Linux 2.6.28 Promises Year-End Presents

darthcamaro writes "Little penguins all around the world are waiting for Penguin-Master Linus Torvalds to deliver some Glogg inspired Xmas cheer in the form of the new 2.6.28 kernel. Among the innovations in 2.6.28 are ext4 as stable, wireless USB drivers, better KVM support and the GEM graphic memory management technology. 'We now have a proper memory manager for video memory, the GEM [Graphics Execution Manager] memory manager,' Greg Kroah-Hartman said. 'This gives Linux much better graphics performance than it previously had.'"

Comments

The new graphics

[chunk08]

Haven't read read the article yet but it does it require doing things differently in drivers or user-land software?

Re:The new graphics

[chunk08]

Hm, sry to reply to myself but according to Wikipedia it seems that the drivers have to be rewritten to support GEM. Still not sure about user-land software tho...

Re:The new graphics

[rmdir -r *]

It's transparent to userland, it should just mean the next generation of drivers will be faster. Of course, if the drivers are done wrong, it'll mean the next generation of drivers will be crashy, so...

Re:The new graphics

[diegocgteleline.es]

I recommend reading this link to get an idea of what's going on in the Linux graphics stack:

"So currently there is not one field where construction done but several. These are 2D Acceleration, Memory Management, 3D Acceleration and 2D Modesetting. And they are all being worked on at the same time to speed things up.

But the problem is that more or less all of these depend on proper Memory Management, which is also the hardest thing to get right.

Now lets look at how Xorg works today; every Xorg driver implements its own way of memory management and provides the DRI1 functionality when it comes to 3D. Furthermore it is responsible for modesetting, which is quite suboptimal, since some perliminary modesetting is already done in kernel, so it can output messages during bootup. The Xorg driver resets the hardware again when it is loaded.

Kernel Based Modesetting

In order to solve this duplication the modesetting code is about to be moved into the kernel, so the hardware can be setup once and for all. But since modesetting involves memory management which is not done properly yet too."

Re:The new graphics

[grantek]

Intel staff were the ones mainly responsible for implementing GEM, so their driver supports it. The open-source ATI drivers recently got a layer of glue to use GEM on the outside without changing much of the TTM-based code that was on the inside. I don't know what nouveau is up to, but the nvidia blob has had a lot of memory management stuff implemented independently for a while now in their X driver.

Phoronix follows a lot of this stuff well.

Re:The new graphics

[dow]

What should be important is that maybe next gen games should be released on Linux as a platform equal to Windows.

I was a long term Linux user, who went to XP just for the games. My gaming rig is waiting an RMA on a PSU, so rebuilt an old system and installed Slackware.

On an older machine with slower drives and a quarter the Ram, the responsiveness of the OS is amazing. If mainstream games were released for Linux I'd have no choice.

Sadly, I mainly use computers these days for relaxation, shopping and play, and if I'd continued as I set out, would no doubt be a full time Linux user... However, as a gamer, I put up with XP64 as a day to day OS.

Re:The new graphics

[mrchaotica]

XP can't do this, though. This is a DirectX 10-equivalent feature; it's comparable to Vista, not XP.

Re:The new graphics

[Antique Geekmeister]

The NVidia blobs remain a big problem. It's not the kernel blob: it's their replacement by setting aside of the OpenGL libraries, used to access the NVidia features. This part of the NVidia process destabilizes every OS that it touches because any updates to those libraries overwrite the NVidia libraries and seriously break your graphical setup.

It's theoretically possible to rewrite the Xorg and Mesa packages to cooperate with this by bundling the Nvidia package and its libraries to a package matching the Mesa components and install one or the other, but no one has yet done so. So NVidia remains a dangerously unstable set of tools to install in any sytem that gets any updates otherwise.

Re:The new graphics

[GigaplexNZ]

...by bundling the Nvidia package and its libraries to a package matching the Mesa components and install one or the other, but no one has yet done so.

Gentoo does an excellent job of managing the upgrades - the Mesa and Nvidia drivers are installed to different locations and symlinks are used to choose the right one, with a nice wrapper script to make it easy to choose what one you want with eselect.

Re:The new graphics

[Antique Geekmeister]

Really! Good, I did not know Gentoo had done this. I'm generally unhappy with such behavior, because it can make package tracking and integration quite nightmarish, but this is a single instance where it could be very useful to work around NVidia's stupidity. I hope that Gentoo has entirely ignored NVidia's truly awful installer and uses their own?

Re:The new graphics

[johannesg]

What should be important is that maybe next gen games should be released on Linux as a platform equal to Windows.

You are delusional if you think technology makes any kind of difference at this point. It is all about market share, and Linux simply doesn't have any in the games market.

Re:The new graphics

[Choozy]

Yes but does it run linux... oh wait..

Re:The new graphics

[Nick Ives]

Debian Sid is fine with the Nvidia drivers too. No conflicts with Mesa as long as you use the packages from non-free. Apt also recompiles the kernel driver every time you install a new kernel!

Re:The new graphics

[morgan_greywolf]

Ditto on Ubuntu. The package management scripts are very intelligent in regards to Xorg and Mesa updates when the NVidia drivers are installed. Kernel updates, Xorg updates, and Mesa updates will all trigger init scripts that re-install the NVidia restricted drivers.

Re:The new graphics

[Elektroschock]

Basically the next generation games consoles will be based on Linux + some API.

Re:The new graphics

[dow]

I still do dual boot my rig... what I meant was that day to day I'm using windows, then if my friends are up for a game theres no need to reboot. Not to mention the way we keep track of where people are playing is X-Fire. If a friend is playing a game somewhere, you can just click to start playing the same game on the same server.

I have a server running Linux too.

If my world revolved around coding and system administration I'd be the same. Apart from the games, the machine is full of free/OSS software, but it doesn't run at the same speed as on Linux.

Re:The new graphics

[GigaplexNZ]

While Ubuntu was good at handling system upgrades with the Nvidia drivers, it was abysmal at handling the circumstance where I temporarily removed the Nvidia card and started using the onboard Intel graphics - I had to uninstall the Nvidia driver before X would start with the Intel driver. Gentoo gave the option to switch which library was active while both were currently installed. Granted, not a common situation but annoying nonetheless.

Re:The new graphics

[Vintermann]

Who is gong to lead the battle cry for such a platform, exactly?

Developers, maybe? I don't know how they presently do it when releasing for all three platforms, but I assume that a common operating system would help.

Re:The new graphics

[brizzadizza]

Why can't a games developer develop in linux and send out the game with some sort of virtual linux installer? Then for the linux crowd the software would be native and for the other crowds, no one would be the wiser.

Re:The new graphics

[spandex_panda]

doesn't work for me ... I usually have to manually uninstall and then reinstall (all apt-get easy) the nvidia-glx-177 package. This may be due to my updating 8.10 from alpha through to release (current).

Re:The new graphics

[3vi1]

Sounds like your DKMS install is broken. There were some problems with the nvidia packages in alpha and they were not removing the proper files they had installed in the DKMS directory.

Have a look in /var/lib/dkms.

Re:The new graphics

[Zoolander]

Mods: this is not a troll.
It's the truth.

Re:The new graphics

[cinderblock]

Then they will be pushed to developing games for linux.

With games developed for a linux based console, it would be an easy switch (if not trivial) to distribute them to linux PCs. That's part of why the windows + xbox releases are so easy to make.

Though M$ likes that they have control over industry and won't give it up without a fight.

Re:The new graphics

[cyphercell]

I would think that both Nintendo and Sony making their games more easily ported to Linux and/or Mac would be a little more like tit for tat than a complete impossibility. BTW Linux runs on a shit-ton of appliances.

Re:The new graphics

[Elektroschock]

The Xbox as a competitor is a reason why they won't take a Windows platform flavour. The enemy of your competitor is your friend.

It doesn't matter if the platform is released but the game console related improvements will lead to changes to the platform and the kernel, just as Windows95 was not really suitable for games Linux will be transformed through its use for gaming.

Re:The new graphics

[Elektroschock]

Right now the Wine developers do not even implement it because they feel its irrelevant.

And the console competitors that are expected adopt Linux as a base OS? Just another argument more to standardise on another platform because they do not want to get locked in. Is WII Xbox compatible? PS? -- ok ok next question.

Re:The new graphics

[Elektroschock]

The rationale is the same as with routers or Mac OS X

Let's put yourself in the seat of Nintendo, Sony etc. Why not copy the Apple OS X strategy? Don't waste your resources on the OS base level, reuse what is there.

This doesn't exclude a proprietary layer on top of it. After all they don't sell software but consoles.

Re:I for one...

[TeknoHog]

...welcome our old Unicode-challenged Slashdot. BÃrk bÃrk bÃrk!

Re:I for one...

[FugitiveMind]

"charset=iso-8859-1"

Welcome to 2000. :|

No answers to your question in TFA.

[rts008]

""We now have a proper memory manager for video memory, the GEM [Graphics Execution Manager] memory manager," Kroah-Hartman said. "This gives Linux much better graphics performance than it previously had."

The video improvements in Linux also extend to power utilization for graphics. Red Hat Fedora Project Leader Paul Frields told InternetNews.com that the 2.6.28 kernel enables reduced power consumption across the video driver subsystem in the vertical blanking routines, which will be helpful to mobile users."

That is all that is mentioned (above quote) about the state of 'the new graphics' in the new kernel.

Re:No answers to your question in TFA.

[TheRaven64]

Although not part of GEM, one related improvement is moving modesetting into the kernel. Currently, when you switch to an X11 VT, X11 requests the console be set back to VGA then initialises it to the correct mode itself. This is really horrible, and doesn't play nice with power management (because the kernel doesn't know anything about the GPU state, so can't easily save and restore it). The modesetting branch in X.org has been defining some clean kernel interfaces for doing this, simplifying both the kernel and the X server in the process (since both previously contained lots of special-purpose code for doing the same thing for each device).

As with GEM, this isn't a Linux-specific thing, it's driven by X.org and being implemented on Linux, *BSD and Solaris.

Re:No answers to your question in TFA.

[geert]

This is on platforms that boot in VGA test mode.

On other platforms, we had kernel mode setting since 1994, through the frame buffer device interface and the frame buffer console. We even got kernel messages printed to the console when running X, but people didn't like them, so they were disabled when the console was in "graphics" state.

Well

[Anonymous Coward]

I heard they were going to include RieserFS but they killed that idea because it would be murder to include it. The developers said they would have to butcher the kernel in order to get it to work.

Re:Well

[Andrew Kismet]

5th unique AC in a row -

I thought it was funny, guess it rubbed someone with mod points the wrong way. And the moment one person mods it to 0 or -1, it stays there.

Re:Well

[poopdeville]

Yeah, that and this process didn't work too well in Mrs. Rieser's job hunt.

Re:Well

[Anonymous Coward]

5th unique AC in a row -

Third AC again.
Looks like you're an AK not an AC. But you are correct the right answer is - funny... funny for three points.

Re:Well

[jejones]

I'd say "overrated" would be more appropriate. Is there any chance for a knob to twist to block all Hans Reiser "jokes"? The first aleph-null of them were kind of funny.

Re:Well

[AbRASiON]

Murder isn't funny!
Unless clowns or huey lewis and the news are involved.

Re:Well

[Anonymous Coward]

You must be new here..

Re:Well

[mr_walrus]

slashdot comments murder the english language all the time.
yet it makes me laugh.
murder IS funny!

Re:Well

[Hal_Porter]

I'd say "overrated" would be more appropriate. Is there any chance for a knob to twist to block all Hans Reiser "jokes"? The first aleph-null of them were kind of funny.

No but you could murder your wife if your computer annoys you.

Re:Well

[Elektroschock]

Indeed it demonstrates the cowardness of the Kernel developers. Ext4 included but not Reiser. A sad expression of the moral state of the kernel project. After they delayed Reiser inclusion for ages.

