Shared publicly  - 
Lennart completely misses the point.

The difference between us is that I respect his right to create whatever software inspires him. I may not choose to use it, but I'm not upset that it exists. It may be an NIH, but scratching an itch is what leads to a lot of great stuff in open source, so good luck to him.

What is hypocritical is his tendency (and that of many of his colleagues who rant about Canonical) to reject someone else's right to do the same. Constant partisan critique OF CODE purely for existing is a bizarrely political approach.

I think it's great that Lennart cares enough about init to make a new one we can all evaluate. I don't buy his constant slagging of upstart, because putting another project down is a really bad way to try to make people like your own, and his commentary on upstart doesn't match what the best people (who care a lot about quality and experience, so much so that they make the distro that most people choose to use) tell me.

I think it's telling that his harshest criticism of upstart and Mir has nothing to do with the code, but rather the company it comes from.

Ask yourself: do you want more innovation? Why would you trash someone's effort to improve on ANY aspect of the free software stack? And why would you pick up a pitchfork and join a mob that is just trashing on someone writing free software? Doesn't that strike you as intensely stupid behaviour?

I welcome critique of any code I write or any code I'm responsible for. I reject the BS slagging of code just because it comes from a company with different values to mine. And that's how it should be.

Summary for the sound bite crowd: pursuing your own interest in creating free software is a GOOD thing even if something already exists in that space. Most of the great stuff we use today, including upstart, python, go, mongo, OpenStack, etc all we're created when there were existing things in roughly that space. NIH is innovation too, so we should celebrate it when people create stuff they want. We should be suspicious when people who do that all the time reject someone else's right to do the same.

Chriss Kalogeropoulos's profile photoChristoph Wickert's profile photoMunawar Cheema's profile photoRyein Goddard's profile photo
Like if sound on Linux where so monolithic or one-project-only thingy, let it be, nothing will happen...
If the arguments I've read so far were applied generally across projects we should all be working to improve XFree86 and, Wayland and Mir should never have happened, for example. 
I can only say ¡GRACIAS! all ... community businesses you support software libre... THANK YOU... because ye the world, technology, human civilization better, more equitable. My opinion of MIR is clear... Wayland has improved because of it and this is very good for all, even for the MIR project and Canonical. And again by extension the world, the planet. I ask for calm, peace and love. A strong greeting from Spain. :)
It is interesting however to read your post again but replacing systemd with Mir and upstart with Wayland in mind. It's the Quid Pro Quo approach that goes both ways.

And where Lennart doesn't have your superpowers as SABDFL it still remains a fact that systemd is used in quite some distributions that are not from the same company/project/group.

So it seems it's not only politics but also code quality/matuity.

The way Canonical/Ubuntu treated Wayland (first commit to using/developing it, remain silent, switch to something completely new) is not comparable to the way for example Fedora/Red Hat decided on upstart v systemd. I hope we can agree on that.

As you can see, I really try hard to stay reasonable because I also believe in choice and competition of code. I can ignore the political rants. And I honestly believe you can too. That's why I find it so hard to understand why you seemingly feel that urge from time to time to lash out at others for doing whyt they do. You must know to what it leads. Hence you do it on purpose. Don't. But if you do, please be fair to others and accept that they will take the bait and lash back. ;-)
I agree with you that much of the outcry over what you folk do is over the top. Then again, perhaps the Tea Party thing was asking for it ;-)

But in dutch we'd say Lennart has some "butter on his head" (comparable to the pot/kettle saying in English). Surely the criticism has to do with the company it comes from, and that is unfair.

At the same time, just the fact that code is GPL (or equivalent) licensed doesn't make it a contribution to Free Software. Linspire and Xandros wrote lots of GPL code, which is now bitrotting away somewhere. As +Aaron Seigo said at some point: are we trying to develop GREAT Free Software or just volumes of it?

Code which is just put up somewhere as "you can use it if you want" is not as valuable as code which is actively contributed to an upstream or in general developed collaboratively.

I wrote about that a while ago:

and I think you're smart enough to understand that.

That doesn't mean you always have to collaborate - companies have deadlines, have work to do. And collaboration should be about mutual benefit, not about giving away. And communities (especially some people in them) can be unfair. I understand that at some point, Canonical just gave up on the collaboration. Or at least, that is what it looks like on the outside...

That's sad and bad, both for you and for Free Software.

So I'm hoping you'll come around. And I am hoping you'll realize that without collaboration, the term "Free Software" means nothing in the end. You won't get the benefits that come from working together in a productive and constructive way. And you're not participating in building the shoulders for future generations of hackers to stand upon.

