NVIDIA Presents Its Driver Plans To Support Mir/Wayland & KMS On Linux

[Temar]

you are living in some fantasy world. linux is selling in ~billion(and growing) smartphones per year, which is far better than all other oses combined.

Oh you are such a smart alec. Unfortunately this article is about Nvidia in the context of Wayland and Mir.

[Temar]

Steam survey has a bias that Ubuntu is the officially supported platform. Even then, we can observe that Mint amount to approx 20% of "Ubuntu + Mint" volume, probably slightly higher outside steam.
This is important because Mint user agent returns Ubuntu as well.

Steam's data is not based on user agents but on exact measurement by its client software which runs directly on the PC itself.

Here is another statistics with exact measurement:

http://www.gamingonlinux.com/articles/unity3d-games-phone-home-with-details-of-your-hardware-software.4452

[philip550c]

maybe in your alternative reality, but here that distro is called android and it has ten digits installed base.

Android is not a distro. Android just uses the linux kernel, so is my smart fridge os considered a distro too? How about the drive thru at mcdonalds?

[philip550c]

This is important because Mint user agent returns Ubuntu as well.

Im pretty sure it doesnt, im running mint and everytime I see a useragent string on anything it never says ubuntu.

[erendorn]

Steam's data is not based on user agents but on exact measurement by its client software which runs directly on the PC itself.

Here is another statistics with exact measurement:
http://www.gamingonlinux.com/article...-software.4452

This is exactly what I was saying, so maybe you didn't get what I wrote.

Im pretty sure it doesnt, im running mint and everytime I see a useragent string on anything it never says ubuntu.

I did not boot my mint partition to check, but this thread seems to indicate it does (no idea for Mint 17):

http://forums.linuxmint.com/viewtopic.php?f=29&t=161065

[ssokolow]

This really reminds me of people who can not accept that Linux has a market share of 1-2%. They make up ridiculous arguments, while ignoring the fact that muliple, _independent_ statistics come to the same conclusion.

You might not like it but it is what it is.

I'll agree that ridiculous arguments are a problem, but some of us (usually the ones who remain fairly quiet) have a different issue with the accuracy of that figure.

If Steam reports 1-2%, what margin for error is imparted by the slice of the market who refuse to run it?
1. If any platform is likely to have Steam-shunners as a disproportionate segment of the market, it's Linux.
2. Let's assume that many users who don't like DRM are like my friends, buying DRM-free downloads that include a Steam key for convenience. The fact that 8.5% of Defender's Quest purchases were made on GOG.com (which refuses to offer Steam keys), suggests that this region of overlap should not be relied upon to consume the aforementioned margin for error.

If Linux's market share really IS at 1-2%, I'm fine with that... but I have a problem with sloppy statistical thinking.

Steam's data is not based on user agents but on exact measurement by its client software which runs directly on the PC itself.

Here is another statistics with exact measurement:
http://www.gamingonlinux.com/article...-software.4452

I don't have hard facts but, given how small the percentage of Linux users is compared to what Steam reports (0.1% vs. 1-2%), I get the impression that they're not correcting for the effects of the many Unity3D games which were never released for Linux. (...and, for all I know, that's a statistic on how many installs there are per platform, rather than users. The source URL is down so I couldn't check.)

[CoderniX]

I can't afford the time to be here 24/7, but I'll try to respond to the posts accordingly:

Why do we keep comparing everything re: Linux to Apple? Apple doesn't know a damn thing about writing good software. Forget about Quartz. Quartz is garbage. Has anybody even tried playing games on OS X? It's a complete trainwreck! Performance is awful, there's no raw mouse input. Linux is already better (technical side) than OS X, we don't need to go backwards.

For one simple reason, because Ordinary People pay $$$$ to have a crappy "Macintosh"es, because it has OSX on them and not Linux or Windows, and because most of the State of the Art Applications are written for OSX and not Linux when it comes to supporting a *Nix platform, and I mean most of the GFX and Sound/Video Compositing and VFX solutions' Software, others can be found as Cross-Platform like the Foundry's and Autodesk's.

Linux has been a second class citizen for a long time, regardless of how good its code is. So let's get out of our denial state, and admit OSX is more well received than any Linux Distro, and only Ubuntu rose with that aspect and on the marketshare level.