Re:Well

[UncleTogie]

I don't understand how it's apparently so fashionable on /. to make disgusting "jokes" about horrific things.

Probably 'cause the best choice in dealing with a horrific situation boils down to humor. The other choices being uselessly wailing, freaking out, or other emotional responses that do nothing to change or improve the situation. Finding the humor, however dark, and getting a chuckle is probably the best defense mechanism we humans have ever developed to stress...

Re:Well

[geminidomino]

Murder isn't funny!

Liar. You can't spell slaughter without laughter, after all!

Re:Well

[Hans T. Reiser]

Will you guys quit impersonating me already?

Re:Well

[garaged]

A lot of people thinks that reiser's code is not really that friendly, besides, Hans is known to work on what he wants, not what is good for people, and the newer version of reisers FS is not finished...

that is enough explanation for me, even when I expected Hans to be inocent, and I actually kind of admire him for his work.

Nice start...

[TheNetAvenger]

Not quite Vista's WDDM abilities in dealing with GPU RAM, but a nice start that people other than MS are actually taking GPU RAM allocation seriously beyond simple context swtiching.

Re:Nice start...

[Anpheus]

Because the UAC window is on an independent desktop that other applications cannot interact with. The only possible flaw is if something has installed itself as a mouse or keyboard input driver, I believe. But doing that will spawn a great big red unsigned driver prompt.

Re:Nice start...

[FishWithAHammer]

It prevents shatter attacks. It can be turned off and a UAC window made to act like any other.

Re:Nice start...

[TheNetAvenger]

Maybe I'm misunderstanding your assertion, but hasn't MacOS X had universal GPU RAM management for many years? I don't think MS has any monopoly on this... it was my impression that it was just Linux that was on Microsoft's heels playing catch-up.

Yes you are misunderstanding, and NO the Mac has not...

OS X uses the 3D GPU as a bitmap composer for the display, and that is it.

OS X's composer is years behind most Linux desktop composers as well as Vista's DWM/Aero. Vista's DWM for example is Vector/Bitmap based, and works with the WDDM of Vista that gives it a lot of power. (WDDM is the new driver model in Vista)

Here are some of the things Apple needs to add to catch up to even Vista.

- GPU RAM Virtualization/sharing (something kind of like what they are trying do with the Intel chipsets and Linux in this article - except Vista does this over the AGP/PCI bus with any Video card and works with or without dedicated GPU VRAM.

- GPU Scheduler - In Vista, the OS, not the applications controls the GPU, and Vista brings pre-emptive multi-tasking to the GPU. (And no this is not like OpenGL applicaiton yielding/cooperative multitasking, as DirectX also does what OpenGL does. This is an OS level management system that opens up a new way of thinking beyond one 3D application on screen at a time concepts that don't depend on applications yielding the GPU. Kind of like the move to the 32bit era where the Intel CPUs offered a pre-emptive scheduler.

(Example: several games on screen at once in Vista, set transparent with a HD video waterfall playing in the background and losing very few FPS in each game and Aero also using the 3D GPU to do its things, like compose the Game Windows with a transparent waterfall behind them and do a shared texture combine write to the video card.) - This is not something you want to try on a Geforce 5200, but it will work, and on newer video cards, even the 7900 series from Geforce, you can do some really amazing things when running multiple games 'viewable' on the screen at once.

- Legacy application 3D acceleration. Apple tried to get this going with 10.4 as an optional switch, but it was too buggy and scrapped as a feature for 10.5. This means that OS X still renders content using good old fashion legacy 2D GPU features or SSE Intel extensions. On Vista, even Windows 3.1 applications get a performance boost as GDI drawing, Font Rendering, and even internal bitmap APIs are shoved through the 3D GPU because it is significantly faster than older 2D GPU rendering methods.

- Vector composer. On Vista when it is running newer WPF applications, instead of the DWM getting a bitmap that is composed to the final render of the screen, the WPF applications tell the DWM/Aero what changes are made, usually vector based (XAML), and the Vista composer makes the changes at the composer level instead of the application having to redraw the application and send a new bitmap of the Window to the composer to assemble. (This is also why RDP (Remote Desktop) on Vista is faster and more featured than XP, as it works at the DWM level and a lot of the operations sent over the network to render the screen are vector based and lightweight, leaving the client to do the heavy rendering instead of passing bitmaps all the time. This is why you can do Aero Glass and WPF 3D over a slow RDP connection remotely.

----
Ok, I am going to stop here, as I am writing this off the top of my head and it would better if you would just visit technet at www.microsoft.com and lookup the Vista WDDM and DWM and WPF technologies.

The whole driver and video changes in Vista were dramatic and borrowed ideas from the XBox 360 development team and do some really impressive things, even though MS didn't put much into the 'cute' uses of it in the UI like they are doing with Windows7.

---

Linux/*nix also has some good composer technologies that make OS X's composer pretty sad in comparison.

With Linux there are still some major driver and kernel level hurdles

Re:Nice start...

[TheNetAvenger]

The independent desktop for UAC is true, and it would require a driver to interact with it. (And yes a big red driver prompt, warning, etc)

Also the UAC 'desktop' runs at TrustedInstaller or System level security, so it takes a system level driver to respond to it. (It is not impossible, but it would be pretty hard for a driver exploit to get past this, as the driver would have no way to 'see' what was on the screen to know where to click, etc.)

A side note, Vista supports multiple desktops (and even varied security) launched in a single user login, even though there is no UI for this feature (One can be downloaded from Sysinternals).

I'm not sure how useful this is, but as the Sysinternal's utility allows, it creates 'virtualized' desktop without the virtualizing, as the desktops are independent full desktop contexts.

So each desktop can run a separate version of Explorer, etc - within the same user login.
(The term 'desktop' here is used loosely as it pertains to GUI process context in Windows.)

This is not a revolutionary feature, just something kind of interesting, as it is separate desktop contexts within a user login within a multiple user environment.

Re:Nice start...

[TheNetAvenger]

Why, every time Vista puts up a UAC requests, does the whole screen go black for a couple of seconds, as if the screen mode has been changed?

Yes it is a security thing.

Some Video drivers - especially older ones, do the blank and screen switch less elegantly than others.

Since SP1 it is something you can turn off, you will still get the UAC, but it doesn't block the screen or switch to a new desktop context for the UAC prompt.

(It is a little less security, but worth the trade off for some people, especially with crazy video drivers that flash their eyeballs on an analog monitor.)

The setting is in the group policies under - Control Panel\Administrative Tools

Re:Nice start...

[simcop2387]

sounds almost like Xnest but without the "nest" (e.g. window), kinda neat.

Re:Nice start...

[644bd346996]

In Vista, the DWM prevents GDI commands from being accelerated, which represents a regression relative to XP. The compositing does mean that moving windows can be smoother, but when the contents of the window change, Vista is liable to be slower (in addition to using twice as much memory for that window due to a pixel format mismatch between DirectX and GDI). Thus, legacy applications are not likely to be able to draw any faster on Vista than on XP.

Compare this with Quartz 2d Extreme, now known as Quartz GL (and not abandoned as you say). Existing applications can have their drawing accelerated at the cost of a potential for glitches or performance problems (due to limited bandwidth from the CPU to the GPU), and new applications can enable acceleration at the discretion of the developer.

Re:Nice start...

[Anonymous Coward]

GDI graphics are not hardware accelerated on windows vista (WDDM 1.0), you can read this everywhere on the internet. However, GDI hardware acceleration will return with Windows 7 (WDDM 1.1). I suspect this is why Windows 7 will be much "snappier" than Windows Vista. Also memory usage will be reduced, because GDI graphics can be kept in the graphics memory of the videocard, instead of as a bitmap in memory and as a bitmap in the frame buffer (and off screen acceleration memory?).

Re:Nice start...

[Anonymous Coward]

Uh, the Intel 386 didn't offer anything that somehow made preemptive multitasking possible.

Re:Nice start...

[Bert64]

Run multiple copies of Xnest at the same resolution as your physical screen within multiple virtual workspaces... You can run them as different users too.

If you have multiple physical screens you can run separate instances of X11 on them with their own window managers and potentially even their own input devices, instead of having one wm spanned over multiple screens...

Re:Nice start...

[Bert64]

You can force it on system wide too, at the risk of glitches and in certain cases it can be slower.

Re:Nice start...

[Rutulian]

Good stuff. To be fair, though, the dismal state of linux graphics has been known and complained about for a long time. The problem is, there have been a lot of political hurdles (none of this was possible when Xorg was still Xfree86, and w/o cooperation from companies that write drivers it would also be useless), as well as technical hurdles. I don't remember all of the details, but there was an article on freedesktop.org talking about all of the different things people were working on. Several projects were started over from scratch several times because they realized midway through that it wasn't going to work. So the end result is that development has been slow, but when it is finished it is going to be "done right" and not slapped together just to ship something out the door quickly.

Re:Nice start...

[TheNetAvenger]

In Vista, the DWM prevents GDI commands from being accelerated, which represents a regression relative to XP. The compositing does mean that moving windows can be smoother, but when the contents of the window change, Vista is liable to be slower (in addition to using twice as much memory for that window due to a pixel format mismatch between DirectX and GDI). Thus, legacy applications are not likely to be able to draw any faster on Vista than on XP.

I know there have been discussions on this before, but it isn't a complete either or...

Some GDI functions are not shoved through the 3D GPU nor the 2D side of the GPU (assuming there is one), and yes you are correct that these are CPU processed.

However, there are a lot of GDI functions that are shoved through the 3D GPU more than compensating for any CPU rendered GDI functions.

For example we could go through a list, and there are some line drawing, bitmap/DIB drawing, and scattered GDI and especially GDI+ functions that are rendered on the 3D GPU. (Font rendering is 3D assisted as well, no matter if WPF or GDI calls the functions.)

Also when you get into GDI+ level functions like transparency/layered windows, these are very much shoved through the 3D GPU for rendering.

And again, these are actually GDI routines being hosted/processed by the 3D GPU, not just the end result of the 3D Composer as you suggest I am conflating.

Re:Nice start...

[TheNetAvenger]

GDI graphics are not hardware accelerated on windows vista (WDDM 1.0), you can read this everywhere on the internet

As I responded above, this is a misconception... Yes the GDI is no longer 2D GPU assisted, but that doesn't mean that NONE of the GDI/GDI+ functions are shoved through the 3D GPU.

Look up font rendering, DIB/Bitmap functions, GDI+ calls that do anti-aliasing all the way to layered and transparent Windows that are all processed using the 3D side of the GPU.

However you are also correct that WDDM 1.1 revisits the GDI/GDI+ functions and add more to the mix that will be 3D GPU assisted, but again not EVERYTHING will get an equivalent replacement as there are times it is just faster to process the GDI function on the CPU.

So WDDM 1.0 is 'some' and WDDM 1.1. is 'more' GDI 3D acceleration.

BTW If I just threw up my hands and said, yes you are 100% correct, you are still proving the point I was making that Apple is light years BEHIND Windows in this regard and has no plans to bring acceleration to legacy GUI drawing function on OS X.

Re:Nice start...

[datadigger]

Uh, the Intel 386 didn't offer anything that somehow made preemptive multitasking possible.

Huh? What more do you need for preemptive multitasking than a timer interrupt ?

Re:Nice start...

[TheNetAvenger]

I'd love to scrap the Direct3D code I have to use now on Vista in order to get decent blitting performance.

Why are you even hitting into DirectX, you should be shoving it through .NET 3.5 and WPF, as this is even accelerated on XP via 3D acceleration.

The whole WPF API set is there specifically so you don't have to drop to DirectX unless you are trying to squeeze the last ounce of performance out of a high end game. You should even be able to obtain a Windows 'texture/image' and use it as a brush via WPF, getting the performance you might be needing that GDI just can't do.

