical have to ruin everything, they used to be cool man, i was so looking forward to a wayland only world, now that cant happen :'( i almost feel like im gonna cry, all those hours of research into wayland almost feel wasted
 
+Jerome Glisse The usleep in the example is just a (useless) leftover from older experimentations. The swapping is vsync-ed internally.
 
+Jamie Williams Simple way to solve the divide, is to make an Android interface so their drivers work with Wayland.  Then get Wayland to a level that justifies its use.  Canoncial may just decide to use that.

I think people around here aren't being honest to what is wrong with Wayland.  Why is it taking so long to be adopted and in the field.
 
+Jimmy Call There'd really just have to be someone to finally start or port a proper Wayland shell, but apparently companies are to dry to bet their money and programming offspring prefers toying with jQuery and stuff instead of NIHing a window manager - like it was common 10 years ago.
 
+Christoph Schunk not judging for whether or not it is good to invent another display server, but I wouldn't go that far in stating there's just bad C++ code unless you say why - I've also looked at it and yes, people MAY have personal coding styles, so has the linux kernel etc, and they don't use mine - but I wouldn't say they write bad C++ code just by starring at it for a few minutes either ;-)
 
Nothing new under the sun: Canonical is again rolling out their own (bad) solution.

The sad thing I believe is that, given Canonical/Ubuntu's weight, if they really pursue this idea it will be hard to ignore it. It means more pain at pretty much every level of the stack to support something that does not bring anything new or interesting compared to Wayland.
 
+Daniel Stone I thought that even if an "evil" client could create a fullscreen transparent window it wouldn't work because the evil client would consume the event so the normal window wouldn't receive them (as there is no way for the evil client to redirect the event to the normal client)?
 
looks like good old kde gnome times. competition is good. so in the end the "best" solution will win. both projects will gain speed now it´s going to be interesting.
 
+Stéphane Raimbault
It's ironic that Canonical made PulseAudio the success it is when all the Nay-Sayers were condemning it to the grave.

PulseAudio is fantastic and was truly needed to take Linux that step closer to a professional operating system.  The only short falls in PulseAudio is the Distributions lacking tools to manipulate channels, especially for recording, like merging channels or choose which one is recorded.
 
i suppose the next reinvent-the-wheel technology to accompany Mir will be "ACLE" or "AGE". whatever that will mean.
 
+Jimmy Call So why not take Wayland and make that a success? With Canonical copyright assignment requirement, I don't see it prosper, as with OpenOffice.org under Sun. Canonical can hold the market and no one will feel like putting effort in improving Wayland and, not being able to improve Mir, progress will be stalled.
 
I suspect they are trying to imitate Apple here. They secretly hope that by Ubuntu 16.04, the only things left of the desktop Linux ecosystem will be the kernel and a few other bits and pieces, with no Linux branding at all. Luckily, I don't think they can pull it off.

It's too bad, to, Ubuntu used to be the best distro.
 
I got turned off from Ubuntu a while back when Mark put out his blog post about Ubuntu not being a democracy. 
 
+Max Eliaser
Are you for real, Ubuntu has the biggest User base out of all Linux makes.

If you ignore important directions you will simply be playing catchup or losing abilities.  Ask yourself in two years when Ubuntu is able to be on most devices, do you want similar abilities with your distro or will you be content x86 offerings.

Microsoft is going down in flames (Surface Rt(Arm)/Pro) because they locked themselves up in x86 and non-open source compilable code to suit other CPU's/SoC's.

Wayland may have promise but it needs to break through with support from Nvidia, etc.  Mir on the other hand will be too tempting for Nvidia to avoid.
 
+Jimmy Call "Microsoft is going down in flames (Surface Rt/Pro) because they locked themselves up in x86 and non-open soruce compilable code to suit other CPU's/SoC's."

- No, Microsoft is going down in flames because they are trying for force the same interface on all platforms and users don't want that!

What makes you think 'Mir' is too tempting for nVidia to avoid?
 
