What method to install the Ati drver has used You ? What method to configure the Ati drver has used You? Sometimes, can coexist in xorg.conf the 2 drivers, the radeonhd and the fglrx. You has verfified that? I prefer to install by compilation executing as "su" "sh ati-driver-installer-9.2-x86.x86_64.run", and the initial configuration with the "aticonfig" tool. Regards 2009/4/10 David C. Rankin :

Listmates,

       There are 2 issues that need to be addressed regarding the fglrx driver and the upcoming 11.2 release. Currently in 11.1, there are many laptop cards supposedly "supported" by the ATI driver that crash or hardlock when the driver is installed. Currently my X1200 is one, the 200 is another from the list (I'm not sure what the technical card architecture is for the "200"). Suffice it to say, this is a significant problem for all affected users. The issues that need to be addressed pre-11.2 are:

(1)     Performance issues and crashes caused by the ATI driver. Recent driver problems: The 9-2 driver release caused my Toshiba 205d laptop to reboot on xdm start and the 9-3 driver caused my laptop to hardlock if glxgears was started. All fglrx drivers 8-10 through 8-12 had performance of less than 60% of the 8-9 driver. Compounding the problem is the fact that nothing prior to the 8-12 driver will even compile or run on 11.1 or later (presuming 11.2 as well)

       Thankfully, there are some cards that seem to do OK with the recent 9-3 driver on 11.1, but it is hit or miss ...and... the fact that it works on "some" does nothing to address and help the users where it "does not" work. Before the 11.2 release, Novell needs to get a handle on what cards will and will not run with the ATI driver and at least incorporate a warning or release note for those with affected hardware.

(2)     Confirm - ATI driver support will be dropped for all R300-R500 ATI cards post 9-3 driver release (per phoronix.com ATI forum). This is another *biggie* Novell needs to get its arms around, to again, either incorporate into the installer a way to not install the driver on non-supported hardware or identify and warn about the lack of support for all R300-R500 cards.

       That means no (nada) fglrx driver support for the R300-R500 series cards which include the: 9500, 9600, 9700, 9800, X300, X550, X600, X700, X800, X850 AllInWonder, and X1050 cards affecting all laptops bought before Q3 2005 (or January 2007 in the case of the X1050). All of which are still *very* prevalent in the group of laptops that make up the installed user base today and will continue to do so through 11.2 release. This issue of dropping support is separate and apart from issue (1) above because the group of cards in (1) are supposedly still supported.

                       Other Thoughts:

       After Stefan Dirsch handed over the reins of driver maintainer to Bob Walmsley I guess he is now the guy that this should be coordinated through. I am glad to try and help, and to run test cases, etc. since I drew the black bean in laptop selection so let me know if/when there are new cases to test. I follow the ATI driver releases and test each one so I have a good handle on where things stand today. The problem is that Novell needs to use its resources to get a working partner in ATI to address some of these issues. Luugi Marsan was in the past the ATI contact, but at present, I haven't a clue who the ATI contact is or even if there is one. Without a good contact, we are dead-in-the-water getting any of these issues addressed and we are guaranteed a significant number of stranded users if these issues are not addressed.

       Right now the 8-9 driver works fine for 10.3 and 11.0, but unfortunately it is incompatible with 11.1. The ATI issues prevent me, and all others affected, from being able to move to 11.1. Hopefully, if Novell will start now to address some of these driver issues with ATI, then by 11.2 release  they have a good chance of ironed out.

       With a desktop, you can always rip the ATI card out and replace it. With a laptop ... you are kind-of stuck.

-- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

-- USA LINUX OPENSUSE QUE ES SOFTWARE LIBRE, NO NECESITAS PIRATEAR NADA Y TAMPOCO TE VAS A PREOCUPAR MAS POR LOS VIRUS Y SPYWARES: http://www.opensuse.org/es/ -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

2009/4/10 Stefan Dirsch :

On Fri, Apr 10, 2009 at 02:37:59AM -0500, David C. Rankin wrote:

...

Listmates,

      There are 2 issues that need to be addressed regarding the fglrx driver and the upcoming 11.2 release. Currently in 11.1, there are many laptop cards supposedly "supported" by the ATI driver that crash or hardlock when the driver is installed. Currently my X1200 is one,

I doubt that 9-3 driver and before will still work on xorg-server 1.6. 9-4 and later won't support R500 any longer. You'll need to switch to radeonhd for 11.2 anyway.  BTW, radeonhd 1.2.5 now enables DRI on R500 by default.