If I am reading your post correctly, you might want to consider tapping into the DWM API as well. From there you can request and process 'thumbnails' of running application windows. And they can be any size from tiny to original to even enlarged version of the source Window Bitmap, and they are also fully live previews, so when you tap the DWM 'thumbnail' of another application you can get as well as directly render the Window to your own form/window at the DWM level. (3rd party applications like the expose' clone use this to scatter the Windows like on OS X with the contents still being live and responding as if they were the original application Window.

There are a few aspects of WDDM that are not API accessible at least in Vista, and hopefully this changes with Windows7 and its WDDM API updates to expose the DWM and in between operations. WPF also has some abilities in this area, but it is still young.

Good luck...

Re:Nice start...

[Anonymous Coward]

It was much easier (and guaranteed to be fast enough) to switch to Direct3D rather than to rewrite the application using .NET and WPF (the app existed before Vista was released), although this is very off-topic with regards to GDI acceleration.

I'm not interested in fetching data from on-screen windows/surfaces so I don't have any use for the DWM API (even though I was already aware of it, including it's limitations). All I wanted was fast GDI blits (in any way) but I couldn't pull this off. ATI and Microsoft employees (reluctantly) confirm that there is no longer any GDI acceleration in Vista, so you'll need to come with something concrete (such as a link) in order to convince me of the opposite.

Re:Nice start...

[ahabswhale]

You've been going on and on about how much better Vista is for 3D over OS X, XP, and Linux and it all sounds great. There's just one problem with all of it...Vista runs games slower than XP. So I guess all that cool new shit in Vista is pointless because XP still kicks its ass in 3D rendering. Maybe I can't run 3 games at the same time in XP but who gives a shit...I can only play one at a time anyway.

Let me know when vista catches up to XP in 3D performance and then come back and impress me with all the new tech. Until then, it's all bullshit.

Re:Nice start...

[badkarmadayaccount]

A kernel implementation? *ducks*

Re:Nice start...

[datadigger]

A kernel implementation? *ducks*

Exactly!

Re:Nice start...

[datadigger]

Parent posted said that 386 offered some new feature that enabled preemptive multitasking. Ie. that it was not possible in earlier CPUs.

I read that, and I wondered which feature that might have been.
It was possible on earlier CPUs: a long time ago I wrote a tiny preemptive multitasking kernel on a 8086 as an exercise.
Preemption of the running task and switching to (dispatching) another one was done in an interrupt service routine, triggered by a timer.

Re:Nice start...

[TheNetAvenger]

to switch to Direct3D rather than to rewrite the application using .NET and WPF

You don't have to rewrite to access WPF or .NET. There are tons of easy ways to tap the features of WPF from a classic GDI or WinForms project, with less work than recreating a blit operation in DirectX like you are doing.

Additionally, have you looked at Direct2D?
http://blogs.technet.com/thomasolsen/archive/2008/10/29/introducing-the-microsoft-direct2d-api.aspx

Good Luck...

Re:Nice start...

[ebichete]

I think the parent poster is confusing a memory protection feature with multi-tasking. With memory protection the independent tasks could be prevented from stomping on each others memory spaces.

To clear somethings up

[Anonymous Coward]

GEM is short for Graphics Execution Manager, it is a graphics memory manager for the kernel written by Intel.

If graphics device drivers want take advantage of GEM, then they need to add some code for GEM in the device driver.
A memory manager for the graphics memory is very useful because it allows direct rendering and direct redirected rendering and such.
This means you can now do things "the real way" which have previously either not been possible, or been done using some dirty hack such as indirect rendering.

Re:To clear somethings up

[squiggleslash]

Kind of a shame, I was hoping they were integrating the Digital Research Mac-like User Interface system for DOS (and the Atari ST) into the kernel, just to annoy purists...

Re:To clear somethings up

[Neil Hodges]

We had one of those when 2.6.9 came out, but not since. A shame.

Re:To clear somethings up

[cp.tar]

This does not sound like a bad idea, though: moving a part of the graphics into the kernel only makes the kernel a little more complex (as it has already had parts of the code in order to use the screen while booting), and at the same time, X can become less complex and thus both faster and less error-prone.

Re:To clear somethings up

[1mck]

I'm kind of getting the feeling that Intel is seeing the writing on the wall, and understanding that Linux isn't going away, and could quite possibly in the very near future take over a considerable amount of the desktop share especially for businesses that want to severely reduce their IT budgets.

Re:To clear somethings up

[Mad Merlin]

A memory manager for the graphics memory is very useful because it allows direct rendering and direct redirected rendering and such.

A definite step in the wrong direction.

One of the things I've always liked about *nix is the separation between kernel and graphics.

No matter how horked X is, I the system always boots in text mode console and work to repair X or a driver, install new software, etc, and even accomplish things with Mutt and links2.

Then, when I'm ready to "go graphical", simply run startx.

You don't really understand the consequences of doing kernel mode setting then. None of your use cases will be impacted by the addition of kernel mode setting, except that you'll be able to more easily get different resolutions out of your virtual consoles (you can already do that with framebuffer consoles, sometimes, depending on the hardware, and what driver you're using with X (if any)).

Re:To clear somethings up

[Anpheus]

Let me put it to you in a way that should impress you: Kernel modesetting allows things like the Windows BSOD and the Mac Kernel Panic, which means that when your kernel dies you can get a direct, immediate error message with details.

Those STOP messages in BSODs are pretty important for figuring out what's wrong with Windows, I imagine with the open kernel of Linux, you could have much more detailed errors.

Re:To clear somethings up

[TeXMaster]

No matter how horked X is, I the system always boots in text mode console and work to repair X or a driver, install new software, etc, and even accomplish things with Mutt and links2.

Then, when I'm ready to "go graphical", simply run startx.

What makes you think it would be any different now? The only difference is that now the kernel provides additional hooks and a consistent interface for managing the GPU resources too, which means that you can finally have a much improved integration between graphical systems (e.g. GPU-specific framebuffer consoles and (one or more) X sessions in (one or more) virtual terminals).

Nothing changes as long as you don't start graphical subsystems. And after you do, the difference is that instead of having a distinct HAL in X you use the system one, so that you get cooperation instead of fighting between code that touches the same subsystem.

Re:To clear somethings up

[Anonymous Coward]

That really does not impress me.

You're not supposed to be impressed, you're supposed to be able to easily fix your graphics (or any other driver/configuration) setup with more-or-less your expected setup. Non-expert users will be impressed by that. Or at the very least less pissed off by the problem they're experiencing.

Cry me a river.

Yeah, fuck all those people who don't want to learn X configuration file formats off-by-heart! But I bet you'll be the first person bitching and moaning when vendor X doesn't provide Linux drivers and vendor Y's software doesn't support Linux. Newsflash genius, it's the masses that bring the recognition and the cash to make the vendors take notice. If you ever want Linux to do all those things that "Year of Linux" spouters have been droning on about for the last decade you're going to have to realise that making Linux useable, maintainable, and fixable by average Joe's with as little fuss as possible is the only thing that matters to the long term future of Linux as a desktop OS.

Re:To clear somethings up

[Kjella]

This allows people who are not so technically proficient to fix their computer without having to resort to using a command line.

Cry me a river.

May you be forced to debug some WTF message without any browser but lynx to help you. Of course, you're probably among the 1% that knows that lynx exists and is able to navigate your way to google and find the answers without a mouse to click. Great that you're built that way, but l33tnix is over there ------------> and the rest us of would like a system that isn't more arcane and user-unfriendly than necessary.

Re:To clear somethings up

[Hal_Porter]

The reason for 640x480 is because safe mode uses the generic VGA driver and 640x480 is widely supported. Actually since XP the generic VGA driver which allows higher resolution modes. No hardware acceleration of course since every manufacturer implements that in their own way.

Actually XP had a cool trick with graphics drivers. If a thread hung inside the manufacturer provided accelerated graphics driver the GDI would switch to the generic VGA driver and pop up a message explaining what had happened and prompting you to reboot. I.e. it could switch from 1024x768 accelerated 32 bit color to 640x480x16 color and keep running, even though a kernel mode thread had hung.

Re:To clear somethings up

[cyber-vandal]

making Linux useable, maintainable, and fixable by average Joe's with as little fuss as possible

I'm still waiting for Windows to get to that point never mind Linux.

Re:To clear somethings up

[Nutria]

You don't really understand the consequences of doing kernel mode setting then.

Why am I not surprised???

None of your use cases will be impacted by the addition of kernel mode setting,

Interesting. Will the GEMified drivers require DRI/DRM? (Everything I've read about it seems to imply that.)

except that you'll be able to more easily get different resolutions out of your virtual consoles (you can already do that with framebuffer consoles, sometimes, depending on the hardware, and what driver you're using with X (if any)).

By using startx, I have no need for VCs. (I tried booting in different vga= modes, but it was too weird for eyes that are so used to 24x80...)

Re:To clear somethings up

[Mad Merlin]

Interesting. Will the GEMified drivers require DRI/DRM? (Everything I've read about it seems to imply that.)

The converse, actually. Using DRI/DRM will require GEMified drivers. GEM doesn't really come into play without DRI/DRM.

Re:To clear somethings up

[Cyberax]

No problem. Just add "nokms" parameter to the kernel command line and it'll start kernel without kernel mode-setting support, in a plain old console.

Re:To clear somethings up

[Yfrwlf]

Actually, Links2 allows you to use a mouse. :P

But seriously, fully agreed, and it's fully possible to have a system that is both user friendly and powerful/stable/featureful/etc. Having GEM in the kernel as a module is fine if the kernel can deal with GEM getting itself borked, which is how everything should be, there should be fail safes galore if things are programmed intelligently from the ground up.

Re:To clear somethings up

[tyrione]

In short, Linux is one step closer to what OS X has done for several major releases.

Re:To clear somethings up

[Yfrwlf]

"instead of having a distinct HAL in X you use the system one"

Wasn't that their point? That it's a separate system, so that if it fails, you'll still at least have the command line?

You can always make the argument that, well, if the code is good, then it should work, so what is the issue here that everyone is beating around the bush about? I think it's stability via intelligent programming. If you have the command line as a failsafe for when X fails, it gives you extra protection against bugs, which will always be there somewhere. You shouldn't just expect code to be written correctly, you should fortify yourself for when things break. If this can still be done even with kernel mode setting and such, if the kernel can switch to a failsafe if GEM or whatnot fails, then that will certainly help. Simplifying software stacks and creating APIs for performance and ease of programming = good. Removing failsafes = scary, unless they added some other failsafe somewhere else or whatnot, or maybe there's already one there.

And no, don't say reformat reinstall, that's the Windows failsafe. :P

Re:To clear somethings up

[timmarhy]

"This allows people who are not so technically proficient to fix their computer without having to resort to using a command line. Cry me a river."

people like you are why linux won't ever dominate the computer industry. try growing up and learning from your mistakes.

Re:To clear somethings up

[Nutria]

If you have the command line as a failsafe for when X fails

Thank you for elucidating my intent. Foolishly, I assumed that people understood what I meant.

Re:To clear somethings up

[Nutria]

people like you are why linux won't ever dominate the computer industry. try growing up and learning from your mistakes.

It is well documented that people who don't know WTF they are doing regularly screw up their computers and allow them to become rampantly infected by malware.

Thus, it is only rational to think that people who don't know how to administer a computer should not perform administrative tasks and have administrative privileges.

If that means that 95% of households that currently have computers shouldn't have them, so be it.

Elitist? Hell yes!

Re:To clear somethings up

[Nutria]

Just add "nokms" parameter to the kernel command line and it'll start kernel without kernel mode-setting
support, in a plain old console.

What about all the gooey goop (Plymouth & Wayland) that RH and Ubuntu are layering on top of the boot process to make it dumber (I mean more user friendly) which specifically depend on kms?

Re:To clear somethings up

[StuffMaster]

I've actually had this happen to me. After a couple of years, my video card decided to 'degrade' its performance. It would sometimes lock up (completely, no numlock or anything) when playing certain games, but after 30-90 seconds, Windows would change modes to tell me there was a problem with my card. Pretty nice if you ask me.