You can use libhybris with Wayland, +Robert Ancell what's so special about Mir?
Since Mir clearly promise to do everything Wayland already does why another project? Canonical weight could be used to convince GPU vendors do EGL drivers for the whole community.
 
This move makes me actually angry. I've always congratulated Canonical for making a very good distro that anyone can use, but now they turn around and shoot the rest of us in the back. I am only a user, but I want GNU/Linux to succeed, not only Ubuntu. I hope Valve gives them the finger.
 
I think there's not a big problem with Mir itself - hey, looking for new ways of doing things is open source by definition. However, there's big cloud over this decision - fact that we already had Wayland versus XOrg split which only started to calm, and that we are not talking about text editors here - it's very critical part of infrastructure. Problems also comes from mistrust and increasing antagonism of Canonical. I fully get they're hear to get profit and I mean them well, but somehow I really doubt this is right way to do this. It's sad of course that recent years have confirmed that Canonical leaving service business model for good.
So my pick is that people are just frustrated and sad to see Canonical go this way. After all, it's have been and still is the major GNU/Linux distribution.
 
+Nathaniel Walker the comment about Windows variants running on non-x86 architecture was a quote from a previous post.
 
+Nathaniel Walker Apple and MS are giant companies with massive resources. Community matters hugely when you're small, and therefore must band together with others to build something competitive. If Canonical has the resources to do this well on their own, then so be it, but if not, they would have been better served by cooperating.
 
+Jerome Glisse Agreed, I didn't say it is correct, that's why it went unnoticed for so long (i.e. it adds a 0.167ms sleep to an already throttled loop).
ja Ko
 
I'am coding in c++ for embedded system but shared_ptr are not good    to speed up applications. :-D
 
+Nathaniel Walker
 I'm glad to hear sense being spoken in this firestorm of folly. Additionally, Canonical's development of Mir does not harm Wayland in any way. They've adjusted their MirSpec article with the following line:

"Please note though that Wayland's input event handling does not suffer from the security issues introduced by X's input event handling semantics (thanks to Daniel Stone and Kristian Høgsberg for pointing this out)"

The Wayland team will continue development as usual and Canonical will continue development of Mir to meet their goals. There should be no ill-will here.
 
as if ubuntu didnt have other problems, they are now developing a display server on their own where there already is one in the pipeline that would work reasonably well... great...
 
+ja Ko What a strange comment. They're not meant to be a speed optimization (aside from development speed). They also are only very, very rarely a speed detriment as the program's bottleneck will usually lie elsewhere.
 
+Jimmy Call Jimmy, pray tell why is Mir irresistible to NVidia as opposed to Wayland? Do you know/see something that I don't from what's been published?
 
+Mattias Eriksson frankly I'm glad C. has dropped GDM.  Here's an exercise: configure GDM to pass a custom option to the X server command line without patching the source code. Bonus points for finding that it used to be possible. Try it and you'll see.
 
+Metta Crawler wouldn't the better way have been to fix gdm? Now they replaced it, and from what I understand, when they did lightdm didn't have accessibility support, no multi stack pam, and no fast user switching support. Some of these things have improved after that, but I still doubt they have all the features of gdm. So while it works for you, I guess all people depending on screen readers had a different opinion.
 
Interesting thing that is. Looking on what they want to do, and what you can do with Wayland, I actually see no technical reason why they wrote Mir except to have their own project in their control. 

I'll bet on a better future for Wayland though, for an obvious reason: Only Unity will support Mir. I doubt that other developers will implement support for Mir. We all know that until we have everything just on Wayland will still be a long path, and I doubt that developers other than the Unity developers will write a Mir port (gnome shell/kwin, projects using the dm directly).
 
+Mattias Eriksson Why doesn't Fedora fix GDM?  They just wrote a patch to set an X server option instead.  Red Hat turned that into a RHEL release instead of fixing it.
 
