Intel Reverts Plans, Will Not Support Ubuntu's XMir

[johnc]

Yeah, I agree. There's no way Intel is going to maintain patches for a one-distro display server like, say... SurfaceFlinger.

...ohh, wait...


I took a look at the code that was backed out. Yeah we're talking about stuff that would cost approximately $25 to maintain over the course of the code's life. A real burden for Intel here.

[dee.]

Look at yourself... rationalizing Intel's actions... Are you proud?

Sometimes, things are just as they seem you know. Your rationalization is not the reason Intel gave for reverting the patch. Why are you trying to make excuses for them?

Intel has a vested interest in Wayland (Tizen). They do not support Canonical's decision [to crate Mir]. That is why the patch was reverted. That is what they said. They did not give any other reason that people have tried to present in this thread. If those reasons were a part of their position then they would have presented them.

It's exactly the right choice. Intel shouldn't have to maintain patches for a one-distro solution. If everyone and their grandmothers start churning out display servers, will Intel have to support and maintain patches for those as well? No. If I go and fork XMir, change one thing so it requires different driver patches, and publish it on my one-man distro, will Intel need to maintain patches for that too? No. Why would Canonical be an exception?

If Canonical wants to do such a horribly stupid thing like Mir, they can do the work for it themselves, not count on others to do their job for them. It's as simple as that.

Do you think Intel's decision is for the good of the "community" as some people call it?

You are just choosing your own dictator... But alas, everyone does reserve the right to choose their own dictator...

Intel's actions set precedence for what could be some quite nasty actions in the future.

Slippery slope fallacy. Intel has shown that they have great open source driver support. They have the best open source drivers of any GPU manufacturer. Yet the moment they decide they don't want to maintain a patch for a one-distro solution, they're the evil empire. Durr....

Intel does not care about the "Linux community". Intel cares about Intel and their decisions are made solely to benefit Intel and not competitors.

Remember the Intel + Microsoft monopoly? Intel wants to recreate that with Linux except this time it will just be Intel.

What happens when Intel management decides that they do not condone the actions of RedHat, SUSE , [Insert other distro or company here that may be a competitor]?

More slippery slope.

People have their problems with Canonical or Mir or whatever else, but they shouldn't let this blind them to the fact that Intel's action was bad and sets precedence for similar actions but which negatively affect more than just Canonical or Ubuntu in the future.

Just remember when it happens that you supported it.

Dude, it was one rejected patch that shouldn't have been submitted in the first place. Canonical should have taken up maintainership of that patch in the first place, because that patch doesn't benefit anyone other than Canonical. It's not like Intel is actively hindering Canonical's plans on creating & using Mir, they're not actively making changes that would harm Mir's performance in any way. They're simply saying that if Canonical takes this road, they're free to do that, but they're not helping. Which is good, misbehaving children need to be set boundaries, otherwise they'll never learn.

And don't kid yourself, this decision doesn't affect anyone other than Canonical (not even them, really, as they can just maintain their own patch). No one else is ever going to use Mir. It's Unity-only, and everyone else will use X or Wayland. That's just the way it is, deal with it.

[jayrulez]

Intel did write the patches. Why don't you pick it up and maintain it? It must be a zero cost since you dare blame Intel for not doing it. Im looking forward to send you standard issue bug reports

Are you plain stupid or are you just playing stupid?

[mrugiero]

You mean writing AND maintaining the driver for its entire life time. Have fun at the bug tracker when it is flooded with angry users pasting shit into reports, and every bug sees a ton of "me too", "FIX it", "I see the same problem on my overclocked PC, but I use a different closed source driver and recompiled with CanonicalCLACompiler0.4.1.pre3.rc2, O=5, arc=ARM65bit. Fix it NOW or I go NetBSD"

Well, I'm no driver developer, yet. However, Intel's maintainer is called a maintainer because that's what he does. I bet he knows if he can maintain something.
Also, what you seem to ignore is that I said several times, within this thread, that I believe such a support actually belongs out of tree, because of obvious reasons. This doesn't change the fact management reverted a maintainer's patch. To my criteria, the maintainer of a driver is the most capable individual of judging if the patch is in conditions to be merged, and if it's trivial or hard to maintain, and act in consequence.

When submitting code you have a duty to maintain that code...

Intel probably understood this. They could have let the code in, but in the end it could have made a bad name for Intel after a few months with regressions and stuff, since they don't seem to be wanting to spend any more time on it.
It's like with Google's Android engineers who wanted to contribute code to linux. The Linux Group didn't believe that Google would maintain their code so they rejected the code. This is a reverse of that situation, with Intel doing about the opposite.

We all know about that, thank you. It was said several times at the beginning of the thread. The point is, the maintainer usually knows how to judge this situations, and knows under which conditions a patch should be outright rejected.

[jayrulez]

When submitting code you have a duty to maintain that code...