Enjoy your weekend ;-)

And yes, there are specific things in the infrastructure you share with especially +Qt Project and +The KDE® Community that we, Canonical and those communities, could work on. I know some patches flow from your company to Qt (yay, that saves you maintenance work!) and I think there's room for more. Doesn't have to be huge, just practical "let's do this together so we both spend less time on it".
I do think there could be more articulation of what positive benefits there are to Mir though. It does server side buffer allocation which is nice on ARM platforms. That isn't a user perceptible benefit though. At the moment xmir is very nearly as good as not running xmir, the smallness of the delta is a hugely impressive feat - but it doesn't appear to have a benefit over xorg. I understand native unity 8 on mir will be good in some way, it has the potential to be better than xmir but I don't know how. I am pretty close to the project, I ran mir on the day it was runnable, but I don't know what user perceptible coolness it will bring. Do we get the desktop cube back? Wobbly windows? A 3d stack of windows with rendered shading and blurring? Apart from server allocated buffers I am not aware of any selling points to Mir, I am sure they are there but I don't think they have been articulated.
+Paul Tansom actually, wayland is written by the Xorg developers - it is basically X12. They just gave it another name. So while Mir is a new project, Wayland is not. So if those arguments were applied, we'd work on Wayland.
"It does server side buffer allocation which is nice on ARM platforms. That isn't a user perceptible benefit though."

Well, if you're an ARM system user, it may be very perceptible. Or not; I don't know myself. But unless there are user-visible performance improvements to be had there would not have been much reason to do Mir in the first place.

Me, I think it's simply counterproductive to slag Ubuntu for scratching an itch. I don't know if I will ever write anything specifically for a Mir-enabled system (seems most differences are hidden away behind our toolkits for app writers) but I can't see its development as a negative either.
Ya know Mark, basically every other Linux distro has embraced systemd other than you, even gentoo, whom also made OpenRC way before you funded Upstart.
Thing is, they gave it a chance and let the winner slowly win instead of locking out the competing technology. I personally love systemd now even if it was a slight learning curve to grok.  You on the other hand (Conoincal et-al) are far too stuck-up to notice anything technically positive nowadays.

Let me ask you this: Do you even know how Mir works? Technically & architecturally; could you even remotely explain Mir's control flow from device input -> graphical output?
+Alan Bell Similar buffer allocation would also be possible in Wayland, just nobody has wanted to do it that way yet. That was one of the misconceptions cleared up in the early days of the "why not Wayland" Mir discussions ...

+Jan Moren The issues for many of us have nothing to do with someone scratching an itch. Mark likes to frame it that way, perhaps because it is then easy to discount the dissent, but it simply is not the issue.

The idea that it only affects desktop shell developers is similarly hokum.

I've enumerated the various concerns in detail elsewhere, both recently and in the past. Unfortunately when the conversation format is "I'll write my bit of text over here, you write yours over there" it is extremely difficult to keep things in context, avoid mindless repetition and have people address real issues rather than look for the easiest rhetorical defense available to them. 

This is why I called for a (virtual) 'face to face' video discussion to be broadcast publicly on the matter.

That people such as Mark seem to not want to sit down and have an adult conversation in a way that could start to clear the fog on these issues is particularly concerning.
Cmon, one person NIH'd systemd and pulseaudio, another NIH'd Mir and Upstart. What is the big deal? Let people innovate, indeed.
I have never heard of blog posts helping anyone discussing an issue, or get different parties to agree on something. Is it so hard to pick the phone or do a private video chat so both "parties" talk whatever they have to say and move on for the best of all community?

There seem to be some obvious grudges from both sides, but there is plenty of room to clear misunderstandings and move on.

I'm a complete fool and ignorant if you will, but I see Wayland people are a bit mad about statements made about their project in a recent past in a wiki, and Mark believes they just want Mir to cease to exist, so maybe some things were not done the best way when Mir was created, is that so hard of an issue that you both can't go in peace privately? 
+Daniel Ferradal While it is tempting to put this down to hurt egos and small minds, there are serious implications and unaddressed problems which are inherent in a Mir/Wayland world that have nothing to do with politics or ego.

I appreciate your call for calm heads, but ignoring these sorts of issues is not having a calm head it is sticking our head in the sand.

Where I do agree with you is that I'd love to see people move beyond the grudges and apparent anger they are holding on their shoulders. Maybe then we can have some real discussions that can lead to actual improvements in understanding and, heaven forbid, improvements in our roadmaps.
+Mark Shuttleworth if I may be blunt with you: focus on what you do. The quality of your products speak for themselves. Good work. No need to listen to cry babies. Focus. 
+Aaron Seigo yes, I saw that Wayland could do server buffer allocation, but until that was cleared up that was a plausible technical reason for doing Mir. I just haven't seen any others, and I would like to.
+Alberto Milone Three things:

* the impact on driver writers, desktop environment developers, applications developers, support teams and ultimately the end user of multiple display systems, and how we can resolve or at least mitigate that

* clarifying the true benefits of Mir and Wayland (not just relative to each other, but in general) as there is a HUGE amount of misinformation out there about both

* attempt to find a commonly agreeable framing for the topic; the NIH claims, speaking on behalf of projects one does not represent, the rhetorics .. it's not helping anything and we can do better

I don't think it works to have just Mir or just Wayland represented, as has been happening in too many blog posts and online forums to date. An open discussion with various POV and agendas represented offers a much better chance for pushing things forward in a positive manner.

I don't think it works to do it behind walls of text. It is too easy to slip into selective response and rhetorics, and text is not the easiest medium for everyone to engage in.

I also don't think it works to have an online shouting match where 9.5/10 people have very little background (if any) in the matters at hand.

