Canonical Posts 15 Mesa Patches To Support Mir

[mrugiero]

How about to stop incompatibilities between upstream MESA and Canonical MESA?
It would be exceptionally irritating for developers if their code worked with one but not the other.

As long there are no API breaks, there shouldn't be any incompatibilities.
Also, remember mesa only exposes OpenGL (and may expose eventually D3D9), and it should remain compliant independently of being Canonical's or upstream, so with the exception of actual mesa developers (upstream or downstream), nobody actually *targets* mesa, every mesa user targets OpenGL, and any incompatibility is a bug or a missing feature, by definition of both concepts.

[bwat47]

I hope the Mesa team rejects the patches.

Canonical gave a big fuck you to the Linux community for going with their own inhouse display server which causes fragmentation.

Mesa should give a big fuck you to Canonical by rejecting their Mir patches.


Sorry Canonical, but Wayland is the future.

Rejecting the patches would just make things even worse.

[LinuxGamer]

That was Mir 0.0.3, they've updated to 0.0.4 now so the picture isn't relevant in the slightest, so it comes across that you're going to post it around like that BSDguy with the tasteful Tux and BSD Daemon smut.

that was all the way up to Mir 0.6 and they're on Mir 0.7 now

[alexThunder]

This.... That's the problem. They are influential, i don't think anyone would deny that - but even with all of their hype / buzzwords / clever marketing / influence / etc ~ they haven't managed to work well with all of these projects, aren't good at PR and are losing/lost respect from these projects/developers that Ubuntu is built on and *needs*... ~ it's a bit of a problematic situation.

Yes exactly. That's why I'd like to see the community taming Canonical instead of fighting or ignoring them. The latter won't produce any benefit to anyone. Again, the fault is on Canonicals side (otherwise, they wouldn't need to be tamed, would they?).

In the situation of CLA - they (Canonical) are the sole copyright holder <because part of being a contributor in CLAland is giving them assignment>... there isn't an "intermediary" like with KDE / Qt clause/license. Under Canonical's conditions, i can't imagine wanting to contribute (unless, i was an employee).

Yeah, I see that. So, some devs won't contribute to it. What's the big deal?

(Btw. Isn't QT's/Digia's CLA similar to Canonical's?)

Worrying is a waste of time. It's already happened, no one seems to be interested in Mir outside of Ubuntu (nor have i seen anyone writing Mir backends for XYZ app). That's why i say they should just maintain the Mir support themselves, rather than adding the burden to developers who are working towards something else.

Yes, they should, but - well you know how things look right now.

Sure, but this isn't anything new and it isn't always convenient / widely available - from what i understand, they (dell) also no longer offer Desktop PC models with Ubuntu, either < and as i said before, i can call dell and have them ship a PC (that is linux certified), pre-installed with any of their supported Distros... >

Still, having these shops around is a pretty obvious effort.

Maybe it was just language / miscommunication, but as i said, not everything that happens in the boardroom is marketing. likewise, marketing isn't the only avenue that led to microsoft dominating the market.

They had an comparably easy start, but they did maintain that via ~market control~ ( ) (I include dirty technical tricks to some degree here - inb4 further misunderstandings).

that's so vague, it's meaningless....

Indeed.

MS is everywhere - no problem getting help. Apple, you take you computer to them (or specialist to you, if needs be. yes, they do that!). Linux you go to a forum, or find someone who can fix the problem (less common than Mac or Windows users, no support from PC shop). and although each platform (of course) comes with their own issues, generally, the problems new-users hit using linux, can be much harder to address...

Yes, getting help for Windows is easier, but that was not the point I was trying to make. If they hit an issue, they'll need external help either way, no matter which OS they're using. If they got someone, who can help them with Linux, that's just fine.

it sure sounded like you were talking about the majority before, but regardless, you haven't really swayed me here / you've said nothing compelling, in the slightest.

Compelling for what?

They are concerned about *losing their userbase*. that is why they would do it - they thought KDE would do it <for them, they were wrong...gee, i wonder why..? lol>...

1. Getting no responses on mailing lists, a lot of negative attention from developers, etc shows that they are already isolating themselves in one sense (whether intentional or not). 2. Focusing on their own solutions rather than working on common infrastructure (Intentionally, isolating themselves) and even worse 3. crapping on other people's work (like Wayland), as they try to pull the rug from under the rest of the community...definitely a good way to piss of developers and alienate yourself.

I really don't think, they consider their KDE userbase as that important. However, " (whether intentional or not)" that was actually the point - whether they're doing this intentionally. I highly doubt this.

They made *Debian Linux* easily accessible to users. You seem to over estimate the amount of work NOT done by Canonical/Ubuntu to make this happen... They aren't doing any of the critical work involved. - furthermore, in the future, they will have very little to do with this kind of development.

I overestimate the work not done by Canonical? Did they do so much? ... Maybe I don't get you (again), but I have a feeling that you wanted to say something different.