So when are they starting a new project on office productivity suite rather improving libre office. Why the hell they want ownership of every other
project they use.  
 
+marz marzoug Just like upstart killed systemmd or unity killed other DEs? It's more likely Mir will also support the wayland protocol for compatibility. After all, Wayland is a mostly working solution already.

Anyways, why would nVidia suddenly jump on-board when they didn't after Ubuntu said it'd use Wayland in the first place. If the entire Linux community + Ubuntu isn't enough for nVidia to step up for Wayland, why would Ubuntu by itself be enough for Mir?
 
+marz marzoug  It's always easy to project amazing qualities and feats upon vaporware -- Microsoft does this all the time.  Unity, which is conceptually a much smaller project based on gnome 3 is only now becoming usable.  There's no way any linux enthusiast can look at this as anything other than pointless market fragmentation in precisely the one area -- the graphics stack -- where linux has historically been a failure.
 
Some questions about the revised Mir spec (https://wiki.ubuntu.com/MirSpec?action=diff&rev2=4&rev1=3).
How is it that this:
"we are aiming for a more extensible input event handling that takes future developments like 3D input devices (e.g. Leap Motion) into account."

doesn't contradict this?:
"With respect to mobile use-cases, we think that the handling of input methods should be reflected in the display server protocol, too."

Next, what does this even mean?  Can anyone explain?
"we consider the shell integration parts of the protocol as privileged and we'd rather avoid having any sort of shell behavior defined in the client facing protocol."
 
Frankly, I don't understand all this canonical bashing. OSS is about choice. If canonical choose to use MIR in ubuntu - hey, that's their decision. If you don't like MIR, don't use it. That's YOUR decision. We have seen similar things before - projects competing for the lead like gcc vs egcs (about 15 years ago), gcc vs llvm/clang (perhaps just some months into the future?), kde vs gnome. Let's just see how this comes out.
 
+Axel Howind  Other people have already explained this, but I will do so again, in case you missed this.  The Canonical bashing is mostly about how Canonical handled the situation; namely by telling everyone they planned to use Wayland right up until they abruptely announced Mir 9 months after development on Mir had started.  If you look at the IRC chats, the Canonical developers who were supposed to be working on Wayland just inexplicably stopped making commits to the codebase; then magically re-appeared months later to deliver a big FU to the Wayland people; however sheepishly the message was delivered.  Yes, choice and competition are a good thing, but this is the one area (graphics) where linux has alway come up short.  The tantalizing possiblity of having a single graphics API for software developers to write to was so close you could taste it; and now with Mir, linux graphics FUD has been fully restored.  I'm not saying Mir won't be better than Wayland, or anything like this as I'm pretty sure no one, including the developers has any idea how this is going to shake out.  I'm just saying that -- based on my almost 20 years of experience with linux, IT, and business in general -- the linux community would probably have been better off if Canonical had put those reources towards improving Wayland rather than striking out on their own to do something similar but just enough different to cause problems for everyone who writes software for end users.
 
I think Canonical made the right decision. Wayland simply is not fitting her bill. That is all that there is to it.

I really do not understand why that should provoke so much hate. I think competition is fine, but it should not lead to such bitter arguments. Linux is about diversity, and Ubuntu is on a different path than most other distro's.

I am glad they chose to go for Mir, as this gives them more control over their development process. It would be absurd to wait for Wayland to become what Ubuntu needs. Let's not forget that Canonical expected to ship Wayland two years ago. No company can plan development and production waiting for a part to maybe or not arrive. Projects like Gnome and Wayland do fine but are way to slow and run by people who  have a different agenda than Canonical which is logical and perfectly fine.

Though people are now blaming Canonical to switch to Mir, the same people would blame Canonical even more if they would steer Wayland in their direction. They would accuse Canonical of hijacking Wayland. So Canonical did the only sensible thing and that is to choose another route. For years Canonical was accused for contributing to little upstream, now they will.