I've since lowered the resolution to avoid this.

Re:To clear somethings up

[Nutria]

The converse, actually.

That's what I meant. Really!

Anyway, is DRM/DRI essentially what the nvidia binary driver and OSS layer currently do?

(I use it, but it doesn't get loaded until needed by X.)

Re:To clear somethings up

[Cyberax]

They also can be disabled. Nobody's forcing you to use them.

Also, I _like_ the smoothness of console switching which KMS gives. It's way too cool.

Re:To clear somethings up

[kelnos]

You'll still have a failsafe command line on your VC when X fails. That hasn't changed, and won't. Quit whining about something you don't understand.

Re:To clear somethings up

[pizpot]

clean the dust from the fan/heatsink, or re-appy the thermal paste on the GPU.

Re:To clear somethings up

[pizpot]

uhhh, try a fresh install of 8.10...

Re:To clear somethings up

[Nimey]

Plus 640x480 at 16 colors is actually part of the original VGA specification from the late 1980s. A card would have to be badly designed indeed (or terribly old) to not support that.

Re:To clear somethings up

[Anonymous Coward]

They're not trying to impress _you_. They're trying to solve the real world problems they're facing.

Re:To clear somethings up

[radish]

That page is a hoot. In general I groan whenever a software homepage has a "Philosophy" section, but the installation instructions more than make up for it...

Installation for Non-Programmers (emphasis mine)
      1. FBUI resides inside the 2.6.9 kernel, so the first thing you must do is to get the kernel, un-tar FBUI in its directory, select the necessary options mentioned in the README, then make the kernel and update your loader to let you boot the new kernel. (I will offer a precompiled x86 kernel later.)
      2. You also need to tell your boot loader to switch to the VESA console during booting. In LILO use the expression vga=792 for a 1024x768 display or vga=789 for 800x600.
      3. Then you boot with the kernel. Next you need to set up the PCF font directory, populating it with fonts from the X distribution, making sure to uncompress them. The PCF font reader is really just a temporary chunk of code so I'm not going to update it to perform automatic decompression. Note, if you aren't sure where the fonts are, type (as root) find / -name "*.pcf*". To make sure libfbui knows where they are, you can use the PCFFONTDIR environment variable (as in export PCFFONTDIR=/path...).
      4. Once you've done these things, just compile the sample programs in /usr/src/linux-2.6.9/libfbui and run them from there. You may find it helps to run a program in a different virtual console using the -c switch.

See how easy? I am a programmer and that's... well... yeah.

Re:To clear somethings up

[Danzigism]

and OS X is doing what BSD has done since 1977..

Re:To clear somethings up

[freddy_dreddy]

So after a decade of mocking BSOD's on windows it's now a sought after feature in Linux? Sorry, don't get it.

Re:To clear somethings up

[GeorgeS]

Good god man...How long has it been since you used Linux? multiple monitors is a pretty simple thing to setup in most Desktop oriented Distros. Even I can do it :)

Re:To clear somethings up

[Hal_Porter]

clean the dust from the fan/heatsink, or re-appy the thermal paste on the GPU.

Yeah, try it.

I had a laptop with Intel integrated graphics that did it. I always thought it was a driver bug until I took the laptop to bits to resolder a broken power connector. The odd thing is that after that the graphics hanging stopped, i.e. it worked. Not sure if it was a loose connector or a badly seated heatsink.

Re:To clear somethings up

[Mad Merlin]

Yes, all of the current X drivers use DRI/DRM, except for the proprietary Nvidia drivers. The proprietary Nvidia drivers apparently have their own memory manager implementation and do not use the DRI/DRM infrastructure, likely because large swaths of the code are shared among all platforms Nvidia supports (Linux, FreeBSD, Solaris and the other thing).

Re:To clear somethings up

[Hal_Porter]

Plus 640x480 at 16 colors is actually part of the original VGA specification from the late 1980s. A card would have to be badly designed indeed (or terribly old) to not support that.

Actually Windows NT based OSs don't need a VGA compatible card. The driver sets a flag in the registry. Risc machines didn't have VGA compatible cards for example. I worked on a medical imaging system based on Windows 2000 that had non VGA compatible secondary displays, they were simple 256 level grayscale bitmap displays.

And before you say "ZOMG Windows Medical Imaging. BSOD", one of the testcases was having the PC OS crash completely. The images would still be displayed and the buttons to control them still worked because that was handled by an i960 running a vxWorks application that was independent of the PC. The i960s pinged the PC driver and if it didn't respond they blanked the PC graphics overlay until it was sane again.

Mind you most cards you buy are Vga compatible and set the registry flag. There's a plan to replace Vga with an EFI graphics standard with a higher minimum supported resolution. Like Vga Windows will switch to that whilst booting, during a crash or after disabling a buggy driver.

Re:To clear somethings up

[Bronster]

Wasn't that their point? That it's a separate system, so that if it fails, you'll still at least have the command line?

Yeah, nice theory - except what actually happens is you have two different drivers both talking to the same hardware, and not talking to each other. The usual failure mode when it really does fail is to confuse things so much that the console driver doesn't work either.

Having X talk to the low level card management drive will mean that when X screws up, the console driver still knows what's going on and can keep driving the card.

Much better.

Re:To clear somethings up

[Nutria]

Sorry, don't get it.

Threw me for a loop, too, at first.

I think the difference in that BSODs (used to?) happen quite frequently in Windows, and that, not the blue panic screen, was/is the real issue. The BSOD was just a visible, and therefore easy to criticize, manifestation of Windows' suckiness.

OTOH, bug-related (as opposed to those caused by failing hardware) kernel panics are pretty infrequent in Linux.

Re:To clear somethings up

[Nutria]

Yeah, nice theory - except what actually happens is you have two different drivers both talking to the same hardware, and not talking to each other. The usual failure mode when it really does fail is to confuse things so much that the console driver doesn't work either.

But a reboot into the text-only console would fix that, which wouldn't happen if Plymouth loaded kms during boot.

Re:To clear somethings up

[Nutria]

They also can be disabled. Nobody's forcing you to use them.

How, if it's initiated early in the boot process?

Also, I _like_ the smoothness of console switching which KMS gives.

Smooth console switching is totally useless to me, probably because I don't use a dm.

Re:To clear somethings up

[Nutria]

if Keith Packard thinks it's a good idea, I'm inclined to trust him

argumentum ad verecundiam, a logical fallacy.

You will still be able to boot headless with a serial line and a dumb terminal, if that's what you want.

That's just a silly argument.

You will still be able to boot into text and run startx.

Unless the distro boot process kms and dri really early in the boot process.

Re:To clear somethings up

[Hal_Porter]

They also can be disabled. Nobody's forcing you to use them.

Not true. _I_ am forcing him to use them

Re:To clear somethings up

[Anonymous Coward]

Still funny how it is GEM and DRI again. Before we had the Graphics Execution Manager and Direct Rendering Interface there was a time with Graphical Environment Manager and Digital Research Inc.

Re:To clear somethings up

[freddy_dreddy]

I think you're right - that it's the frequency of the event which throws sand in my face. Still, don't get why this is worth the attention.

Re:To clear somethings up

[JohnFluxx]

For those wanting a better reason: instant user switching.

At the moment if you press ctrl+alt+f1 from X it switches to a console, but takes about a second to do so. This is because the video card has to be torn down and then setup again. Likewise if you have two X sessions running, as different users, switching between them is slow.

This will make the switching almost instant.

Re:To clear somethings up

[Bert64]

Most disturbing is the number of windows users, even people working in the it industry, who just continue using the generic (and extremely slow) drivers rather than installing proper ones...
You don't even get a warning telling you that the proper drivers aren't installed.

Re:To clear somethings up

[Bert64]

Windows is much further away than linux, which is also far enough....

At the end of the day, both systems are far too complex for average joe, as are most devices people use... How many people know how to fix their cars or their tv etc?

People really need simple appliances that don't break, and do a small number or predefined tasks very well, games consoles are a good example of this, having one device that does everything results in a jack of all trades kind of device, it does everything poorly and nothing well, and it's too complicated for most people.

Re:To clear somethings up

[Bert64]

Or you can use a LiveCD to fix your linux install...
While it's rare to require the command line on windows, this is primarily because the cli isn't powerful enough to make it viable for most tasks.
Most tasks in linux are possible without the CLI too, but when you ask an expert for advice they will often tell you the CLI way because it's usually easier to explain... Trying to baby someone through a gui where they might have changed the color scheme or moved things around is very difficult, telling someone "type this" or giving them some text to copy/paste is much easier, assuming they can read.

It's not uncommon to require registry editing to fix windows problems, a task which is more arcane and complicated than anything on linux... The CLI on linux has man pages and help flags for individual commands, and text files you might need to edit usually have helpful comments... What does the registry have? a bunch of arbitrary text strings and undocumented numbers?

Re:To clear somethings up

[TeXMaster]

Thank you for elucidating my intent.

Re:To clear somethings up

[Elektroschock]

Yes, but still the USB that worked previouisly fails...

Re:To clear somethings up

[Elektroschock]

Well, you really don't need to touch the kernel and can use it precompiled. With new Desktop Environments as LXDE there is a fair chance that the complexity mistakes made with KDE and Gnome will not be made again. Ideally your hardware provider gets your hardware run with the kernel.

Re:To clear somethings up

[Elektroschock]

So the solution would be to standardize SVGA?

Re:To clear somethings up

[TheRaven64]

Please stop talking. You have posted a huge amount of nonsense in this thread already, and every subsequent post only makes you seem stupid. Ignorance is excusable, since it is correctable. Willful ignorance - remaining ignorant in the presence of people trying to educate you - is not.

A kernel already contains code for setting the mode of the graphics card. The X server also contains this code. There are two instances of the same code, maintained by different groups, doing the same thing, often at almost the same time. This is not, and never was, a good solution. It exists because XFree86 was used on a wide number of architectures and operating systems, and this was the simplest way of initialising VGA-like adaptors. Now, they have defined a set of standard interfaces for setting the mode of the display.

The result is that there is now simpler code in both the kernel and the X server. This will lead to fewer bugs, not more, since the result is less code, being used by more people. You are advocating returning to a state where every device driver had significant portions of the code duplicated in the X server and the kernel. This lead to code duplication, lots of complex and buggy work-arounds because neither the kernel nor the X server knew the state of the hardware at times when it needed to, and complex kernel-userland interfaces which amounted to passing on commands directly to the device from what should be untrusted code.

Re:To clear somethings up

[Nutria]

I understand the benefits of kms, but I also see the downside consequences of the combination of

1. the ability to do graphical boot, and,
2. certain popular distros' love of graphical boot.

Re:To clear somethings up

[Rutulian]

the whole OS world is moving towards microkernels

Uh, says who? There isn't a single working mainstream microkernel anywhere. The only active project I know of that seriously attempts to use a microkernel is GNU/Hurd, and I haven't heard anything from them in quite a while (they were supposed to be using the L4 microkernel, but I guess that fizzled for some reason).

Linux isn't "pushing graphics to the kernel." It is doing exactly what it should have done a long time ago. It is making the kernel handle the job of allocating and managing hardware resources. Graphics drivers have long been left out of the kernel for a lot of different reasons, and that has been a problem for graphics development on Linux. This is just a start, but it is a good one. This is a memory manager for the GPU. Don't worry, OpenGL and Xorg aren't going to make it to the kernel anytime soon.

Re:To clear somethings up

[Rutulian]

How, if it's initiated early in the boot process?

There is a grub flag...I think it is "quiet splash" that these distros use. If you take it out, you get a plain console boot up. I do it with every install because I like to see what is going on when the system boots up.

Re:To clear somethings up

[Rutulian]

This has nothing to do with the ability to do graphical boot. That has always been possible. A nice consequence of this new code is that graphical boot becomes easier and more stable, but that's not the purpose. The purpose is to hand over low-level control of the graphics hardware to a single resource, the kernel. That is all.