And if you want to see success in OS sales and see how the world works, you look at MS, not Apple. Apple is a fifth-rate PC vendor behind Acer. ACER!!! A-C-E-R!!!

You're getting it wrong again, OSX has always been offered for dirt cheap money, rather they sell a full complete bundle, a combo of Hardware & Software, and no one is doing that to be fair to Apple, although I hate the guts out of them, yet I still use OSX like right now :/

OSX now is for free, and you don't have to pay a penny for it, while MS Windows is ... is what? Is ..., is what? You know the rest, and I remember that more than %80 of the Windows Desktop found on PCs around the world are pirated, although the article was quite old, I bet it's worse now. Microsoft can't make money out of their comfort zone, i.e. North America and some EU countries, as I've visited other countries and Windows was sold for less than a $1 as a pirated burned DVD, and I'm not speaking about 3rd World Countries, where even Government Departments use Pirated Copies, oh, and I've seen that with my own eyes, so let's get down that analogy.

Windows has never been marketed and sold because it has superior code either, rather people have used to its familiarity, and because it has ranges of Applications for whatever you want to do on the OS.

You can get $100 Acers but that won't happen for any Mac though, so again your analogy isn't suitable either, because if Mac goes down to those prices, I bet no one will beat them to it, and if you sell zillions of $100 Acers, then count the Support, Warranties, paying the employees and the research guys, while Apples would have to provide less support, pay less people, fix less Warranties and may be make more net income comparably, let alone that Apple is always after Fame & Glory, and would careless about the other stuffs.

The only thing Apple is good at is selling low-quality hardware at premium prices, and finding people to fall for it. I read a report that Apple takes 50% of PC profits (despite practically no selling marketshare). That says it all, folks. We know how to calculate "profit": revenue minus expenditures. You're not going to find that kind of irrationality in the Linux-buying market.

Just read this, and found out that you agree to what I've just written above, thus you should get an idea why we can't deny that we have to compare Linux to Apple, since it's Crappier yet it does a million times better than Linux, and it's not about people falling for their prices. I need to use Photoshop, AfterEffects, and unfortunately I need to use Xamarin when the client wants a cross-platform mobile App, which makes me opt-in for OSX, since I don't want anything MS OS on my computers.

As for Canonical. To me they started off as the "Hey, we're the opposite of the corporate Red Hat crap" company. And everyone was like, "Yay! Linux for human beings!" IOW, going against the grain was their attitude. That turned into the, "Hey, we're going to be like Apple. Who told you you could put that icon there in the system tray? No! Remove it!" company. And everyone was like, "WTF just happened?" Suddenly it was less about the Ubuntu community and more about some design fairy-dusters in Canonical using the distro as their pet project. We get these UI tweaks and new icons and meanwhile look at all the bugs marked "Critical: Triaged" in LaunchPad.

Nothing changed for me, Mark stated from the very beginning:

"No. This is not a democracy. Good feedback, good data, are welcome. But we are not voting on design decisions."

The above included everything else for me, and I knew where I'm going to base my whole workflow on, and I remember getting extremely upset when Unity was introduced, and I switched to all sorts of Distros, only to find myself running back to Ubuntu, and if I have to jump to something else, that would be Elementary then Mint comes third, and all of them are Ubuntu based.

So I'm kind of on the opposite side where I think Canonical needed to buck the "Linux community" (read: RedHat) and do their own thing, just like Google did with ChromeOS and Android, but at the same time needed to do it in a way that actually addressed real user needs.

But they're not doing that anymore.

Now they're doing phones.

Nope, they're doing Desktops, Phones, Tablets, and TVs.

And let's be serious here... this has got to be the last straw for Shuttleworth. I can't see him going on when even this venture fails. And make no mistake, the market trends are clear: they will make no money on this venture and accumulate no significant market share.

Only time will prove how right or wrong you are. The Chinese must be paying for it, other OEMs would be too, and a good strategy would be getting the Phones to be marketed well, so the funds coming from that way can fund the Desktop part, since it's a one stand solution.

I'm the wrong guy to ask. I'm still committed to the idea that X is the wave of the future. Though admittedly all this talk from the GPU vendors about Wayland, Mir, etc. is making me a little nervous about my prediction.