I also think moderation and having a 3rd party to help hold the context of the discussion as it progresses would be immensely valuable.

Does that seem reasonable to you?
+Jos Poortvliet +Aaron Seigo from what I understand (please correct me if I'm wrong here), the purpose of such a debate would be to "discuss Mir and cast aspersions on those who disagree in the process". My question is: why do you even need to discuss the existence of a project or the decision to choose it over the already existing alternatives? What's the point?
.. and one other item: I don't think it works to do it in private. The Free software public has been handed a woefully confusing bag of conflicting statements of varying degrees of accuracy from people who are involved as well as people who are not involved. This needs to also be an exercise is setting the public record as straight as we can manage to.
+Mark Shuttleworth So you "respect" +Lennart Poettering's "right to create whatever software inspires him", but when he does, you complain it "is hugely invasive and hardly justified" and he "NIH’d just about every important piece of the stack they can get their hands on". That doesn't sound like "respect" to me at all.
+Aaron Seigo "* the impact on driver writers, desktop environment developers, applications developers, support teams and ultimately the end user of multiple display systems, and how we can resolve or at least mitigate that"

This is the only point I can personally see a use for. As long as the answer doesn't end up being "drop X and adopt Y instead".

"* clarifying the true benefits of Mir and Wayland (not just relative to each other, but in general) as there is a HUGE amount of misinformation out there about both"

To what purpose? Can't we live with the fact that they have a different approach and that they are trying to solve different problems?

"* attempt to find a commonly agreeable framing for the topic; the NIH claims, speaking on behalf of projects one does not represent, the rhetorics .. it's not helping anything and we can do better"

I definitely appreciate your concerns, especially as to how to resolve this while keeping it civil. However, I don't think there is anything to resolve here.

If a project such as Ubuntu or Fedora or Gentoo, etc. decides to switch to a certain component (be it the init system, the display server, the audio server, the desktop environment, etc.), whatever the reasons behind this choice are, they should be respected. Even if it were a matter of personal preference, or of having complete control over a home brewed project, or of technical superiority (a kind of subjective topic) of one piece of software over another, I don't think others should be upset.

Am I being unreasonable?
Respect does not mean you're not allowed to criticize.

Btw, why is NIH bad?
This is so cool, just like watching daughter[1]'s kindergarten class discussing who's clay thingie is the best in class.. :)
+Jan Wildeboer RH adopting upstart in RHEL 6 then switching to systemd for RHEL 7 seems a perfect analogy of our choice in regard to Mir. But there's a key difference - RH put upstart into production before switching, wayland hadn't yet been formally adopted.

Yes, it was a less than ideal result: we looked at the problem early and decided we liked what wayland was doing, but then wayland went in a direction that gave us cause for concern and the graphics team at canonical decided they could do something lean and clean called Mir. I think the teams initial desire to justify that by critiquing wayland was flawed both in plan and in detail, but that team has delivered what they promised in Mir, which is a lean graphics stack that we can use across phones and desktops. It would have been better for me if wayland had been right for us as I could focus our investments on the user visible parts, but I think it's great for free software that we're willing to put up the alternative.

While we erred on initial critique of wayland, we corrected that. Lennart persists with fud about upstart.
+Jos Poortvliet that statement seems to imply that Wayland is acceptable because the developers are the same as That leads to the assumption that Mir is not acceptable because it is not developed by those same developers. Only some developers moved from XFree86 to, so should that have been accepted? Where do you stand on whether Gnome should exist? Should it have been stomped on the way people seem to want to do with Mir?

I doubt that is what you mean, and I have a tendency to play Devils advocate, but I can see why +Mark Shuttleworth and others get fed up with criticism that seems to say that, in spite of the philosophy of open source software, they should tow the line and do it the way everyone else does. If you take that to extreme we should all be using Windows ;-)
+Mark Shuttleworth Thanks. That indeed does help me in understanding the whole thing. Even though I might not agree on all counts, it is a better way of communicating.

Ultimately we all want the same - more free software and less proprietary lock-in. Let's not get distracted too much from that, shall we? ;-)

A cold beer is always waiting for you and your friends in Munich, BTW!
+Mark Shuttleworth i have loved the fantastic integration delivered by canonical over the years, and have never bothered about the lack of hardcore upstream work. Upstart was a first welcome project, unity gives me mixed feelings, but Ubuntu touch is a fantastic direction IMHO. What worries me is that canonical starts to be very isolated and will have to maintain a huge stack by itself, as you refuse to use the same parts as the other big Linux players. Of course it's your right to use a different stack, it is just worrying for users as it is a big move from "very little upstream work" to "let's reimplement a full graphic stack and maintain our own init". It's canonical right to do so of course.
What is not right is to spread false facts about Wayland as "technical reasons" for Mir, than pretend that the rest of the world is bashing canonical for political reasons (and using this to reject technical discussions).

As simple user, I wonder if graphic drivers will be properly shared between mir and Wayland, if apps will run flawlessly on both Wayland and mir and so on. If canonical wants to do it's own stuff, that's fine with me, but if it leads to more incompatibilities, I want to know it and I will probably avoid advertising them (which would be a pity as I really look forward to your "convergence" stuff, as former maemo user).
Linux is filled with really abrasive people writing amazing code. Leonart is one example. Stallman is another. Hell, Linus, himself, isn't exactly a cuddly bunny. 