If you don't want a framebuffer console, you won't be forced to have one (I already responded in another post how to get a regular boot console if your distro defaults to a graphical one). And you will still benefit from the new code because Xorg will use it and won't get borked up as easily. Do you remember the DOS days? Do you remember every software package writing it's own drivers, and endless tweaking of CONFIG.SYS to try to get certain programs to run w/o conflicting with each other, which would happen anyway and lock everything up? It was ridiculous. A kernel to manage low-level hardware events is the right way to do things.

Re:To clear somethings up

[UncleTogie]

Windows is a lot closer to it than Linux on most fronts.

Not even close.

I work at a PC shop. We have users try to reinstall Windows all the time. Y'know what they can't wrap their heads around? Drivers. If I had a buck for every time someone said "well, I reinstalled Windows, but now I can't get on the internet and my screen colors look funny!"

A PC that's hardware-spec'd to Linux? Drop the disk in. Install. Enjoy.

Regardless, why should Windows be expected to get there first anyway?

Maybe 'cause with each OS iteration Microsoft tells us that they're already there... even though they're lying through their teeth... See my point above.

I think that sums up the attitude of far too many Linux devs/supporters - content to be second best on the desktop and hope that this "Free" thing will take it the rest of the way. It really won't though.

"Free" has caught the attention of quite a few of our clients. Most of them had Joe Q. Relative build them a PC on the cheap, drop a pirated copy of Windows on there, and leave them to the mess they created. When we tell 'em their choice is to buy Windows, or get a functional equivilant for free, LOTS choose Linux. Heck, just OpenOffice itself [another fine FOSS project] has quite literally saved our clients thousands of dollars.

The trick is to let 'em know an alternative exists to begin with. Not everyone will jump at it, but the boulder's still at the top of the hill. Let it pick up some steam first.

Re:To clear somethings up

[UncleTogie]

For recent versions of Ubuntu, fire up Synaptic {or use apt-get, or any other preferred method} and get the package called Startup Manager.

Beats the hell out of tweaking GRUB entries.

Re:To clear somethings up

[freddy_dreddy]

Hey dude, I really really like the fact that you took the time to comment (elaborately) on my post, but I typed moving towards, not owning and selling. OSX is based on Mach, MS's Singularity microkernel (and to some extent MinWin) is also a move in that direction. That's pretty much the whole PC market - which I bet nobody agrees with me here. Moving GPU hardware management to the kernel is a step in the opposite direction.

Re:To clear somethings up

[cyber-vandal]

Because Windows is already supposed to be there perhaps - or did you not see all the Microsoft publicity pretending that it is? Even though I'm now on Xmas Day creating a BartPE disk for the next time I'm asked to rescue a shambles that used to be a Windows install.
People don't find WIndows easy - they just know how to do stuff through experience. Once something happens that's outside that experience they come to people like me to help them work it out.

Re:To clear somethings up

[Rutulian]

Well, saying MacOSX is based on Mach is like saying Linux is based on Minix. They are related, but that is as far as the similarity goes. The OSX kernel is very much a monolithic kernel, and I have seen no indication by Apple that they plan to transition toward a microkernel design.

Singularity is an interesting project, but it is still an experiment by Microsoft Research. There are no commercial products with it on the horizon, although there have been hints. I would put it in the same category as L4 and other similar microkernel projects. There might be a niche market in some embedded devices in the future, but that's hardly the "whole OS world."

The thing is, microkernels are being debated by some, but for most the debate is over. There are advantages to microkernels, but not significant enough to offset the cost in performance, not to mention the cost of building a new system from scratch. Look at something like grub2, which is just a bootloader, nevermind a kernel. They decided they needed to rewrite it from scratch in 2002, and they are still working on it. It is probably usable in most distros now, like Ubuntu, but nobody trusts it because it hasn't been proven yet. The original grub (and lilo and just about every bootloader) had thousands of tweaks and workarounds to deal with various hardware, BIOS, and OS quirks. A new bootloader may be simple to write in principle, but it takes a long time for something like that to mature.

So, that was a bit of a digression...anyway, I think the future is pretty clearly going to be in virtualization. Some fairly elaborate stuff is possible, more than simply running an OS on top of a hypervisor, and I think we will start to see it in the next few years as virtualization becomes less of a buzzword and more of a standard practice .

Re:To clear somethings up

[freddy_dreddy]

Well, saying MacOSX is based on Mach is like saying Linux is based on Minix. They are related, but that is as far as the similarity goes.

Well, you may want to mail the people at Apple and ask them to stop all this disinformation.

Putting grub2 as an example that microkernels will never work is something I don't get. What about Cosmos ? A long way ahead but I don't see any problems on their horizon. Sorry, I'm strongly opposed to these fat monolyths, but that's probably part of the education.

not to mention the cost of building a new system from scratch.

I completely agree on this front, but the reason is not that it's not worth it. The reason is that there is no formal working model. In 2011 Intel is releasing an 80-core and we're currently seeing a trend to push computational efforts to the GPU. Trust me: you're not going to pull of a monolythic OS that doesn't know the exact amount of cores there will be on the die - or how many GPUs there are on board - and that hardware is where we're going.

All major CPU Co.s are pushing for these kind of dies because it increases the yield on the wafer in a tremendous way. The first appearance was Cell, which had an SPU turned off, but this was to increase the yield and offer a uniform product, not because the SPU had a defect. The trend is going towards production of dies which have as many cores as the amount of defects allow, every die can then be sold - but there won't be a homogenous product (the amount of co-processors won't be 16,32,64 but 40-ish or 70-ish). Monolythic kernels are lost on that, too much traffic is wasted in getting everything in sync. Compare it to the internet nodes: decentralisation means flexibility.

Re:To clear somethings up

[Rutulian]

Well, you may want to mail the people at Apple and ask them to stop all this disinformation [apple.com].

If you read a marketing blurb, expect to be deceived. Try this instead. In a nutshell, OSX borrowed some ideas from Mach, but it isn't the Mach kernel. It isn't a microkernel, at least not by the strict definition. At best it is a hybrid kernel.

I didn't mean grub2 to be an example of why it couldn't work, just that it's a lot of work to make it work. If you rearchitect the Linux kernel as a microkernel, and then have to spend the next 10 years retweaking everything to get it to the same performance levels it is at today, have you really gained anything from doing all that work? Don't get me wrong. Some things need to be rewritten (grub did). But you have to measure the cost against the gain. I think it is pretty hard to argue against a well-established, and properly designed and maintained monolithic kernel, like Linux. It's not like Linux is plagued by horrible problems that can't be solved without redesigning it.

Trust me: you're not going to pull of a monolythic OS that doesn't know the exact amount of cores there will be on the die

I'm not sure what you mean by this. SMP and cluster computing (grid, cloud, whatever you want to call it) have been going on for years (decades, even) with monolithic kernels. I don't see why multi-core computing would be all that much more complicated. And like I said earlier, virtualization is where everything is going. Intel's 80 core processor will run a bare metal hypervisor and that's it. Everything else will be running in virtual machines, although it will likely be a lot more sophisticated than what we see today.

Re:To clear somethings up

[freddy_dreddy]

Try this [wikipedia.org] instead.

Right back at ya (History:Mac OS X is based on the Mach kernel) ... I'll have to do some reading on this, but the uni library is closed. You're on the money re marketing.

I'm not sure what you mean by this.

Bear with me, I'd rather discuss this in a bar than via text.
In a normal die more than 90% of the surface is lost on metal interconnects, this is for 1 core. Putting multiple cores next to each other means you have to compromise on core interconnecting. This is normally limited to neighbouring cores (although they're working towards system in package and the likes via optical interconnects). I have two objections with a comparison of multicores with SMP and cluster computing:
- in cluster computing, communication between CPUs slows the system down - so you avoid it, minimize the packets or reduce the amount of packets by stretching the amount of computation in one job per CPU or by applying smart caching that reduce sending results back and forth. So: to avoid idle waiting they normally make sure that the queue is stuffed with commands during the wait. But: clusters and SMPs have a backdoor: you can interconnect CPUs arbitrarily (like here: too much traffic? make an extra bridge or lane).
In multi-cores, this is not an option, it's like traffic intersections: you can't just build a bridge or tunnel everywhere, everything is planar.
Secondly, you're restricted with what a core can do: you can't just stuff the queue as deep as with SMP and clusters. So you get these extremely complex scheduling and reservation formalisms - not to mention context switching.
- secondly: the clusters I worked with are built case by case, you can scale & design hardware infrastructure according to the jobs you're planning in the next 2 years or so. In multicores you have the opposite problem: designing an OS which must handle each quantity of cores optimally for any given piece of code.
When I wrote doesn't know the exact amount of cores there will be on the die, this is the future: cpus are going to be sold with variable number of cores, the defect on a die will be isolated by disabling the core where it's located. This makes yield skyrocket per wafer and allows relaxation on defect density per wafer (cheaper wafers). So, an OS will have to figure out how many cores there are, where the holes are (disabled cores) and make a plan to manage this optimally. This is the kind of job a microkernel is meant to do: local arbitration and communication, per core. Like the internet: design it so it flows around defects. You're calling it a hypervisor, but why on earth would you design an OS which consists of both a micro-kernel-ish part (hypervisor) and an monolythic part (what you mention that runs on it) when you can do the same with a microkernel.

Regarding Linux: Linux isn't plagued by anything, but truth be told: its architecture hasn't changed a lot in 15 years. There will be a lot of opposition to a redesign and most of the anti-arguments will be irrational - but they're welcome to surprise me on that ;)

Sorry for flooding you with text - I actually hate explaining things by typing.

For me, it's something else

[bogaboga]

When it comes to Linux, for me it's the other stuff the Linux does not do very well right now.

Let's agree: "Linux" as implemented by the many distros right now is ugly out of the box! Compare that with Apple's OSX or even Windows XP out of the box. With Linux, you first have to look for those Microsoft web fonts before you call a potential convert to have a look! Sad indeed.

Multimedia handling is still wanting on Linux. To make matters worse, even Linux advocates will prefer to create video files on Adobe's [proprietary] flash instead of .ogg! This makes you wonder which master Linux fan-boys serve. Heck, we can't even eat our own food?

One positive thing for now: KDE 4.2 is very very promising when to take a spin of it. Great work is being done as I write this. Gnome on the other hand will get there but the pain will be quite a lot before it does.

Re:For me, it's something else

[Ritz_Just_Ritz]

Some of the incarnations might be ugly, but the building blocks for "making it pretty" are there. A big Achilles heel is the graphics situation. Even with REALLY fast video cards, my Linux desktop seems to lag Windows in terms of having fast graphics. Without that, it's hard to ratchet up the eye candy for people that care about that sort of thing (and I'm not really one of those).

Nice to get ext4 into "stable" though.

Cheers,

Re:For me, it's something else

[Korin43]

I suppose "ugly out of the box" is a matter of opinion. I tend to like how Fedora looks out of the box.

Re:For me, it's something else

[Repossessed]

With Linux, you first have to look for those Microsoft web fonts...

Er... what? I've never done this before.

Re:For me, it's something else

[Just Some Guy]

Let's agree: "Linux" as implemented by the many distros right now is ugly out of the box!

I've actually never seen Linux, except for a few messages during boot. What's the shape of the invisible system that interfaces with your hardware?

Hint: You're talking about distros. We're talking about a kernel.

Re:For me, it's something else

[Aladrin]

"Multimedia handling is still wanting on Linux. To make matters worse, even Linux advocates will prefer to create video files on Adobe's [proprietary] flash instead of .ogg! This makes you wonder which master Linux fan-boys serve. Heck, we can't even eat our own food?"

We'll just assume you're right about 'Linux advocates', even though I seriously doubt most (let alone all) of them prefer flash.

What does it say about OGG that flash is preferred over it? It says it's not good enough. Don't expect people to use something that's not good enough just because it's 'free'. Being right is a lot more important than being 'free'.