Whether Canonical can pull it off is their business, but I have full confidence. Canonical has a very good track record in delivering on schedule. I think what they are trying to do is very exiting and will create a necessary extension form the desktop to mobile devices. The desktop may not be so important in the future as it is now, as lots op people are moving to mobile devices.

Seeing what MS is trying to do it would be a fatal error for Linux not to develop the same kind of integration. I think it is very exiting that we can develop applications for cross platform. Mir will be able to use existing mobile drivers which is great, as drivers is often an issue and I think that Canonical might very well try to make the driver support for the Desktop in such a way that Wayland can use it too.

What I dislike about the Linux community is these over-emotional tirades without any attempt to see other peoples take on the matter. Linux is about freedom of choice. Freedom of Choice is a dead letter if you do not allow others to deviate. I think these destructive comments do more harm than good.



 
 
+Patrick Goetz

It is perfectly normal to cooperate until the decision to break is irreversible. Why would you risk troubling relations by uttering doubts. That is not how companies go about.

"The tantalizing possiblity of having a single graphics API"? Since when is it a tantalizing idea to have one of anything in Linux. Linux is about choice and let the solution win by being adopted. Free market. You ideas would fit Apple or MS, not Linux.

The thing is that Linux developers often regard their projects as babies. I can understand the deep disappointment of losing the support of Canonical. But please lets be a little professional about this.

All these accusations of Canonical trying to falsely blame Wayland. The remarks were not aired by Canonical but by people who are looking for any material that can burn and be used against Canonical. It is all so childish, so immature. Don't get mad, get even. Show us why Canonical made a great mistake in not choosing Wayland. Then maybe Canonical will return. They are practical people, not fighting silly flame wars.
 
+Maga Laan  > Since when is it a tantalizing idea to have one of anything in Linux?

Since you want 3rd party application developers (particularly game developers) to port their software to linux.  Since you want graphics hardware manufacturers to write good drivers for linux.  No big thing.  Just the difference between being successful and not successful on the desktop.

> Then maybe Canonical will return. They are practical people

If this were true, they would have dropped upstart for systemd for the 13.04 release.
 
+Patrick Goetz

The only reason we have games coming to Linux is Ubuntu. The only reason hardware manufacturers will be writing good drivers for Linux is Ubuntu. Not Wayland, Ubuntu. They do not care about Wayland, they already said they would not support it.

Why Ubuntu? Because Canonical is a reliable business partner and has a long term vision. Ubuntu has an excelent track record for delivering on time. Wayland just the opposite, it is more like Waitland.

You are not seriously suggesting we wait another few years for the people of Wayland to produce someting tangible but likely not what Ubuntu needs?

We have been waiting far to long. I am glad Canonical took matters in her own hands. Ubuntu will do it much quicker on her own and the end result will be better suited for what they want.

And who knows, maybe Ubuntu will do it in such a way that drivers are kept out of the display server, which would be a much better approach.

Open source is not monolithic development you want, it creates a free market. It offers choice. If Wayland wants to be viable it will have to proof herself by the product.

And it is wrong idea that more could be accomplished if all people work together on one project. We have seen that with Open Office. Since it was split, development goes much faster. Having all these in-fights between people with different ideas does not help progress.
 
+Patrick Goetz But wayland/MIR would just be the display server. As an application developper I wouldn't even bother if it's wayland or MIR. The "single graphics API" would be OpenGL/ES. Or you just stick with some toolkit like GTK or QT. These toolkits (and maybe others) will have to be changed, yes. But honestly, when was the last time you did graphics programming using xlib directly?
 
Yes, I understand this.  The current situation in the world of PC games as I understand it is that game developers develop for MS DirectX because the DirectX drivers supplied by graphics cards manufacturers are more robust than the GL drivers.  And because most games are being developed for DirectX, the graphics cards manufacturers devote most of their resources to these drivers.  A vicious circle.  You can write OpenGL/ES programs all day long, but if the drivers are buggy, no one will be happy.  It's precisely the driver issue which is of most concern here.
Add a comment...