But just because the person may be a less than a stellar example if humanity, that does not affect the quality of their code. Systemd is a great system, and I have tried a few init systems, init, sysV, upstart, and systemd. 
I get that Aaron defends two similar projects is a waste of programmers effort, as a lot of the code is doing the same thing and is reinvented.

But, being really pragmatic about it: I believe that sometimes we need to start from scratch. is an amazing project, but it just doesn't fit well in our current context. I was really excited about Wayland when it was proposed. But companies need to make a profit, and it's unfair to the community to prioritize anything in the development to help a companies project, as it's unfair to companies that has contributed a lot to open source projects, to have to wait for other project leaders to decide whether some feature is a priority or not. Canonical has a vision about what GNU/Linux can bring to the personal computer changing paradigm. It might be right for everyone, maybe some people, or not right at all, but if their needs conflict with the community's vision and priority list, than I think it's only fair that they get out of the way and do it on their terms. It's GPL, so if they get some breakthrough or great idea, we'll just implement it in Wayland, without "unnecessary" Mir stuff. Of coure it slows down the development of both solutions, but too many discussions about the roadmap do that to.

Like, OpenOffice and Evolution mail are examples of projects that are just too big for a community driven overhaul. And we know they badly need it. So when people decide to act and have a go at it from scratch, it's natural to see different ideas about how to do it.

In the end, it's more productive to spend time coding, instead of arguing. Someday, there may be a merge, or a realization that one of them took the wrong path, but that's really hard to predict. Alternatives are good. Politics and personal grudges are really bad. 
None of this really matters in the long run. If Mir is successful along with Ubuntu - and there is every reason to think that it will - it will be supported and not ignored. People can mash their gums all they like about whether it should or not exist and not change anything.

Now if any engineers out there would care to make a substantive contribution to the debate as to the relative merits of Mir then I am all ears.
+Alberto Milone As I've already stated, I think the utility of these textual back-and-forths featuring an all-comers format have ceased to be useful in discussing the issues. As such, I'm not going to torture the matter further by getting into the details here.

The questions you raise are precisely the sort I'd like to discuss in a public, face-to-face format, however.
"I get that Aaron defends two similar projects is a waste of programmers effort,"

That is not my issue with Mir.
+Aaron Seigo
 Why you don't post a message in the Mir mailling list to speak about this technical points ? A public debate seem more like politic fight and I think it will not be really helpful.

Speaking directly with people engaged technically in Mir about this problem, open the discussion can help a lot.
Sure +Aaron Seigo, I get that it's much deeper than just that, but ultimately, in practical terms, that's what happens when we have lots of projects with overlapping purposes.

Questions about vision, design and development priorities are too open, very subjective.

I do agree with you, though, about the limited benefits of this back and forth internet discussions. Also, driver support is a serious issue that the Floss community has fought long and hard to advance on with vendors, and should not suffer from this differences.

As a user and app dev, I feel lucky to have guys like yourself, +Mark Shuttleworth and others trying to make better software. Thanks for caring. 
Canonical and Mark have decided to go with Mir, other people decided to go with Wayland, publicly debating about Mir/Wayland's flaws or benefits is probably useless at this time, +Aaron Seigo  and other people who are worried about compatibility issues should continue to discuss with +Mark Shuttleworth and Canonical and make sure that this split will impact minimally the community instead of asking for public debates (I'm skeptical about public debates, I never see those ending with a person declaring "I was wrong" or with a majority of the people clarified that one side is clearly right, looks to me like a waste of time and anything fuzzy that would be said in such a debate would be fuel for even more useless debates and bad blood). 

I think that being practical about it should be more important that trying to demonstrate that you are right and the other person is wrong. In a sense it's irrelevant if Wayland is better than Mir or the other way round, Canonical chose to develop and use Mir, they probably have their reasons, I can understand why a company that has specific goals and deadlines would want to control the development or some parts of their software stack. I can also understand that Aaron and other people feel that Mark and Canonical employees misrepresented some aspects about Wayland, but that is not that important, Mark and people from Canonical have the right to be wrong like any other people (and Mark already admitted that there was some misinformation put out about Wayland initially) it would be nice if all parts would restrain from making comments about what other people are doing and concentrate on what themselves are doing, the code should and will eventually speak for itself.

I can also see the point that some people are afraid that the code will be irrelevant, that some other extraneous items like marketing and popularity would eventually decide, that's where I guess most of the Ubuntu-hate is coming from, people are afraid that whatever mistakes Ubuntu makes would impact the community. That might be true, but I think it's an overblown worry and Canonical and Mark still have the right to make their own choices and mistakes.
+Frank R. Schmidt NIH is bad if it is the sole reason to do something again.

And NIH'd software I would call the stuff that were solely done because of NIH.

pulseaudio, wayland, upstart, systemd are not NIH'd.