Re:For me, it's something else

[Kjella]

Multimedia handling is still wanting on Linux. To make matters worse, even Linux advocates will prefer to create video files on Adobe's [proprietary] flash instead of .ogg! This makes you wonder which master Linux fan-boys serve. Heck, we can't even eat our own food?

Playing flash videos as in downloaded videos with a .flv extension is no problem. In fact, I don't remember last I had trouble playing any codec in a normal container format like avi, mpg, mkv, mp4, mp3, aac and so on even though Blu-Ray/HDDVD playback still needs work. The problem is flash, the universal crap plugin. If all the video sites could start using a x-flashvideo mimetype for that and leave x-flash for flash games and other ugly stuff that needs the real flash, half the issue would be solved. Of course the HTML5 video tag would be even better, but...

Re:For me, it's something else

[The Master Control P]

I recently got KDE 4.2 installed and took it for a test drive. First, something about kernel 2.6.18 that I'd been using resulted in system crashes - not even Magic SysRq keys worked. Putting in an overdue new kernel fixed that and I took it for a test-run that ultimately lasted 2 days. My system isn't particularly great - 2.4Ghz Athlon64, 1GB DDR533, nVidia 7300gt, generic PATA hard drive - but KDE 4 ran like a champ. I got full and perfectly smooth compositing/transparency.

It is as you say very pretty. Endless configurability of effects/compositing/multidesktop/etc via right-clicking title bar and hitting "window behavior." However, I had to flip between 2 or 3 menus to setup my desktop switches! One to enable switching when the mouse is pressed against a side, a second to enable the plasmoid (grid/sphere/cylinder/etc) for the effect I selected, and a third to select the side I wanted to press against. That is a Bad Thing, but fixing it comes down to rearranging some menus. With compositing enabled, I of course had great fun playing with transparency and wiggly-windows and such.

I liked the "draw on desktop" widget greatly. Now, put a color picker/small pallet and line-width picker in the corner and you've got yourself a sale! This is good as both a silly toy and to let me jot down or highlight stuff for other users. Add a selectable "poweruser menu" (and name it that) that does some more sophisticated stuff and you've got a competitor for the electronic-whiteboard thing they used in some of my classes. Oh, and try to work on it's frame/capture rate. It's one of those things that's got to run smoothly.

The new colors for Konsole were nice, a bit more sedate and less glaringly saturated than 3.5's Konsole default. Kwrite was beautiful, loved the new syntax highlighting. Konqueror... Loved it to death. Canvas element support was very nice (yay, I can play most javascript games now), and at last my preferred browser has the "last session unexpectedly exited; restore?" dialog. Konqueror as a file browser - disappointing. The default is a nauseating exercise in "How much space can we waste with giant icons and whitespace." Seriously, monitors don't get too much past 1920x1200 before you go dual-panel - KDE 4 fullscreen showed maybe 120 icons whereas I just selected 250 in Konqueror. Personal opinion, I know; It was a fairly short exercise to kill the whitespace and shrink back to smallish icons.

Configuration and customizability were great, typically so of KDE. We need a tutorial here; I don't have any problem navigating through about 30 tabs to set things up my way, but new users would be either overwhelmed or never know how much their desktop could do. Even a short popup explaining what Akonadi is and why I just spent the time to let it onto my MySql installation would be nice. And why did it have a problem creating its personal database after I gave it privileges to do so? Anyway, there's so much KDE 4 can do - most people will need a guided tour.

The bad... Okay, it's simply not acceptable for my keyboard to stop working about once a day and force me to exit to console and startx again. I don't care why, not acceptable - fix it, now, or this isn't going anywhere. This is ultimately what made me end my test-drive and decide not to move my emails/etc over yet. I did however discover that KDE4's desktop restore is fantastically faster than 3.5's - kudos.

There really needs to be a "taskbar always on top" option. I spent some time looking and if it's there it's hidden so it might as well not be there. Every user of Windows and previous versions of KDE and Gnome will expect the "max size" button to cover everything but the taskbar and for good reason!

Oh yeah, scroll wheel can change window opacity - good. Letting it scroll a window down to 0% opacity and disappearing - bad, no matter how good it would be for pranks.

Ultimately, I miss my "spiffy new desktop" but it's got a few glitches (fix the no-keyboard thing and I'm back) that are enough for me to hesitate. Final verdict: Tried 4.2 beta 1, waiting for 4.2.1 so I can get my asciifish screensaver back!

Re:For me, it's something else

[poopdeville]

However, I had to flip between 2 or 3 menus to setup my desktop switches!

And this is why I chose XMonad over Gnome or KDE at work. xmonad.hs FTW.

Re:For me, it's something else

[Just Some Guy]

Flamebait? No way. I've heard a global estimate of 30,000,000 desktop Linux users versus several billion "hidden" Linux devices, from servers routers to televisions to cell phones. This story is about the 99% of Linux computers that don't have a desktop interface and where such a thing wouldn't even make sense. That's why I said the OP was off-topic - we're not discussing anything remotely related to a GUI or other user interface.

Re:For me, it's something else

[pizpot]

KDE is cutting-edge.

Re:For me, it's something else

[slash.duncan]

Umm... I enjoyed the review, but... it wasn't KDE 4.2, as KDE 4.2 isn't even out yet, not even in rc, altho 4.2 beta 2 is (on Dec 18th, according to kde.org). So a bit of your "bad" should be fixed for the real 4.2, and not have to wait for 4.2.1. As one example, you mentioned the lack of an always-on-top taskbar option. From what I've read that's still 4.2.

I've not tried the 4.2 prereleases (in any form) yet, and 4.1.x isn't yet ready for this power-user, tho it was better than 4.0. Missing configurable hotkey support in 4.1 makes it a no-go for me, as I have the custom keys on my Logitech all setup in 3.5 and only use the kmenu for those apps used so seldom it's better to browse the menu for them. There are other problems... dual monitor support and panel issues among others, that make 4.1 a pain. But 4.2 is supposed to fix most of that and from your report and others it seems it's on its way to do just that.

Of course, KDE screwed the numbering and announcements up early in the 4.x cycle, but while unfortunate, that's history now and can't be changed, and 4.2 should, by the accounts I've read, be finally worth a normal .0 moniker, good enough for those like myself that tend to run leading/bleeding edge of a lot of stuff including their distribution, tho those that prefer to run stable may prefer to wait for 4.2.1 at least, and possibly 4.3.1, in no small measure because major KDE add-ons like amarok and k3b are only now looking about like 4.0 did for core -- IOW wait a couple more versions or be prepared to give up functionality from the 3.x versions.

Re:For me, it's something else

[lysergic.acid]

true. i don't think any of the mainstream distros are truly "ugly" out of the box, but you have to admit, they don't quite have the same level of polish as OS X or Vista (or even XP in some cases).

i think part of the problem is that FOSS has primarily been a software movement. so while there are a lot of world class programmers contributing to the FOSS movement, there's far less aesthetic/design expertise to go around in the FOSS community.

it's well known that both Apple and Microsoft spend huge amounts of money polishing the look and feel of their operating systems--for instance, recruiting visual interface designers with masters degrees in interactive media design, conducting focus groups to fine-tune the aesthetics and user experience (down to the tone of the language used in system messages), defining an extensive set of human interface guidelines for developers, contracting interactive design firms to create system icons, and even hiring world-famous musicians/composers like Brian Eno to create system sounds, etc.

this kind of attention to detail towards aesthetics just doesn't exist within the Linux development community because the lack of contributors with professional visual arts/design/music backgrounds.

Re:For me, it's something else

[Tubal-Cain]

With Linux, you first have to look for those Microsoft web fonts before you call a potential convert to have a look!

That's never been my experiance (I stopped installing MS TrueType fonts when I realized they didn't do anything for me), but there is <whisper>Red Hat's Liberation fonts</whisper>

Re:For me, it's something else

[Antique Geekmeister]

Much of that is a licensing issue. There is no available Linux commercial license for MP3 players, for example, that I've been able to find, and there are _especially_ no available licensed DVD players. There are certainly useful download sites such as the Penguin Liberation Front for getting the tools, but the creators of those technologies refuse to license them for Linux, so major distributions _cannot_ include them in their basic distribution or they face very dangerous lawsuits.

Look up the history of the libdvdcss library for more details of this kind of licensing craziness that directly interferes with the consumer success of Linux.

Re:For me, it's something else

[JamesTRexx]

I think you mean Kutting-edge. :-)

Re:For me, it's something else

[Rutulian]

Actually there is the Fluendo MP3 codec which is licensed and legal for use and free distribution. For DVD, there are legal players like LinDVD and PowerDVD, but they aren't free. I didn't realize this until I did a search just now, but apparently you can buy PowerDVD from the Canonical store (for Ubuntu), and other distros probably have something similar.

Re:For me, it's something else

[Antique Geekmeister]

I thought there were major patent issues with MP3 tools, which may be software copyright available but run into major patent issues in the US. (Wikipedia has some references for this on the MP3 page.) It looks like Fluendo is dealing with this as well as they can legally, which is a _good_ thing.

PowerDVD is providing players for Linux? Also good! When I looked for such tools a few years ago, I could find nothing available legally. Note that most distributions are understandably unwilling to pay even a few dollars to include such components in their base package, and I can understand this. I loathe software patents and consider the DVD encoding and encryption nuttiness to be extremely abusive by the manufacturers. But while I may want to commit civil disobedience, I don't want to do so in my workplace or help my employers do so without their knowledge, so I'm glad to have legal such tools available.

It's Christmas!

[Chris_Jefferson]

It's Christmas! Be sure to go to bed, get up, and spend the day with friends, family and food. Do you really need to update your kernel today? Why not let other people find out if there are some terrible early bugs in it?

Re:It's Christmas!

[Hal_Porter]

Santa provided this kernel. If you install it and if kills your box you must have been naughty, if it works well you must have been nice.

Re:It's Christmas!

[ghostunit]

I have no friends, family nor someone to cook for me you insensitive clod!

Re:It's Christmas!

[Hal_Porter]

I hear Cheetos do a special "Turkey and Cranberry" flavour at this time of year.

Re:It's Christmas!

[shutdown -p now]

It's Christmas! Be sure to go to bed, get up, and spend the day with friends, family and food.

What do I do if my family (i.e. wife) and frieds are all compiling the new kernel already, and the only food in the vicinity is a half-empty pizza box on top of my PC?

Further enlightenment into 2.6.28

[sega01]

If you haven't been following every commit's short log, you may find http://kernelnewbies.org/Linux_2_6_28 useful. I for one, would like 2.6.28 for Christmas.

Oh oh GEM is copyrighted

[Orion Blastar]

It is also another open source project OpenGEM based on the original DRI GEM. GEM was a Windows like 16 bit interface for DR-DOS and MS-DOS like Windows 3.X was. Apple sued them and they had to change their look a feel, and Atari used GEM as a GUI for TOS.

Re:Oh oh GEM is copyrighted

[John Hasler]

I yhink that you are trying to say that "GEM" is a registered trademark. One cannot "copyright" a three-letter string.

Re:Oh oh GEM is copyrighted

[Hal_Porter]

It is a shame since GEM and its ilk were a hell of a lot more efficient than X Windows.

GEM for Linux? Darn

[OrangeTide]

I was hoping someone was bringing back GEM from Atari ST and IBM PC to Linux. Oh well

Re:GEM for Linux? Darn

[Orion Blastar]

Check out this OpenGEM project maybe you can run it under DosBox or some other emulator?

Re:GEM for Linux? Darn

[OrangeTide]

Yea I'm aware of FreeGEM/OpenGEM. But porting it to Linux was non-trivial so I gave up. Running it under qemu was not so great, how can I use it to manage my ext4 partition that way? DosBox is a thought, although there is a lot of weirdness in DosBox for filesystems over some smallish size. Depending on what API the applications use.

Barely on v.2.6.28? Sheesh, Windows way past that!

[Anonymous Coward]

Heck Bill Gates has his OS way past Windows version 2000! The Linux will never catch up.