...

the 200 is another from the list (I'm not sure what the technical card architecture is for the "200").

Don't know either.

You can view an unofficial table with details of the ATI video chips: http://en.wikipedia.org/wiki/Comparison_of_ATI_Graphics_Processing_Units The X200 and X1200 do'nt appear as "Mobility radeon". They appear 2 versions of the X200 onboard and it has the RS400 family chip: Radeon Xpress 200 IGP Mar 11, 2005 RS400 Radeon Xpress 200 IGP June? 2005 RS482 And the Radeon X1200, has not the R500 chip: Radeon X1200 IGP Feb 28, 2007 RS690C -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

2009/4/10 Stefan Dirsch :

On Fri, Apr 10, 2009 at 09:53:48AM -0300, Juan Erbes wrote:

...

2009/4/10 Stefan Dirsch :

...

On Fri, Apr 10, 2009 at 02:37:59AM -0500, David C. Rankin wrote:

...

Listmates,

      There are 2 issues that need to be addressed regarding the fglrx driver and the upcoming 11.2 release. Currently in 11.1, there are many laptop cards supposedly "supported" by the ATI driver that crash or hardlock when the driver is installed. Currently my X1200 is one,

I doubt that 9-3 driver and before will still work on xorg-server 1.6. 9-4 and later won't support R500 any longer. You'll need to switch to radeonhd for 11.2 anyway.  BTW, radeonhd 1.2.5 now enables DRI on R500 by default.

...

the 200 is another from the list (I'm not sure what the technical card architecture is for the "200").

Don't know either.

You can view an unofficial table with details of the ATI video chips:

http://en.wikipedia.org/wiki/Comparison_of_ATI_Graphics_Processing_Units

The X200 and X1200 do'nt appear as "Mobility radeon".

They appear 2 versions of the X200 onboard and it has the RS400 family chip:

Radeon Xpress 200 IGP         Mar 11, 2005    RS400

Radeon Xpress 200 IGP         June? 2005      RS482

RS4xx is supported by radeon driver, but currently broken on openSUSE 11.1. Bugreports are filed.

...

And the Radeon X1200, has not the R500 chip:

Radeon X1200 IGP      Feb 28, 2007    RS690C

supported by radeonhd driver including DRI. RS690 is a R500. Confusing I know.

Yes, confusing, but in the table, with the R500 chip, start the series with the number X1300: Radeon X1300 Oct 2005 RV515 http://en.wikipedia.org/wiki/Comparison_of_ATI_Graphics_Processing_Units#Rad... http://en.wikipedia.org/wiki/Radeon_R520 The Radeon onboard X1200, is newer than the X1300 series: Radeon X1200 IGP Feb 28, 2007 RS690C Regards -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

David C. Rankin wrote:

After Stefan Dirsch handed over the reins of driver maintainer to Bob Walmsley. I guess he is now the guy that this should be coordinated through. I am glad to try and help, and to run test cases, etc. since I drew the black bean in laptop selection so let me know if/when there are new cases to test. I follow the ATI driver releases and test each one so I have a good handle on where things stand today. The problem is that Novell needs to use its resources to get a working partner in ATI to address some of these issues. Luugi Marsan was in the past the ATI contact, but at present, I haven't a clue who the ATI contact is or even if there is one. Without a good contact, we are dead-in-the-water getting any of these issues addressed and we are guaranteed a significant number of stranded users if these issues are not addressed.

David, I can understand your frustration in this matter, as for my part, I currently have the ability to test the latest software on a Dell Studio 17 with an ATI Mobility Radeon HD 3650 in 64 & 32 bit mode. If there is a need for more varied (hardware) testing I am willing to invest in a separate Computer with assorted Video cards. I am always up for suggestions to make the process better and beneficial for everyone. Linux user 359911 openSUSE 11.1 (x86_64) Linux 2.6.27.21-0.1-default x86_64 - Dell Studio 1737 4.2.2 (KDE 4.2.2) "release 112" -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Stefan Dirsch wrote:

I don't see what the openSUSE community can do here. ATI is only interested in bugs, if they affect and being reported by big vendors like e.g. HP. But these big vendors are only interested in using SLES/SLED.