I am not sure about Mir. But, if in doubt, I would say it is not NIH'd either.
I concur with +Mark Shuttleworth . To be frank, I started using Linux just because of Ubuntu. I was introduced to it when I was in my pre final year of my Engineering and till date I am still using it. I am really thankful for you can Canonical as a whole.
+Mark Shuttleworth  IDK man ....... maybe there is a little truth about some of this.  As you and other have asked many people to walk away from the community.  Please don't take this the wrong way but flame wars help nothing. As a sound man I listen to steely dan when doing sound checks because they have great recording quality.  one of the 1st songs that I listen to is Rikki Don't Lose That Number. It is a alright son do you like it ?  
+Sebastien Lemarinel Of the three things I listed as topics, and the reasons I gave for wanting to discuss them publicly, not one can be resolved usefully by posting to the development mailing list.

I have spoken with Mir developers, as have many others. It's not like we're avoiding each other :)

+Adrian Mihailescu Speaking directly with Mark and Canonical has been tried. The current situation is what has resulted. I believe that doing the same thing over and over but expecting different results is insane.

"Aaron and other people feel that Mark and Canonical employees misrepresented some aspects about Wayland"

That isn't a feeling, it's a fact. Mark has owned up to that fact. Unfortunately, it was not the only such case of talking of behalf others that occurred .. but most importantly, it is indicative of the root social problem.

(Note: there are non-social issues as well .. this is not a monodimensional issue)

" it would be nice if all parts would restrain from making comments about what other people are doing and concentrate on what themselves are doing,"

That would be a good first step. Mark's blog entry from the other day demonstrates there is a long ways to go ...

"the code should and will eventually speak for itself."

Just like how Betamax won out over VHS.

"that's where I guess most of the Ubuntu-hate is coming from,"

For me this is not about hate; I can't speak for others of course.

"people are afraid that whatever mistakes Ubuntu makes would impact the community."

It's not about mistakes, but decisions. And yes, the decisions actors make within the Free software ecosystem impact it in correlation to their influence.

" That might be true, but I think it's an overblown worry and Canonical and Mark still have the right to make their own choices and mistakes."

Nobody is saying they don't have that right, and I don't understand why people such as yourself keep making statements about things nobody is suggesting.

What I would like to see is proper information in the public about this. I'd prefer to do that with Mark, even if we don't agree on all things. I think that would be a better way to go about it than the shouting at people from blogs approach.

People are woefully misinformed on the topic, and I see that in the comments being posted in this very thread.

How great would it be if people had a good handle on the situation so each can make an informed decision.

Can you help me understand why that would be a bad thing?
+Mark Shuttleworth  I have seen you say "...then wayland went in a direction that gave us cause for concern" before but what does this mean? To me the creation of Mir mostly seems to be an accident. You or your team didn't know enough and went with your own thing. If not that it only leaves a "political" motive for it's creation. It doesn't have any technical reasons anyway.. not that I and most other know of. If that was the case it would have been mentioned at this point. Which it hasn't.

I would be much less concerned if Canonical didn't use CLA's. It goes against the purpose of GPL. Mir would need to fork and become OpenMir for others to be able to use it properly. Or have I missed something when it comes to this? To me this feels like Canonical is against openness. Why is the CLA's important for you? I would be less concerned if CLA's was not used on core components.

I am scared this will make a split in the Linux community.. with two different desktop Linux (so to speak). Is Mir really that important for this to happen? I would be lightened if I am wrong in this. This is the greatest fear I have. If this would not happen and app developers don't get extra work and such.. then I would have no issue at all with you taking the Mir route. It feels like Ubuntu is piece by piece becoming more like Android.. which is not app compatible and have a complete different community.

Ubuntu have been important and awesome in the past. I used it between 2005 and 2010. During this time it was the best distro.. but something changed after that. If that reason could come out in the open. Then maybe everyone could accept it more.
I have stayed with Linux because of Ubuntu and I have not been let down yet. I will only use Ubuntu...only Ubuntu. It does what I want just right. So what if the various Linux tribes don't like what you are doing. Linux gives us the freedom not to have to justify that to naysayers. Ubuntu "just works" and that is good enough for me.
Tech aside, we the users have been waiting for Wayland since 2008. The original Wayland "manifesto" said something along the lines that there was too much cruft in X11 and it needed to be gutted to be lightweight with no complicated APIs [1,2]

Five years later, I would argue those lean and practical goals are out the window, and when Mir showed up as competition, Wayland put the pedal to the medal and actually released something to show they weren't last year's vapourware?

Is Mir the best solution? No comment. But, at least it's goal oriented, which is what Wayland was supposed to be. At least Mir is perspective.

All free software is good period.
+Aaron Seigo 

> Just like how Betamax won out over VHS.

And now we use neither, but somehow the technology and concept of video broadcasting has improved exponentially, barely resembling anything like it's analogue past.

Herein lies is the problem. You see this as a competition, Canonical sees it as trying to figure out how to run a successful business, and the future won't care so long as someone ships something.
Not all users are involved enough to know there are options outside of Ubuntu and its default software. Many Ubuntu users I've met don't even know there are other desktop environments.

So when you wind up with a majority of users, and many of those users are unable to evaluate the alternatives, what you give them is what they'll live with. And I'm sorry to admit that not all free software is good- some of it is poor quality and should not be used.

