Intel Reverts Plans, Will Not Support Ubuntu's XMir

[TheBlackCat]

A brand new day, a brand new FOSS drama

This isn't exactly new. Where have you been for the past six months?

[seba]

This isn't exactly new. Where have you been for the past six months?

My comments were regarding the original post title: "Intel Reverts Plans, Will Not Support Ubuntu's XMir" which I am just seeing today...

[mrugiero]

I don't actually think so. Ubuntu seems to have a pretty good following building up in China, they have their own localized variant (Kylin, or somesuch) and there are some Ubuntu-preloaded computers quite prominently displayed on some Chinese stores.

Also if that's what they wanted to do, they could still use FreeDOS. Certainly, some people who already own a copy of windows might buy the Ubuntu version just to be spared from paying for a redundant copy of windows, but if that was their only aim, they could do just as well with FreeDOS or something similar, no need for a fully functional OS. Also, if that was the case, they most likely wouldn't offer support or guarantees for the OS (not sure if they do that now, but I suppose that should be relatively easy to check if someone is really that interested).

It's a valid point. This doesn't change the fact in China there's a lot of pirates, who might just buy it to install a pirated Windows. There is probably a big lot of legit Ubuntu users, too.

[Bucic]

First Amnesia A Machine for Pigs gets released (including Linux) http://store.steampowered.com/app/23...snr=1_4_4__118 and now the news. What a great day!

[Pajn]

Or have "no distro-specific patch" policies like most sane upstreams do.

That would be a valid policy yes, however that isn't the case here.
The patch were approved by the maintainer of the repo.
That means the guy who is usually fully able to say that
I want this or I want that. It isn't like it were approved
by some guy who didn't know what he were doing.

If it were a "no distro-specific patch" policy he wouldn't
have approved it in the first place.

[Vim_User]

So, no one should push anything upstream ever? Because that's exactly what you just said. Why is it only shifting burden of maintenance upstream if Ubuntu submits a patch, but its sharing and caring if anyone does? The double standards are astounding.

Don't put words in my mouth. Shifting things upstream is OK if those things are not single-distro solutions and/or if the submitters of those patches are willing to maintain them. Canonical has not a good track record with maintaining such things and Mir/XMir is single distribution, so any sane upstream will deny the merge of those patches.

[dh04000]

Don't put words in my mouth. Shifting things upstream is OK if those things are not single-distro solutions and/or if the submitters of those patches are willing to maintain them. Canonical has not a good track record with maintaining such things and Mir/XMir is single distribution, so any sane upstream will deny the merge of those patches.

Vim, look at the comment directly above you. Its an appropriate response to your comment.

[TheBlackCat]

That would be a valid policy yes, however that isn't the case here.
The patch were approved by the maintainer of the repo.
That means the guy who is usually fully able to say that
I want this or I want that. It isn't like it were approved
by some guy who didn't know what he were doing.

If it were a "no distro-specific patch" policy he wouldn't
have approved it in the first place.

This is not necessarily true. I could easily see a situation where, for example the maintainer thought that the patch was small enough that it didn't matter, but others thought that it would set a precedent for accepting other, larger, more intrusive changes later, so they decided just not to go down that route altogether. Or the policy is not strict, but there was a disagreement regarding which side of the line this falls on.

We don't really know what the exact reason was for accepting the patch or removing it. It may turn out that it was an entirely political decision. But we don't know enough to say that yet.

[mrugiero]

Don't put words in my mouth. Shifting things upstream is OK if those things are not single-distro solutions and/or if the submitters of those patches are willing to maintain them. Canonical has not a good track record with maintaining such things and Mir/XMir is single distribution, so any sane upstream will deny the merge of those patches.

Worth mentioning is that it's also a matter of scope (of course, the particular case of Intel's management reverting the patch is kind of politics and nothing else, assume I talk in the general case and not this); if your patch is only useful in a specific context (for example, a single distro, but could be a specific CPU, too), it might make sense to keep the patch floating around and not in the main repo, since every change could at some point have side effects, which won't bring any positive effect to people outside that context. In the case of distributions, it makes more sense, IMO, to have them in an out of tree repo from the distro, so it's easier for them to update this changes periodically, instead of having to always send it upstream, go through extra revisions (IMO, they should be already tested to work correctly on their context, so for context specific patches this extra, generic revisions may be unnecessary) and troubles just to get it updated for them.

[mrugiero]

This is not necessarily true. I could easily see a situation where, for example the maintainer thought that the patch was small enough that it didn't matter, but others thought that it would set a precedent for accepting other, larger, more intrusive changes later, so they decided just not to go down that route altogether. Or the policy is not strict, but there was a disagreement regarding which side of the line this falls on.

We don't really know what the exact reason was for accepting the patch or removing it. It may turn out that it was an entirely political decision. But we don't know enough to say that yet.

Are you kidding? Haven't you seen the notice Intel gave for it? It wasn't about being distro specific, it was clearly stated the 'course of action' of Canonical wasn't approved, and that it was the reason. Also, if there is disagreement it usually turns that way in the discussion in the mailing list, when the patch is proposed.
What you suggest is not better than BO$$ claims that there is an ultra secret OMG IT'S GREAT technical feature that makes Mir better than Wayland, just Canonical can't reveal it yet.