Intel probably understood this. They could have let the code in, but in the end it could have made a bad name for Intel after a few months with regressions and stuff, since they don't seem to be wanting to spend any more time on it.
It's like with Google's Android engineers who wanted to contribute code to linux. The Linux Group didn't believe that Google would maintain their code so they rejected the code. This is a reverse of that situation, with Intel doing about the opposite.

Look at yourself... rationalizing Intel's actions... Are you proud?

Sometimes, things are just as they seem you know. Your rationalization is not the reason Intel gave for reverting the patch. Why are you trying to make excuses for them?

Intel has a vested interest in Wayland (Tizen). They do not support Canonical's decision [to crate Mir]. That is why the patch was reverted. That is what they said. They did not give any other reason that people have tried to present in this thread. If those reasons were a part of their position then they would have presented them.

Do you think Intel's decision is for the good of the "community" as some people call it?

You are just choosing your own dictator... But alas, everyone does reserve the right to choose their own dictator...

Intel's actions set precedence for what could be some quite nasty actions in the future.

Intel does not care about the "Linux community". Intel cares about Intel and their decisions are made solely to benefit Intel and not competitors.

Remember the Intel + Microsoft monopoly? Intel wants to recreate that with Linux except this time it will just be Intel.

What happens when Intel management decides that they do not condone the actions of RedHat, SUSE , [Insert other distro or company here that may be a competitor]?

People have their problems with Canonical or Mir or whatever else, but they shouldn't let this blind them to the fact that Intel's action was bad and sets precedence for similar actions but which negatively affect more than just Canonical or Ubuntu in the future.

Just remember when it happens that you supported it.

[powdigsig]

When submitting code you have a duty to maintain that code...

When submitting code you have a duty to maintain that code...

Intel probably understood this. They could have let the code in, but in the end it could have made a bad name for Intel after a few months with regressions and stuff, since they don't seem to be wanting to spend any more time on it.
It's like with Google's Android engineers who wanted to contribute code to linux. The Linux Group didn't believe that Google would maintain their code so they rejected the code. This is a reverse of that situation, with Intel doing about the opposite.

[dee.]

No one 'needs' Tizen to take off, there are plenty of other Linux based smartphone platforms on the market already or have similarly been announced such as Jolla, Firefox OS, Android, OpenWebOS, Ubuntu Mobile and SHR.

Any one of these platforms has the capacity to meet the same needs as Tizen.

Right, but if we really want to see some mainstream success on the laptop/desktop side, with an actual Linux OS, not some silly cloud toy, then Tizen on ultrabooks is probably the best shot there is of that happening.

[dee.]

It's a pity that most of the Tizen project is hinging on the success of the smartphone platform, which Samsung is looking uneasy about supporting in the future.
The lack of a FOSS SDK and the fact that several applications shipped with the OS fall under the non-FOSS Flora license is also worrying, compared to other open source projects such as FirefoxOS, not to mention the closed source Bada libraries that will be included with Samsung's handsets.

There also seems to be a lack of planning especially in regards to the official development frameworks available.
Firstly it was going to use EFL and C++ for native applications and HTML5 for the rest, with the EFL frameworks being depreciated temporarily, then were brought back.
There also hasn't been any word as to whether Qt applications will be officially supported by Tizen and therefore allowed into their walled garden market.

Their use of Wayland is impressive, but their execution of this project may have killed it before it even took off, the same way LiMo, Maemo and Meego went.

About the SDK, it APPEARS they may have done away with the onerous license it used to have, the page on which the SDK license used to be does not exist anymore, and there is no such license included within the SDK anymore, only file on "license information" in it lists the licenses of the included packages. So if it's true and they decided to can the crappy license (whose legality was also questionable, considering the SDK contained GPL code in it) then that's a good thing.

The Flora license doesn't worry me too much either. Arguably, Samsung had to use such a license, because giving a 100% patent grant to everyone would have hindered their legal battle against Apple. Apart from the patent grant part, the license is identical to the Apache license. Also the Flora license only applies to some user-facing apps which can easily be replaced with free alternatives if it comes to that, nothing crucial to the functioning of the system.

[johnc]

Intel is already supporting X11 and Wayland. Having developers on staff cost money so one has to focus on a few important key projects. You simply can not support everyone's pet project out there.

Ubuntu abandoned community developed Wayland of political reasons because they wanted full control to drive the project in THEIR desired direction. So I is just fair that THEY develop and support their own creation.

Intel makes more money in an hour than Ben Bernanke can print in a whole year. I don't think they're pinchin' pennies over there. However I do agree that this would have been a better explanation. Of course it wasn't "The Management's" explanation... theirs was something about not condoning the actions of Canonical. Sufficiently vague yet gets the point across, in a passive-aggressive sort of way.

[skies]

Intel is already supporting X11 and Wayland. Having developers on staff cost money so one has to focus on a few important key projects. You simply can not support everyone's pet project out there.

Ubuntu abandoned community developed Wayland of political reasons because they wanted full control to drive the project in THEIR desired direction. So I is just fair that THEY develop and support their own creation.