So we all have a responsibility to seek the best solutions and innovate in a meaningful way, because our users depend on us. So long as you go forward in the spirit of making things better for everyone, without unnecessarily stepping on anyone's toes, hardly anyone has right to complain.

But you must recognize how much influence Ubuntu has over the Linux desktop at this moment in history. When decisions are made that may cause more than a passing dependency issue, you need to address peoples' fears rather than call them irrational.

We are afraid; Afraid that what you're doing may inhibit our efforts at simply maintaining basic compatibility. You don't have to agree with us, or even care what happens to us when commercial vendors see 'Ubuntu' as 'Linux'. But please address the uncertainty that non-Ubuntu users are facing, rather than thinking to yourself that it's just our ego on the line.

An arm extended in friendship and understanding has a lot more value than arms folded in protest.
+Aaron Seigo said "Just like how Betamax won out over VHS."

Actually there are people who make a good case that VHS was a better overall product, just an example: 

"How great would it be if people had a good handle on the situation so each can make an informed decision."

It's a complex issue and fairly technical, I doubt most of the public will get a handle of the situation even after a public debate. And whom do you want to convince, the public or Mark? If talking to him privately didn't do much good, how would a public debate solve the problem? Canonical would still go forward with Mir. I think I start to understand Mark's using of the "Open Source Tea Party" term, it started to sound a lot like "But Obama doesn't want to negotiate with us"... I don't see what's there to debate or negotiate, to me your proposal of public debate is very akin to political grandstanding than something meant to fix anything.
+Mark Shuttleworth 'putting another project down is a really bad way to try to make people like your own'. Quite.

+Alan Bell +Aaron Seigo Wayland does not, and never ever has, mandate client-side buffer allocation. The Raspberry Pi Wayland EGL implementation (for which all the code is public) does server-side buffer allocation just fine, with no changes to Wayland.

When Mir was announced, I was on-site with a customer, watching a triangle rotate on a Wayland compositor using server-side buffer allocation on ARM (again, no changes to Wayland itself required) while reading Google+ on my laptop. Again, no code changes required.

It's a(nother) total red herring.
I will never understand the hatred of any project that doesn't go with the flow. Isn't this the entire point of Free, Open-Source Software? It's certainly not a soapbox for everyone who doesn't like Ubuntu.
I'm still confused by this whole mess. I would love to see +Mark Shuttleworth and/or the guys from both projects work together and straiten out the public and give us reality instead of the constant FUD which seems to be everywhere. Here, I'll give you my fears so you can address this.

Mir does it one way, Wayland the other. Both offer a backwards X11 thing. Developers, wanting to get their projects out there must maintain more than one track of code to have both, so instead rely entirely on X11 mode and never bother to take advantage of any other features either offer, thus basically negating the point of having them in the first place.

That is my fear. That's what this FUD causes. I do not understand what's going on beyond the fact that only Canonical seems to be interested in Mir, which tells me there's yet another reason to be frustrated with Ubuntu. That's not what I want. Is that a very simple look at it? Yes. I'm very likely way off. I'm not alone. What is going on with this? I want the real info and to have some clarity.

Please dispel the FUD! All this post says "Let's just agree to ignore each other" and that is so backwards from what needs to happen.
Ubuntu did and does a great favor to Linux by being (one of the) most user-friendly distributions, and making things just work. For that, I respect Canonical a lot.
However, my biggest concern on this mess is regarding device drivers, which is already a very complicated subject for Linux. It could get worstby having things like: "NVIDIA supports MIR, but not Wayland". Or even worst: "NVIDIA supports officially only MIR, but Wayland may or may not work well". That would be the greatest disservice Ubuntu / Canonical could do to Linux in general.
If not, I'm totally fine with both display managers, and wish MIR great luck.
+Luis Fernando Planella Gonzalez
 I think that's the legitimate concern that many of these conversations aren't getting around to, for various reasons. It's become so politically charged that the real consequences of Mir adoption aren't being evaluated rationally.

That being said, there's no reason another distro couldn't be just as easy to get set up as Ubuntu. Of course, the main reason Fedora or SuSE aren't suited for this is due to ideological objections to closed software (and some legal reasons concerning codecs and fonts). As we have rallied around Ubuntu for years now as the savior of Linux, a lot of that passion can't be easily displaced. So you see Ubuntu derivatives rather than new projects that try to do things better from the ground up.
+Jack Gandy you are wrong in this assertion. MIR was defined since the beginning as being focused on Unity. Maybe things change with Unity 8, as it will be based on Qt5, but currently, it is a hassle to install Unity 7 in other distributions. That's why is not in the default repositories of any of them. Also, if you read the comments after the 13.10 release, you'll see that enabling the Gnome 3 PPA breaks Unity, which reveals some fragile balance to be maintained (in terms of dependencies) in order to Unity to work...
+Luis Fernando Planella Gonzalez
 Oh- I never supposed anything like that. I'm only saying it's possible another Linux distro could easily replicate the easy out-of-the-box experience Ubuntu is lauded for, as many of the elements of this exist in isolation. Arch's amazing stable repository and optional AUR, new app centers for Mint, GNOME, or elementary, and a strong focus on design all exist. Until now Ubuntu's the only distro that seems to combine these things for users, but I'm sure another distro could.