If it's up to me, I'd rid X from all the rubbish that keeps Hardware older than 5 years working, and start anew on the same cleaned up code base.

?????????????

Don't bother to answer. Meantime Mir will reach the 1.0, the world will be already on wayland for quite a while.
Consider this scenario:
Middle of 2015, I will play some AAA game through my nvidia, using their blob on top of wayland. Then I will take a quick break alt+tab-ing to the browser just to see the nth article about "when Mir gets more acceptance/approval and will be more used than Wayland?".
:-)

The world has been on X for as old as some members' ages here, and sometimes way older than them, yet it's being deprecated, hence when Mir gets in the Playground we shall see what would happen, isn't it?

[erendorn]

If it's up to me, I'd rid X from all the rubbish that keeps Hardware older than 5 years working, and start anew on the same cleaned up code base.

This looks like what the (GNU/)Linux world has been doing for the past 5 years, through the graphics stack refactoring, and projects like wayland/mir.

The world has been on X for as old as some members' ages here, and sometimes way older than them, yet it's being deprecated, hence when Mir gets in the Playground we shall see what would happen, isn't it?

Wayland and Mir bring whole new paradigms to the display system, and X has many issues from being designed for hardware of the last century. That's why it's being deprecated.
But Mir doesn't bring anything compared to Wayland, so the comparison is pointless. The only thing it could bring is "swiftness" for Canonical (at the cost of less reusability of other projects), but even that it failed to do, given that Mir has not come to market any before than other Wayland compositors.

Why they did not just make their own in-house Wayland compositor I simply cannot fathom. They would have gotten the best of both worlds... maybe they will, in a similar way to upstart.

[Anarchy]

This looks like what the (GNU/)Linux world has been doing for the past 5 years, through the graphics stack refactoring, and projects like wayland/mir.


Wayland and Mir bring whole new paradigms to the display system, and X has many issues from being designed for hardware of the last century. That's why it's being deprecated.
But Mir doesn't bring anything compared to Wayland, so the comparison is pointless. The only thing it could bring is "swiftness" for Canonical (at the cost of less reusability of other projects), but even that it failed to do, given that Mir has not come to market any before than other Wayland compositors.

Why they did not just make their own in-house Wayland compositor I simply cannot fathom. They would have gotten the best of both worlds... maybe they will, in a similar way to upstart.

What's there for you not to fathom? The bulk of serious bugs and complex problems are drivers, X.org and toolkit related. Moving to waylaid will not solve anything because then they'd be dependent on upstream to fix errors, make sure they follow guidelines, implement new stuff, etc. It's going to be the same mess like it's with X. Having their own compositor, display server and tightly integrated commercial toolkit, i.e. qt, this leaves only the drivers out of their direct control.

There is a reason why Apple controls everything from top to bottom. It gives them absolute control over the software and hardware, which makes for an exceptionally stable experience. Canonical goes for the desktop crown, so they want to control as much as possible of their software stack.

[erendorn]

What's there for you not to fathom? The bulk of serious bugs and complex problems are drivers, X.org and toolkit related. Moving to waylaid will not solve anything because then they'd be dependent on upstream to fix errors, make sure they follow guidelines, implement new stuff, etc. It's going to be the same mess like it's with X. Having their own compositor, display server and tightly integrated commercial toolkit, i.e. qt, this leaves only the drivers out of their direct control.

There is a reason why Apple controls everything from top to bottom. It gives them absolute control over the software and hardware, which makes for an exceptionally stable experience. Canonical goes for the desktop crown, so they want to control as much as possible of their software stack.

Wayland is only a protocol, so there's very, very little "upstream" (only protocol to C conversion), and it's stable already. And the whole current featureset of Mir is covered by it.
The only thing that is different between a wayland and a non wayland compositor (both being written by the same company), is that for the non wayland one, you need to write the full toolkit support yourself (they will use the same drivers, the same compatibility code for android and X, most likely the same libs for input, which does not leave much else).

Last time I checked, Qt was not under Canonical's control (??), so they will need to have a downstream Qt fork (plugin) anyway, so yes, I don't get what the benefit is for them (while Qt of course already has wayland support too, so maintaining a downstream of that would be much less work).