Again, to me Ubuntu is just a (convenient) distro, packaged and set up in a worry-free manner. Still, for adoption this is important (in fact, Ubuntu is somewhat popular among Linux distros).

(some) Ubuntu users think that Ubuntu *equals* linux, which it does not, it's just one flavor... and what i meant by that is that Gnu/Linux isn't dependent on Ubuntu's success, contrary to many ubuntu kids claims. Ubuntu can't succeed (and has failed in the past) because Gnu/Linux hasn't been and (to some degree) isn't ready for the masses, even now. Maybe in a couple of years, once X is replaced and some other bits are smoothed out in the stack... either way, it won't have much to do with Canonical.

Was I saying the opposite? If so, I didn't intend to.

Sure, it's nice to see more developers / titles come to the platform. but as i said, there are other industries, that are very important in order for Desktop Linux to gain any widespread use.

Yes.

ninez, I think our opinions don't differ essentially (at least I hope), so I'd like to sum up my point of view (quoting all these less important passages gets annoying, don't you think?):

Canonical is behaving terribly towards the Linux Community, who actually made them strong. They get a lots of flames, but for good reasons. However, they have a strong influence on end-user/desktop Linux and they're growing - not too fast, but they still do. What they're doing right now may lead to fragmentation, which could (or already did) damage the Linux ecosystem and I don't see them stopping this so soon. To me it seems, the only way out is, when the Linux Community swallows the bitter pill and stops fighting them so much (although they have every right and reason to) and instead, starts cooperating - just to tame Canonical or integrating them back. Otherweise, since Canonical still is the only company seriously investing in the end-user/desktop Linux, we'll either have a MS/Apple/Canonical situation in the future (instead of MS/Apple/Linux) or we'll remain with MS/Apple and that's it. I don't consider any of these outcomes as desirable.

I may be totally wrong (I somewhat whish I am) and things may go completely different - time will tell - but right now I'm concerned.

---

Can I be BO$$, too? Can I? I promise I'll be annoying as hell! Pretty please?

You're welcome Heh, we're slowly building an army of BO$$es :> Beware of us, LinuxGamer!

[intellivision]

that was all the way up to Mir 0.6 and they're on Mir 0.7 now

That's incorrect, trunk is only at 0.0.5. No code has been contributed to the 0.06/7 placeholders as they're just that, placeholders for bug tickets.
So now we know you're lying or not well researched, how can we trust what you say?

You also didn't put up any evidence that proves mrugiero and BO$$ are the same person, so I assume you're going to shut up about that instead?

[mrugiero]

Contributor agreements are contributor agreements. Wrong is wrong. Having an intermediary like the one between KDE and Qt is just layered problems not a solution. First of all it is an agreement between Qt an KDE. Not Qt and Qt Free Users. Having discrimination like this is worse than having nothing. Second the intermediary is close to powerless. Qt can do whatever they want and evolve to evilness, all they need to do is to publish a new point release every year for the already published Qt Free code base. The agreement is so full of Qt cleverness that an evovled Qt Free can be WORSE for KDE. The Ubuntu users were the last to know about Mir, Qt Free users would get the same treatment. Qt have every right to differ more between Qt and Qt Free, and there is nothing you can do about it.

On one hand, the post you quoted contains several wrong bits, for a start, is not a copyright assignment, is a right assignment where you give Canonical the right to sublicense your work. You still have your copyright, and you can still relicense *your* bits of the code (but they would be useless by themselves, as a patch is only useful when there is a code base to apply it to) or do whatever you want with it, with the possible exception (there are no known succesful cases) of reverting the GPL license on them (i.e., you can not retroactively make it closed source, but since you are the copyright holder for that, you can make a closed sourced copy or derivative). As Canonical retains this right assignment from any and every contributor, they have this right over all the code, so they can effectively make closed source derivatives of the code. They can not (and not even hypothetically) retroactively erase the GPL, not only because the CLA doesn't give them the right, but because they aren't, and can't be, the legal copyright holders of any licensed work (i.e., contributions from external devs).
The same is valid for Digia.

So, in the end, what happens? Both are guaranteed to keep open source, at least every contributed bit (they can rebase their work as closed source, in the case of Canonical, while there is another agreement between KDE and Digia and whatever company that buys Qt in the history that prevents this in their case), and both assign a different set of rights to Canonical and to the developers, so even when there is a pragmatic advantage in the Qt case, both are good or wrong on the same degree, ethically: the company has got a right over the whole code base that nobody else has got.

[mrugiero]

That's incorrect, trunk is only at 0.0.5. No code has been contributed to the 0.06/7 placeholders as they're just that, placeholders for bug tickets.
So now we know you're lying or not well researched, how can we trust what you say?

You also didn't put up any evidence that proves mrugiero and BO$$ are the same person, so I assume you're going to shut up about that instead?