Re:2009

[windsurfer619]

This is really getting old. How do these guys still get modded funny?

Re:demographics

[Anonymous Coward]

Fucking religious fanatic.

What about linking to the changelog?

[diegocgteleline.es]

The changelog is available aswell...you might aswell have waited a bit to the final release!

I know these changes are awesome

[moniker127]

But why does it seem like every new kernel update has all these awesome new features that will change everything, but when you ask how they will affect people who USE the operating system, people are like "what do you mean? who cares?"?

I'm not an idiot, or someone who never uses computers, but what does the user of linux GET from these thousands of man hours?

Re:I know these changes are awesome

[Paradigm_Complex]

You're correct that the vast majority of improvements in the Linux kernel - when taken by themselves - are unlikely to change anything for any specific end user. These become significant when you add them all together. Odds are slim that any one person will ever use some new hardware support being added in a given kernel update, or some notice some change that ups battery life by couple a percent. However, when you compare the hardware support or battery life of a modern Linux distro to one even a few years old the change is drastic.

There is a huge number of examples I could give, but a recent event really stands out for me. Just a couple days ago a friend was having computer problems (couldn't read a DVD) and wasn't sure if it was a hardware or software issue. A simple check was to boot off Linux off of a USB flash drive and see if it worked (it didn't - ends up the DVD was funky). What's amazing here is that on a completely random system - built as a Windows gaming machine without Linux in mind - a Linux install which has never seen this hardware before performed flawlessly. It booted off of the USB drive faster than the (clean, relatively minimal bloat) XP did from the hard drive, detected and automatically connected to wifi, et al. Everything just worked.

Adding support for a few new webcams or wifi adapters or some new memory management or power stuff isn't going to make a difference. Doing that repeatedly for years, however, and all of a sudden you've got the best hardware support (out of the box anyways) and best performing OS around.

Re:I know these changes are awesome

[Spy der Mann]

You're correct that the vast majority of improvements in the Linux kernel - when taken by themselves - are unlikely to change anything for any specific end user. These become significant when you add them all together.

You're right. The problem is that there are still many hurdles for end-users to adopt GNU/Linux.

As an example:
* Easy virtualization with DirectX support. Kinda fixed with Virtualbox 2.1 (supports experimental OpenGL, but still no cigar)
* Ugly fonts. Fixed whenever all distros install the Liberation fonts by default.
* I want to receive and send animated MSN smileys! (Don' know about AMSN, but it kept crashing on me the last time - and pidgin doesn't support saving animated smileys yet)
* Games, games and more games. (When will Wine support 100% Windows apps?)
* Support for all video card vendors. Sigh.
* Support for all network card vendors. Sigh.
* Support for all webcam vendors. Sigh.
* Support for all other hardware vendors. Sigh.
* Partitioning setup problems, unexplained hangs while booting/shutting down, WTF? (Blame the distros)
* And while some particular kernel/software feature is already available, it usually takes a few months before it's available on the user's distro's repository. And trust me, the user does NOT want to get into command line to compile the latest version of his favorite software. I simply got tired of hunting for dependencies. Even as a GNU/Linux supporter, I've had enough.

I believe that there'll come a time when all the above features become "good enough" that a great majority of users will be satisfied with that. Then GNU/Linux adoption will reach the critical mass we're all expecting.

Re:I know these changes are awesome

[John Jamieson]

* Games, games and more games. (When will Wine support 100% Windows apps?)

When will any verson of Windows support 100% of windows apps?

Re:2009

[Shetan]

The moderators are drunk on Christmas spirits.

Linus just released it

[diegocgteleline.es]

http://lkml.org/lkml/2008/12/24/105

It doesn't really matter what day it is, or what holiday (if any) you're
celebrating, because even if you sit at home, alone in your dank basement,
without any holidays or friends, I bring you a tiding of great cheer: you
can now download Linux-2.6.28, and compile it to your hearts content!

Listen to the cheerful grinding of your harddisk as you reboot into an
all-new kernel - and I'm sure that if your computer could smile, it would
have a big silly grin on its non-existent face. So as you sit there in
your basement, give your computer the holiday cheer too.

In fact, even _if_ you have friends or family, leave them to their endless
toil over that christmas ham or turkey, and during the night, when they're
asleep, you can give them that magical present of a newly updated
computer. When they wake up tomorrow morning, tell them how you saw Santa
crawl down the chimney with his USB stick in hand, updating the OS of all
good boys and girls.

Ho, ho, ho,

                Linus "almost Santa" Torvalds

Re:Linus just released it

[Anonymous Coward]

Oh, my... The man is insane.

Re:Linus just released it

[Anonymous Coward]

Insane in the mainframe!

Best Christmas Gift, in the Kernel way

[DiegoBravo]

A new and single sound stack (valid for the next 10 years); with the added promise of discontinuing (deleting from the main tree) all the others by 2010.

Re:Best Christmas Gift, in the Kernel way

[Anonymous Coward]

Ummm... looks like ALSA is what you want and will be around for the foreseeable future. If you're talking about pulse, it's a user-space application that allows a much richer set of controls over system sound (i.e. redirect it over the network, per-application control, etc).

OSS has also been removed AFAIK except with respect to maintaining an OSS-compatability layer overtop of ALSA for old non-ALSA drivers.

Re:Best Christmas Gift, in the Kernel way

[3.1415926535]

While we're at it, could we add a new unmaskable fatal SIGPULSEAUDIO that the kernel automatically delivers to pulseaudio whenever someone tries to run it?

I guess SIGKILL would also work.

Re:Best Christmas Gift, in the Kernel way

[walshy007]

A new and single sound stack (valid for the next 10 years); with the added promise of discontinuing (deleting from the main tree) all the others by 2010.

There is only one kernel sound stack, alsa. What your probably referring to is userland audio bits, pulseaudio, jack, etc. Unfortunately each one has different goals and different typical uses, typical users often use pulseaudio but jack is the only real choice for musicians and people who do atypical things with audio.

a one solution fits all thing really would not fit all, it would inflict horrible pain upon those who have different usage habits than the average joe.

But still, this is not a kernel issue, the kernel has alsa, but a userland one

Re:2009

[_Sprocket_]

It's because every year is the year of Linux. Its just funny that some people haven't realized it yet.

Re:These Intel guys

[SHaFT7]

exchange sync the first time over wifi, had to cable up for that.

ext4 is stable enough -- for others

[Alain Williams]

"The ext4 filesystem, the successor to the ext3 filesystem, has been marked stable enough for people to start using and relying on,"

Forgive me for being a cynic -- I am going to wait until others have really tested & debugged ext4 before I trust it with my own data.

Re:ext4 is stable enough -- for others

[Anonymous Coward]

How's Sarge running for you Mr. Luddite?

Re:2009

[Anonymous Coward]

The moderators are drunk on Christmas spirits.

Which is only proven by the fact that they modded you insightful for that very comment.

Not sure about GEM

[Eravnrekaree]

Im not so sure about putting graphics stuff in the kernel? Why? Why not make it a part of X and thus platform independant. Now we will have a class of drivers locked to linux. Great, just what we need, incompatabilities.

Re:Not sure about GEM

[Creepy Crawler]

Akin to that idea: so you think that regular memory handling should be done by the shell? That is the analogue to X handling graphic memory.

Re:Not sure about GEM

[MostAwesomeDude]

Graphics stuff must be in the kernel at some level. The reason for GEM is that the entire system needs to have unified memory management for GPUs, just like for CPUs.

Also each GEM-capable driver has to support legacy mode. Linus was *very* clear on that point. So, starting with 2.6.29, each KMS or GEM driver supports non-KMS and non-GEM mode. (Some drivers, like the Radeon drivers, are all-or-nothing, so running KMS without GEM won't work.)

You're probably a BSD guy. Which is fine. Nothing wrong with that. Unfortunately, your upstreams have shown a rather lackluster interest in actually participating in these DRM changes. While there are a few guys working on porting this stuff, most of us are not BSD guys and are certainly not required to make it work across kernels. We're trying to make it as open and clean as possible, though. (DRM is actually built from a shared core that has Linux and BSD wrappers.)

And really, you don't want drivers in X. That's what we've done for a long time, and frankly, it sucks. Poor memory management, poor direct rendering. Lock contention, kernel sareas, GETPARAM/SETPARAM insanity. Each new feature requires kernel modifications and new ioctls which then have to remain working for a decade despite Mesa being the only real consumer of those ioctls. (nVidia doesn't use our DRM. They got this stuff working a long time ago on their own code. That's right, the closed-source drivers do this.)

Sorry for ranting, but that's the way it is.

Re:Not sure about GEM

[mortonda]

Im not so sure about putting graphics stuff in the kernel?

Because the whole job of the kernel, aka the Operating System, is to interface between hardware and user land, and ideally to do so in the most efficient manner useful to the user space programs.

Re:demographics

[PCeye]

Writing to the Slashdot community on Christmas eve that you're not thinking about the next kernel notes... sounds like you need a 12 step program for Linux or Slashdot.

couldn't disagree more

[patiodragon]

"Let's agree: "Linux" as implemented by the many distros right now is ugly out of the box! Compare that with Apple's OSX or even Windows XP out of the box. With Linux, you first have to look for those Microsoft web fonts before you call a potential convert to have a look! Sad indeed."

When I go to use one of the newer Start buttons in windows that is all "mapped out" instead of a nice, simple, list, I don't know where anything is. Very frustrating. And what about Macs? Because stuff gets BIGGER when I mouse over it, that makes it easier to understand!? I hate the way KDE looks, but I don't pretend that Gnome is any better. It is my opinion.

Your whole argument starts off like a statement of fact when it is really only an opinion.

Re:Barely on v.2.6.28? Sheesh, Windows way past th

[db32]

I hear the next one is 7. I think this illustrates why they should not be in the OS business, they can't even put simple integers in the right order let alone do complex calculations.

H Tree Indexing

[KermodeBear]

I looked over the Wikipedia article for Ext4, and it mentioned that Ext4 uses an H Tree for directory indexing. I looked over the H Tree article, but it is sparse, and I wasn't sure how it differs from a balanced B Tree. Could someone more mathematically inclined explain, or point me to some better information?

Thanks in Advance. (o:

Re:H Tree Indexing

[Nimey]

The "B" in B-tree stands for "balanced", fyi.

Re:H Tree Indexing

[sedman]

Actually, it stands for binary as in two branches. You can have an unbalanced b-tree.

Re:H Tree Indexing

[EvolutionsPeak]

No, a B-tree is not a binary tree. It is an m-ary tree where all the leaf nodes are at the same depth, so it is balanced.

Re:H Tree Indexing

[badkarmadayaccount]

No, a B-tree is not a binary tree. It is an m-ary tree where all the leaf nodes are at the same depth, so it is balanced.

You missed something.

Re:Heh

[siride]

Except that this isn't at all equivalent to that. That would be the equivalent of moving the X server into the kernel, not just some directly hardware facing parts of the drivers.

Re:Stable??

[siride]

There are stable branches: older kernel releases. They keep getting bugfixes and security fixes for some time.

An awful lot has happened this year.

[Zombie Ryushu]

Religious people tend to always find a religious meaning, let me be the first to give a secular reason to be thankful for this season. However slight and insignificant, Linux has gained ground this year. More hardware and applications work this year than last year, and once again, I am preparing my yearly X-mas LAN war like so many other years before. (Hey, I don't believe in religion/God, but I do find this time of year wonderful for partying. And thats exactly what I will do.)

I'm starting to re-evaluate my dire forecasts for Linux's future. In 2005, and 2006, I was fairly certain Linux users were in the edge of the Abyss. Linux may never become a huge player in the desktop market, but maybe just this once, the hundreds of thousands of rabid Windows and and to a lesser extent OSX users that are all hounding for the death and mutilation of the Linux OS and the charter it represents are held back for the time being.