I'm fully aware that Mir and Unity are Ubuntu-specific. Those aren't necessary for a simple, modern user interface on Linux- I'm sorry if I didn't write concisely enough to make that clear.
I agree with you, but please dont refer to them as Tea Party, There are many Tea Party members and supporters who love Ubuntu and use it every day including me.  Tea Party believes in financial responsibility and freedom.
complete agreement, choosing/preferring something else instead of one thing is completely okey and on that no one should have any problem. 
Both of them (new and the older one) are best, choose what you think is better with you or for you. Period.
One thing I never see discussed in these brouhahas is the different "psychology" that I believe underlays many of Canonical's criticized decisions. Basically, if you have aggressive schedule goals, as a company you put your schedule at great risk if you use open source products that need to be customized to work in your context, without having any reason to believe that the project involved will work on the priorities you need at the pace you need. Sometimes, NIH makes achieving schedules possible, because it gives you a level of control you need. 

Part of the issue is how it's considered "rude" to defacto-fork a project that doesn't have an upstreaming pace that supports your schedule needs -- just look at how much heat Google has taken over having a custom kernel. If I were a Canonical, I would think twice about making my projects critically dependent on e.g. Wayland, just in light of this factor. 

What can defuse all of it is eliminating the 0-sum thinking, and moving to a abundance mentality ... Canonical not using Wayland doesn't steal something from the Wayland community (0-sum), it could very well open up new ideas that Wayland can later benefit from.  Or not ... so what?
+David Bucci I think a lot of the hostility comes from people not agreeing with that exact justification. That may be why you don't hear of it very often- many people are opposed to what Canonical's doing outright, solely because it exemplifies why we prefer not to rely on for-profit organizations for software.

There is one ideology among the community- doing it 'right' is more important than doing it quickly, even at the cost of marketshare. People may not like it, but it's a dominant concept that people aren't going to let go of overnight. So the justification of, 'well, we need control to do this quickly with the features we want' merely suggests an unwillingness to collaborate from the psychological perspective of these developers.

My real issue with this situation is that, regardless of whether Mir is any good, there's plenty of social upheaval going on that could be avoided or remedied. This adversarial mindset is starting to drive contributors further apart, which is an unnecessary side-effect of Mir's creation.
+Jack Gandy ok, something for me to think about ... but to clarify, I'm not suggesting "do it fast" is the justification, I'm saying "_do it on time_" is, when "on time" is defined as on a schedule that lets us succeed as a business. I think you get that's what i'm saying, but wasn't 100% sure from what you said. If "open source" has to mean "not caring if it's on time, only caring that it's 'right'", well ... that removes everything but as-is use from the practical business world. Stating that as a core ethic comes pretty close to saying "open source development shouldn't ever be involved in business".

As to not doing Mir-like things to avoid social upheaval - that's a little post hoc ergo propter hoc, imho. My core point is that the social upheaval comes from people applying zero-sum thinking to things like Mir. I'm saying that the better solution is for the people "upheaving" (so to speak :-) to stop thinking zero-sum. Not sure if that came through clearly in what i wrote.
+Mark Shuttleworth if you are determined to push Ubuntu forward to more markets than the other distros have been (phones etc), why do you care criticize and not just follow the Google way ? They forked almost the entire user space and for many years they had a half finished product full of bugs but at least they did not blame others for their problems which is what you are trying to do. Do you remember the Compiz fiasco and how your team fucked up a very nice project? They admitted that the code they wrote was a hack and unmaintainable and dropped it. Are you sure the same thing is not going to happen with Mir, which is times 10 more difficult project? I dropped using Ubuntu because the desktop has become terrible, worse than windows and the efforts are focused on system bits instead of creating nice applications for user space. How about implementing an office suite like libre office (which sucks) but much more modern and better? This looks like a manageable goal for you people
+David Bucci
Oh yeah- I agree. I don't mean to say that Mir necessarily causes this upheaval- I'm saying that the way the conflict is being treated is causing unnecessary external ripples that could be avoided on a social level. I'm not saying Mir itself is the cause, and that people should just pack up their ideas if they don't coincide with the current direction of things.

So long as we can go forward with the assurance that Mir won't sabotage Wayland on the compatibility front (I'm not sure if that's really in Canonical's control, however), then I agree that it's just an irrational 'you did something different, therefore you're taking away from us'. I think it would be naive to assume that Mir's impact won't be harmful at all to other projects, since Ubuntu has a great deal of influence. But I'm hopeful that it isn't harmful, although again, I'm not sure if Canonical can guarantee that. They can't tell Steam, for example, to please focus on Wayland and Mir equally- companies may ignore Wayland simply because it's not used by Ubuntu. That's the main fear, from what I can gather- extra work for developers, which can cause favoritism for whatever is considered 'popular' (we all know this, as we suffer from Windows's popularity).