Didn't they release 0.0.7 a few days ago? Also, in my test box I have 0.0.6 installed.
Anyway, it's not that corrupted in my box, so I assume he's either lying or it's hardware dependent.

[intellivision]

Didn't they release 0.0.7 a few days ago? Also, in my test box I have 0.0.6 installed.
Anyway, it's not that corrupted in my box, so I assume he's either lying or it's hardware dependent.

I was going off what was released in trunk, here: https://code.launchpad.net/~mir-team/mir/trunk

If what you're saying is right, I'm guessing that they haven't released the code for 0.0.6+ yet.

[mrugiero]

I do not claim copyright are to be transfered. Licenses are.
I do not claim published GPL code can be revoked.

I didn't mean you claimed that. I meant the post you were quoting contained the first statement, and since you didn't refute this I thought you might ignore it. The second was just to clarify, just in case.

The agreement between KDE and Qt solves nothing. It has no guarantee or promise for Qt to be free. It simply states that KDE can make a relicense of "Qt Free"(not Qt) IF Digia fails to do a yearly point release of Qt or "Qt Free". That will not happen unless Digia abandons Qt. What KDE and Qt is doing is discrimination. No other stakeholder have these rights. There is no such thing as asymmetric freedom.

Yes, that was the point. I misinterpreted the agreement, though, but the point was that you are right in being just the same. If it's wrong for Canonical, is wrong for Qt. If it's right for Qt, it's right for Canonical. I'd personally avoid committing any code to them if I were that skilled (which I'm not), but wouldn't avoid the use, since this is mostly unfair to devs and doesn't change things much for end users (I mean, we, as users, have in both cases the same rights we'd have with plain GPL; the ones having in some way their rights stolen are the developers).

[ninez]

Yes exactly. That's why I'd like to see the community taming Canonical instead of fighting or ignoring them. The latter won't produce any benefit to anyone. Again, the fault is on Canonicals side (otherwise, they wouldn't need to be tamed, would they?).

no, i disagree with the taming part. Letting them manage their in-house solutions seems like a better plan.

Yeah, I see that. So, some devs won't contribute to it. What's the big deal?

(Btw. Isn't QT's/Digia's CLA similar to Canonical's?)

A few developers not contributing isn't a big deal. Canonical controlling our replacement for Xorg, doesn't seem like a good plan... I think Mathew Garrett's article got into some of this stuff nicely; http://mjg59.dreamwidth.org/25376.html?thread=959520 <you should even note the 1st comment too, on Qt's clause... and although the other commenter here says Qt's is worse, i've heard otherwise from many sources over the years. it sounds to me like Canonical's CLA is worse, in terms of free software>

Yes, they should, but - well you know how things look right now.

?

Still, having these shops around is a pretty obvious effort.

It's not the first linux effort and won't be the last. Things are going to go in that direction when the stack catches up.

Yes, getting help for Windows is easier, but that was not the point I was trying to make. If they hit an issue, they'll need external help either way, no matter which OS they're using. If they got someone, who can help them with Linux, that's just fine.

sure that's fine, except getting (external) help is less likely and not everyone can afford "down time". So that kind of makes what your saying moot.

Compelling for what?

compelling; definition 2. Not able to be refuted; inspiring conviction

your comment i was responding to, wasn't compelling...

I really don't think, they consider their KDE userbase as that important. However, " (whether intentional or not)" that was actually the point - whether they're doing this intentionally. I highly doubt this.

your naive. They wouldn't be paying employees to support it (since KDE won't) if that wasn't the case (that kubuntu user's weren't important) - please, use some "common sense" here. :\

ninez, I think our opinions don't differ essentially (at least I hope), so I'd like to sum up my point of view (quoting all these less important passages gets annoying, don't you think?):

Canonical is behaving terribly towards the Linux Community, who actually made them strong. They get a lots of flames, but for good reasons. However, they have a strong influence on end-user/desktop Linux and they're growing - not too fast, but they still do. What they're doing right now may lead to fragmentation, which could (or already did) damage the Linux ecosystem and I don't see them stopping this so soon. To me it seems, the only way out is, when the Linux Community swallows the bitter pill and stops fighting them so much (although they have every right and reason to) and instead, starts cooperating - just to tame Canonical or integrating them back. Otherweise, since Canonical still is the only company seriously investing in the end-user/desktop Linux, we'll either have a MS/Apple/Canonical situation in the future (instead of MS/Apple/Linux) or we'll remain with MS/Apple and that's it. I don't consider any of these outcomes as desirable.

I may be totally wrong (I somewhat whish I am) and things may go completely different - time will tell - but right now I'm concerned.

Where we fundamentally disagree: You believe that "desktop Linux" can only be successful if Canonical succeeds, thus you expect others (ie: the entire community / distros) to be a push overs for Canonical, because if we don't -> then We are stuck with Apple or MS.

Ubuntu is investing in Ubuntu, not "Desktop" Linux.