More Mir Talking Points Come Out Of Canonical

[Serafean]

That sounds like a great argument... to use Windows.

Seriously i don't know how many times I need to repeat this. If Canonical can't "play nice" with others in the open community, they don't deserve the same level of respect as other community projects. I'm not saying they're all bad or anything like that, I'm saying judge each of there action for it's merits, not "well Canonical did something really dumb and potentially destructive for no reason, but it's okay cause we have to follow what they do cause they're the 'big kid on the block'". I mean, if that's your attitude, why even use open-source software in the first place?

Hear hear...

Steam works fine on my Gentoo system. For some reason it doesn't detect the S3TC extension on OSS drivers though; but hey, I'm running git versions of almost everything .

Back on topic : I read the post on G+, and as an avid follower of Wayland, I still call bullshit on their reasons. (referenced as on the G+ page) :
1) X is old and crappy to work with. Seems to be the consensus.
2) "We didn't want to use Weston." The reference compositor is reference for a reason. You do your own after that. In fact, rolling your own compositor could help highlight some deficiencies in Wayland.
3) "At the time Mir was started, Wayland's input handling was basically non-existent." So starting from scratch instead of helping out seemed like a good idea... Deep Thought Material.
4) "We need server-side buffer allocation for ARM hardware; for various reasons we want server-side buffer allocation everywhere." Can't comment, no idea there.
5) "We want the minimum possible complexity; we ideally want something tailored exactly to our requirements, with no surplus code." First, this isn't going to happen, a Wayland-like product will eventually grow out of it. Second, if it does happen, you'll be the only users of Mir...

If I'm wrong somewhere, come at me...

Oh, and I hate blueberries

[TemplarGR]

Funny thing. I just read half an hour ago on ars tehnica a couple of guys using arch and complaining about instability because of the rolling release model and how it always took more work to just get it working, especially after a few months away from it. Arch is not targeted by Valve and it probably has something to do with that. Arch is also not very popular, also probably because of that. The fact that Steam works is just luck, but who knows, maybe tomorrow it won't work. Right now Ubuntu 12.04 is safe for 5 years knowing that steam will work on it from now on. I'm no Ubuntu fanboi as it has plenty of bugs, but for now it's the most popular and it looks like it only will get more popular when vendors will add or improve support for it. That is why I was suggesting people to move to it. If Arch was the most popular I would suggest people to move on it. Get it? It's not the fact that it's Canonical, but that it's the most popular.

Well, the funny thing is that for the type of distro Arch is, it is fairly popular. It's target group is power users, not casual users. But, let me say a couple of things:

1)Arstechnica has gone downhill. Most of the people writting there don't know shit. That is the reason i never visit it anymore...
2) There is no instability because of the rolling release model. None. Zero. The guys saying this crap are useless n00bs who shouldn't use a distro for power users anyway...
3) No sane desktop user/gamer will stay on 12.04 for 5 years. It is too much, because there will be no real updates, just security fixes. You wouldn't like to use a 5 year old web browser, would you?

[russofris]

Protecting your property is justified.

Except that it is not property.... You get that, right?

F

[AdamW]

I find it interesting that Canonical are all but openly saying that NVIDIA and AMD are on board - that's what all the insinuations about "Graphics vendors have been happy to engage and ensure it works well on all architectures" and "our engagements with NVIDIA and AMD" are about - but I haven't seen a single public statement from NVIDIA or AMD yet. If I were Mark I'd be a lot more careful about that. People are always happy to sound as positive as possible in negotiations, right up until the point they have to commit to doing stuff that will cost them money...

[AdamW]

Being relatively new to linux (less than a year) I think it's funny that it seems any opinion someone puts on here is strongly countered by someone else saying how they are totally wrong.
Let me check something.

Blueberries are awesome.

Blueberries are disgusting and actively harmful to human health. Go join your friends in AL QAIDA.

[deanjo]

2) There is no instability because of the rolling release model. None. Zero. The guys saying this crap are useless n00bs who shouldn't use a distro for power users anyway...

Now that isn't true. A rolling release is more likely to find regressions quicker then a periodical release. It's just the nature of the beast. I am not however saying that regressions don't happen in periodical releases either because they do as well it is just that you don't see them as often as on a system with upgraditis.

[curaga]

If they keep this up, I will run out of both popcorn and peanuts.

[AdamW]

What's the specific problem with Canonical's CLA? The GNU copyright assignment is similar and nobody seems to give a shit. IMHO, the GNU is no more trustworthy than Canonical. A permissive license would have been a better choice for Mir, though, especially as they want to support GPU vendors which usually don't like GPL family licenses.

That said, explicit effort to support common graphics stacks, open source and closed source, seems to be the most important difference between Mir and Wayland. As far as I know, Wayland/Weston developers have said (or at least implied) they don't particularly care about the blobs and won't collaborate with GPU vendors or go out of their way to make blob support happen. But no matter how much some people wish for an open source graphics world, the blobs are here to stay for some time.

Mainly, FSF is a non-profit which specifically exists to further the cause of free software. Canonical is a private company: the only thing you can absolutely trust it to do is act in its own interests. Which you don't really know what they are. So assigning copyright to Canonical is a very different act from assigning it to FSF.

Two, people don't really care about the FSF copyright assignment because there's no particular reason ever to use it. You don't even have to if you want to be a GNU project, which mostly people don't care about any more anyway. I don't even remember the last prominent new GNU project. It used to be useful to be a GNU project to get hosting and promotion and stuff, but that was at least a decade ago...

(edit: now I read up on it, you don't even have to assign copyright to the FSF to be a GNU project. You can choose to do so or not.)

(edit2: in the case of existing GNU projects which are still important, also see What Rahul Said.)

[TemplarGR]

Now that isn't true. A rolling release is more likely to find regressions quicker then a periodical release. It's just the nature of the beast. I am not however saying that regressions don't happen in periodical releases either because they do as well it is just that you don't see them as often as on a system with upgraditis.

Nope. You are just confusing bleeding edge release with rolling release... A distro can be rolling release and not bleeding edge.

I agree that a bleeding edge release is more likely to find regressions, but most of the time they are fixable. And those that need such a distro have the will and the knowledge to do so.

On the other hand, a rolling release isn't obliged to throw untested code to stable branches... Ubuntu could be rolling release, and have all updates tested properly before deployed... Nothing would change for the end user, except that his system would be more up-to-date especially for apps, and that he wouldn't have the need to dist-upgrade every 6 months...

[cardboard]

Nope. You are just confusing bleeding edge release with rolling release... A distro can be rolling release and not bleeding edge.

I agree that a bleeding edge release is more likely to find regressions, but most of the time they are fixable. And those that need such a distro have the will and the knowledge to do so.

On the other hand, a rolling release isn't obliged to throw untested code to stable branches... Ubuntu could be rolling release, and have all updates tested properly before deployed... Nothing would change for the end user, except that his system would be more up-to-date especially for apps, and that he wouldn't have the need to dist-upgrade every 6 months...

Rolling release doesn't mean "up to date" either.