Now, Linux user must not grow complacent or overconfident. But for this season, for this year and this next few days, we can all be a little less afraid than we are normally.

Personally I would have gone with option C...

[ipX]

But I'm nonetheless glad to be compiling 2.6.28 right now. From the Linux kernel mailing list (emphasis mine):

Perhaps more interesting is simply the release scheduling issue. I'm getting slowly ready to do a real 2.6.28, but I don't think anybody really wants the merge window to be around the holidays. So the question is really whether to (a) just make the -rc's go on a few more weeks, and do 2.6.28 after xmas I like this, because alledgely people are debugging things, and we'd get a more stable 2.6.28. or (b) release in a week or two, but just allow for possibly extending the merge window due to people being drunk on eggnog.. I like this because let's face it, we get more and better bug information after releases, and everything _should_ be ready for merging *before* the merge window anyway. or (c) some other crazy scheme that somebody comes up with in a drug-induced stupor. So I haven't quite decided on that thing yet, but I'm open to suggestions. Linus

Re:2009

[jalefkowit]

"The future is already here. It's just not evenly distributed." -- William Gibson

Re:2009

[Eighty7]

whoa

Glogg inspired?

[anothy]

Little penguins all around the world are waiting for Penguin-Master Linus Torvalds to deliver some Glogg inspired Xmas cheer in the form of the new 2.6.28 kernel.

wow. i've often looked at linux and thought the programmers must've been drunk while writing it. now we know.

Re:Glogg inspired?

[anothy]

sigh. xkcd is so amazing. i think it ought to be a new mandatory part of every slashdot story submission that you pick which xkcd comic goes with the story.

OT: 2.6.27/28 config files?

[twilight30]

Hi everyone,
Ever since 2.6.27.x came out I have not been able to compile from source and have the internet connection work correctly at all.

Basically I try to take old source configs and run them in the new kernels, but I get the same result.

Even binary Ubuntu kernel builds fail to run internet connections correctly...

Apparently this item may be related to it:
http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=fd6149d332973bafa50f03ddb0ea9513e67f4517

(regarding the reordering of TCP options... how do I fix it?)

Any advice very gratefully appreciated ...
M

Re:demographics

[Antique Geekmeister]

Not all Linux users are Christians, you know. I know several devoutly pagan Linux advocates, and quite a few Jewish ones.

So Linus

[VincenzoRomano]

should put a red and white dress as well as a white beard!
Oh oh oh oh!

Re:H Tree Indexing .. Will it speed ls & rm?

[Mandrel]

I hope this fixes the two annoyances I have with Linux:

1. Doing an ls on a directory containing 1000 ext3 files on my quite modern computer takes nearly 10 seconds.

2. Deleting a multi-gig file such as a TV recording locks up the OS so badly that other apps freeze. If mencoder is recording TV it will fail to keep up with the stream. AV sync is lost, ruining the rest of the recording.

Re:Heh

[TheRaven64]

To clarify:

DRI1 is a big lump of ring-3 code and a small chunk of ring-0 code. The ring-3 code issues commands, the ring-0 code validates these commands (to make sure that DMA is being safely used and so on) and passes them off to the device. With DRI1, you have some things, like setting the graphics mode, which are implemented in the kernel's (X-independent) driver and then implemented again in X.org. This leads to copy-and-paste errors and all sorts of other problems (like power saving, since the kernel driver now knows less than it needs to about the state of the GPU).

The ring-3 blob is also responsible for setting up memory mappings for DMA, which is bad for three reasons. Firstly, it means that every driver is implementing almost identical code. Secondly, it means each driver is implementing code that depends a lot on kernel interfaces, making porting harder than it should be. Finally, it is bad for security.

With DRI2, the kernel controls the IOMMU (if there is one) and sets up memory mappings so that the driver just mmap()s regions of device memory or physical memory. There are quite a lot of related improvements, although I'm not sure exactly which ones are part of the DRI2 'brand.'

The non-GL acceleration interface is cleaned up a lot, to make supporting compositing and spline rendering easier and the GL state tracker is moved out of the driver and into a separate bit of (ring-3) code. This simplifies the drivers even more, since their interfaces to the system is now stateless (DRI2 drivers are a fraction of the size of DRI1 drivers). This means that they all share a lot more code, which is good for stability (since it's now code that's tested by a lot more people) and good for cost (it's now cheaper to support X.org). The nice side effect of removing the state tracker is that it's possible to plug in new ones, for example OpenGL ES or DirectX. the WINE guys have been talking about porting their DirectX back end to issue DRI2 calls directly, rather than OpenGL, so that there is no overhead - DirectX and OpenGL are both first-class user APIs on top of the low-level API.

The DRI2 stuff is a good example of open source done right. The result will be that each of the concerned parties (kernels, driver writers, X server, GL state tracker) all have a simpler codebase to maintain as a result of using standardised interfaces to other code. This leaves nVidia out in the cold - they are stuck maintaining their own memory manager, mode setting code, GL state tracker, and so on, while their competitors aren't. Developing a Free Software GPU driver costs a small fraction of the cost of maintaining the nVidia driver, which to my mind is a far better way of advocating Free Software than bludgeoning people with reciprocal licenses (the DRI stuff in X.org is MIT licensed, only the Linux kernel part is GPL'd).

Re:I for one...

[TheRaven64]

2000? What operating systems / browsers didn't support UTF-8/16/32 in 2000? More like welcome to 1993.

Re:H Tree Indexing .. Will it speed ls & rm?

[JohnFluxx]

Um, there's something seriously wrong with your machine.

cd /usr/share/doc; time ls | wc -l

real 0m0.021s
user 0m0.008s
sys 0m0.004s

That folder contains just under 2000 files/directories. Repeating on /etc (with about 500 files) gives similar results.
Likewise deleting files shouldn't take so long

Re:H Tree Indexing .. Will it speed ls & rm?

[Mandrel]

That's interesting. I got a doc dir time that, while not quite as fast, was still under a second.

I don't think I've got something fundamentally wrong. This computer has a recent kernel, and 4GB of RAM.

One likely cause of the ls slowness I reported is the large average size of the files in that directory, which I think requires inode chains to be traversed. Will ext4 speed this?

The other possible cause is the software RAID1 I've got that ext3 fs (/dev/md0).

I'd be interested to see how long it takes for other systems to delete a 10-20G file, and its effect on other processes.

Better than having unmanaged access to hardware.

[spaceturtle]

Well, at the moment we don't really have the command line as a failsafe. When the X server crashes it seems to lock up the keyboard so Cntl-Alt-F1 doesn't switch to the VT (though I can usually ssh in, and restart X, but then I could also ssh -Y and restart X with a remote GUI, so the "commandline" doesn't really help here).

The problem is the currently we have three different things that can directly mess with the video hardware: The framebuffer, DRI and the X server, and so any of these can cause trouble. This can lead to worse than triple the number of bugs because interactions between these can cause trouble.

Its similar to how having a database server tends to be more reliable than having clients directly accessing the database files. Yes, the database server adds a single point of failure, but that is better than having 20+ nodes each of which can horribly corrupt the database files. While in principle GEM could fail, so could the hundred other modules in my kernel. And a bug in GEM is unlikely to be as serious as a bug in ext3/4.

Innovation?

[heffrey]

Whilst there are good and worthy new features here it's going a bit far to call it innovation. Incremental improvement is more like it, and that's no bad thing.

Re:H Tree Indexing .. Will it speed ls & rm?

[JohnFluxx]

I also have raid 1, but it doesn't take longer than a 10th of a second to list a few thousand files

Re:demographics

[PCeye]

"I can't help but wonder who'd care enough about this topic to be writing serious thoughtful comments on it on christmas eve!..."
"...the point is that even I am not spending christmas eve thinking about the next 2.6.28 release notes"

My comment is relative to the AC post above (as I replied to that posting).

"Not all Linux users are Christians, you know. I know several devoutly pagan Linux advocates, and quite a few Jewish ones" ...thanks, I guess, ... Not sure why you feel you must this out since I am referencing one comment.

Re:demographics

[Spy der Mann]

Not all Linux users are Christians, you know. I know several devoutly pagan Linux advocates, and quite a few Jewish ones.

What about those of us belonging to the Church of Emacs?

There's no system but GNU, and Linux is one of its kernels O:)

Comment removed

[account_deleted]

Comment removed based on user account deletion

wireless usb drivers

[Sark666]

Does that just mean the addition of new drivers or a revamp of the existing? I have some no name wifi usb that uses zd1211rw and it's pretty easy to make it fall over.

It works, but if I copy numerous small files it'll stall in quick order, but with one large file it's usually fine. Same with a movie, streaming it and just playing is fine, but skim through it too much and it'll drop the connection.

I use ubuntu intrepid, in the last version, not only would it drop, but a lot of times I could not bring it back to life without a reboot, even removing/reloading the module would result in errors.

So it's a little better, but it's still pretty unreliable.

Re:wireless usb drivers

[Paradigm_Complex]

Does that just mean the addition of new drivers or a revamp of the existing? I have some no name wifi usb that uses zd1211rw and it's pretty easy to make it fall over.

I'm not sure about this case, but in the past it has been both. My wireless usb stick (rt73) WAS supported in Ubuntu Dapper (and Edgy?) but after that it stopped working (required me to throw the firmware into /lib/firmware - once I do that it works fine).

From that experience I gather that yes, it is possible for them to revamp currently working drivers; however, it would probably be easier for you to just buy a supported card. May be a bit late to ask for it as a Christmas present, sadly.

Re:H Tree Indexing .. Will it speed ls & rm?

[BlackCreek]

I hope this fixes the two annoyances I have with Linux:

1. Doing an ls on a directory containing 1000 ext3 files on my quite modern computer takes nearly 10 seconds.

AFAIK Adding directory name indexing should solve that (other than if you have a laptop and need to spin the disk up before listing).

Heh

[amoralyrr]

that's cool =) I'll try to update tomorrow and convert my /home/ to ext4

Re:2009

[tiananmen tank man]

The future is now, old man!

Re:H Tree Indexing .. Will it speed ls & rm?

[Mandrel]

Is this for a first fetch or for a cached fetch? I certainly get very quick directory listings while they remain cached.

Re:H Tree Indexing .. Will it speed ls & rm?

[Mandrel]

I ran dumpe2fs /dev/md0 | grep features, and the dir_index option is enabled.

Re:H Tree Indexing .. Will it speed ls & rm?

[JohnFluxx]

First fetch

Re:H Tree Indexing .. Will it speed ls & rm?

[amorsen]

You probably want to turn directory indexing on. This is the default for newer versions of mkfs.ext3, but you have to switch it manually for older installations.

You could also switch to ext4, of course.

Re:H Tree Indexing .. Will it speed ls & rm?

[BlackCreek]

Do you have fancy ls options enabled ? Namely "--color"?

On my system (FWIW reiser3 with indexes):

echo 3 >! /proc/sys/vm/drop_caches

time \ls /usr/share/* > /dev/null
\ls /usr/share/* > /dev/null 0.06s user 0.01s system 3% cpu 2.224 total

echo 3 >! /proc/sys/vm/drop_caches

~ % time ls /usr/share/* > /dev/null
/bin/ls -h -N --color=auto -F -v /usr/share/* > /dev/null 0.02s user 0.36s system 2%cpu 16.000 total

Re:demographics

[Antique Geekmeister]

Linux is its only kernel. HURD was a false prophet. (I betatested HURD: it never worked.)

GEM was Copyright 1988 by Digital Research Inc

[bmullan]

Graphics Environment Manager (GEM) was the 1st PC (non-apple) Graphical Desktop but DRI lost again to Microsoft and Windows 1.0 despite a great GEM development environent that included a true multasking, multitreaded real-time scheduler in the DRI FLEX-OS GEM. Gary Kildall was a great guy and it was a shame he lost to Bill Gates in the early PC days. >>>GEM/3 Desktop Release 3.11 Copyright (C) 1988 - Digital Research, Inc. November 3, 1988