One more thing- the doing it 'right', not shipping features prematurely 'on time' ideology isn't the predominant open source attitude. It's just one major strain of discourse I've noticed. Many projects find a balance between pushing deliverables and keeping things stable (like GNOME's app 'previews'). Sorry for the long comment- I hope I addressed your points adequately.
+Chriss Kalogeropoulos Why don't you implement a new office suite?  The fact is, it's extremely difficult, which is why StarOffice,, LibreOffice, and many, many others have produced perfectly good office suites, but when the unit of measure is "pixel-perfect compatiblity with every version of Microsoft Office, even though every version is different", there's no way to succeed.

I have two points to make.  The first is that these things are not simple, or they'd have been done years ago, and because of this, someone has to decide where to dedicate resources.  LibreOffice is working on an office suite.  Wayland is working on a powerful, flexible Linux-only display server.  Mir is working on a powerful display server that won't suffer on ARM platforms.  Everyone is working on what they think is most important at this time.

The second is that there is no "you people" in Free Software.  If you want to help, choose your favorite code editor (there are dozens and dozens) and help.  If not, feel free to try to form informed opinions but Free Software is a meritocracy, and those who do the work get to choose their assignments, not the peanut gallery.
+Nathan Haines
Indeed. Not to mention that people will work on what they're passionate about. You can't tell people to work on something like a new office suite simply because it's 'important'. Developers will work on what they believe is important, so it's more effective to find someone who cares about office software than to convince a distro to move its resources away from what it does best.
+Nathan Haines Nice comment Nathan !!! But keep in mind that i am a user not a developer (although i work as a developer in my day job) and i am not interested in developing an office suite or any other suite. I am just expressing my opinion on a public post of +Mark Shuttleworth and that's it. Sorry if you did not like it but i felt the same way when my favorite OS (ubuntu) became a total mess due to bad management decisions. You might argue or dissagree but the truth is that Ubuntu popularity is dropping fast and the reason is not Wayland or SystemD. And i  know that development of an Office Suite is hard but its not harder that developing a display server where a ton of sybsystems need to co-operate in perfect harmony
+Jack Gandy I though that Ubuntu developers are getting paid to work on what the management feels its important not what they like. Working on hobby projects is good only if its not critical for the survival of the company. And lets all understant that Ubunty and Canonical are different. Mir was not developed because it was important for the distribution(ubuntu would work great with wayland too), it was developed because the management had a different agenta or policy that did not match well with the rest of the Linux ecosystem. What does the fact that everyone else is using systemd instead of upstart tells you ?
+Chriss Kalogeropoulos But by your own admission, you're a developer.  So why not stop what you do in your day job and work on that office suite or on Wayland?  Well, it's a silly thing for me to ask of you.  But I say it's just as silly for you to ask that of +Mark Shuttleworth or Canonical.  If you're not trying to solve the problem, how can you tell others how to spend their time and their money?

You're probably a very talented develop in your field, and you're probably a really terrible developer when it comes to graphics subsystems or office suites.  This is another reason that Mir developers couldn't (universally) work on a decent office suite.  They're graphics developers.

I use Ubuntu every day.  It's not a mess.  It's a very well-integrated collection of software with a beautiful user interface.  Ubuntu popularity isn't dropping as fast as you suggest, and it isn't a popularity contest in any case.  No one else is attempting the convergence functionality that Ubuntu is, and that should be proof enough that “different needs” might be a valid reason to use something different.
Everyone's usage of a thing is a perfect condition for one to make an alternative. Every distribution has its own agenda. Linux and open source give diversity. Let's just use it!
Mark (and everyone) has every right to spend his money as he pleases. It is also my right to comment on a public post and express my opinion, that's it. I fail to see what good did Ubuntu the use of upstart. I also fail to see what good it will do the use of Mir. Probably i dont have the info or the skill to see it, but as always time will tell. I have high hopes for Ubuntu  in general. But as you very well know, its a shame to spend resources on projects that are almost finished (like Wayland) and do nothing for projects that need assistance (for example a nice enterprise level mail client). Examples of the past (compiz etc) have shown that nothing good have come out of this.

But as i said, time will tell
+Harald Glatt That does not make sense. Mark is referring to Lennart, even in post, but Lennar never called Ubuntu to replace upstart.
Mark don't let the naysayers stop you innovating. If it wasn't for Ubuntu I would never have used Linux and now Ubuntu is a permanent part of my workflow. 
Ubuntu is awesome, I just wish I could customize the dash/launcher more.  I know you gotta fight bugs so making things customizable might decrease stability.  I have tried other DEs so far and Unity is closes to what I prefer.  The top menu bar saves screen space, a launcher is the best to display favorite programs/switch between them, and the dash makes sense now that computers are faster; a menu is much slower.

If you would be so kind as to read my list of ideas though.  Optionally when clicking on a program icon in the launcher it should bring the window forward and also minimize it if clicked again.  Not a huge deal, but I don't see why it just does nothing if the window is already in focus.  The Dash is awesome, but it would be cool to be able to set it up so which ever function you want as the default is the default.  For example I pretty much only use the dash to search applications I already have installed so it would be cool if I could set my applications "scope" as the "home".

Any way I love Ubuntu thanks a lot Mark & the Ubuntu Community.
Add a comment...