Hmm.. Stefan, I'm not sure I follow here. Isn't openSuSE the forerunner for what will end up in SLES/SLED? If so, it just seems to make more sense from a quality assurance perspective to address any issues at the beta (openSuSE) stage before they end up having to (react) to problems with SLES/SLED. The proactive/reactive issue. Moving the bug identification/resolution a bit further upstream if you will. The absolute crushing blow ATI is dealing its users by dropping support for all R300-R500 cards is simply amazing. Nvidia still provides support for all it cards with the legacy driver. If there is ever a contact at ATI it would be worth suggesting a similar "legacy driver". That just seems like common sense. But then that too is uncommon in todays world. (i.e. a global economy built on "credit default swaps" -- give me a break...) Hopefully we can find someone with a fglrx.com email address that will entertain some candid feedback and suggestions instead of the current "linux driver feedback" which from what I can tell == /dev/null -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Friday 17 April 2009 05:02:33 am Stefan Dirsch wrote:

Then you can play the proxy between the openSUSE community and ATI and will notice that ATI ignores the feedback of beta testers as well.

How many people are there that we can contact? One bird doesn't make spring. They can ignore that. -- Regards, Rajko -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

David C. Rankin wrote:

Listmates,

There are 2 issues that need to be addressed regarding the fglrx driver and the upcoming 11.2 release. Currently in 11.1, there are many laptop cards supposedly "supported" by the ATI driver that crash or hardlock when the driver is installed. Currently my X1200 is one, the 200 is another from the list (I'm not sure what the technical card architecture is for the "200"). Suffice it to say, this is a significant problem for all affected users. The issues that need to be addressed pre-11.2 are:

<snip>

(2) Confirm - ATI driver support will be dropped for all R300-R500 ATI cards post 9-3 driver release (per phoronix.com ATI forum). This is another *biggie* Novell needs to get its arms around, to again, either incorporate into the installer a way to not install the driver on non-supported hardware or identify and warn about the lack of support for all R300-R500 cards.

<snip> A glimmer of hope form the AMD/ATI web site: <quote> AMD has moved a number of DX9 ATI Radeon™ graphics accelerators products to a legacy driver support structure. This change impacts Windows XP, Windows Vista, and Linux distributions. AMD has moved to a legacy software support structure for these graphics accelerator products in an effort to better focus development resources on future products. The following products have been moved to the legacy software support structure (including Mobile and All-in-Wonder Variants): ATI Radeon 9500 Series ATI Radeon 9550 Series ATI Radeon 9600 Series ATI Radeon 9700 Series ATI Radeon 9800 Series ATI Radeon X300 Series ATI Radeon X550 Series ATI Radeon X600 Series ATI Radeon X700 Series ATI Radeon X800 Series ATI Radeon X850 Series ATI Radeon X1050 Series ATI Radeon X1300 Series ATI Radeon X1550 Series ATI Radeon X1600 Series ATI Radeon X1650 Series ATI Radeon X1800 Series ATI Radeon X1900 Series ATI Radeon Xpress Series ATI Radeon X1200 Series ATI Radeon X1250 Series ATI Radeon X2100 Series </quote> -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Hi, I only now saw this post... 've been working with hardware vendors for a while and their primary objective with drivers is cost reduction. A major cost for them is regression testing for new drivers. So if they can avoid testing they'll do that. So if they release the 8-9 driver, and that works well with a Toshiba 205d (and whichever chipset release in there), then they recommend to just use 8-9 with Toshiba 205d and to not update the driver, ever, unless being told. Only if they find a significant issue with 8-9 on toshiba 205d (and have committed to fix issues there to Toshiba), they'll retest that chipset with the current driver (for the sake of an example let's say 9-3). Then Toshiba 205d users (i.e. the chipset in there) are reccomended to update to 9-3. Why do they do that? If they don't update the driver for Toshiba 205d, they don't need to test on 205d. This approach is a given. We will not change them to start regression testing. That's just how it is. Now what we _can_ do is, we start building the different ati driver versions in parallel in an obs project: project: ati-fglrx packages: ati-fglrx-8.8 ati-fglrx-8.9 ati-fglrx-8.10 ati-fglrx-9.1 ati-fglrx-9.2 .. and then the user chooses The Right Package. Two approaches have been discussed in the Driver Backports and distribution LF workgroup http://www.linuxfoundation.org/en/Driver_Backport to make this user choice automatic, and to automate the rare case that ATI reccomends to update the driver on some systems. option #1 was to make the driver KMP exact in specifying the supported pci-ids. option #2 was to create a repository per system, a 'driver kit'. So you have a project Toshiba, with Subproject 205p. This project 'aggregates' the ati-fglrx-8.9 package. And if ATI would reccomend to upgrade 205p to, say, 9.3, then the aggregate would be changed to do this migration. You, like all other toshiba 205p users, have your machine connected to this project. The nice little tool "Jockey" from our friends at ubuntu could be used to make this repo addition automatic, e.g. based on the smbios id of the machine type. S. full quote of the old mail below... "David C. Rankin" writes:

Listmates,

There are 2 issues that need to be addressed regarding the fglrx driver and the upcoming 11.2 release. Currently in 11.1, there are many laptop cards supposedly "supported" by the ATI driver that crash or hardlock when the driver is installed. Currently my X1200 is one, the 200 is another from the list (I'm not sure what the technical card architecture is for the "200"). Suffice it to say, this is a significant problem for all affected users. The issues that need to be addressed pre-11.2 are:

(1) Performance issues and crashes caused by the ATI driver. Recent driver problems: The 9-2 driver release caused my Toshiba 205d laptop to reboot on xdm start and the 9-3 driver caused my laptop to hardlock if glxgears was started. All fglrx drivers 8-10 through 8-12 had performance of less than 60% of the 8-9 driver. Compounding the problem is the fact that nothing prior to the 8-12 driver will even compile or run on 11.1 or later (presuming 11.2 as well)

Thankfully, there are some cards that seem to do OK with the recent 9-3 driver on 11.1, but it is hit or miss ...and... the fact that it works on "some" does nothing to address and help the users where it "does not" work. Before the 11.2 release, Novell needs to get a handle on what cards will and will not run with the ATI driver and at least incorporate a warning or release note for those with affected hardware.

(2) Confirm - ATI driver support will be dropped for all R300-R500 ATI cards post 9-3 driver release (per phoronix.com ATI forum). This is another *biggie* Novell needs to get its arms around, to again, either incorporate into the installer a way to not install the driver on non-supported hardware or identify and warn about the lack of support for all R300-R500 cards.

That means no (nada) fglrx driver support for the R300-R500 series cards which include the: 9500, 9600, 9700, 9800, X300, X550, X600, X700, X800, X850 AllInWonder, and X1050 cards affecting all laptops bought before Q3 2005 (or January 2007 in the case of the X1050). All of which are still *very* prevalent in the group of laptops that make up the installed user base today and will continue to do so through 11.2 release. This issue of dropping support is separate and apart from issue (1) above because the group of cards in (1) are supposedly still supported.

Other Thoughts:

After Stefan Dirsch handed over the reins of driver maintainer to Bob Walmsley I guess he is now the guy that this should be coordinated through. I am glad to try and help, and to run test cases, etc. since I drew the black bean in laptop selection so let me know if/when there are new cases to test. I follow the ATI driver releases and test each one so I have a good handle on where things stand today. The problem is that Novell needs to use its resources to get a working partner in ATI to address some of these issues. Luugi Marsan was in the past the ATI contact, but at present, I haven't a clue who the ATI contact is or even if there is one. Without a good contact, we are dead-in-the-water getting any of these issues addressed and we are guaranteed a significant number of stranded users if these issues are not addressed.

Right now the 8-9 driver works fine for 10.3 and 11.0, but unfortunately it is incompatible with 11.1. The ATI issues prevent me, and all others affected, from being able to move to 11.1. Hopefully, if Novell will start now to address some of these driver issues with ATI, then by 11.2 release they have a good chance of ironed out.

With a desktop, you can always rip the ATI card out and replace it. With a laptop ... you are kind-of stuck.

-- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com

-- Susanne Oberhauser +49-911-74053-574 SUSE -- a Novell Business OPS Engineering Maxfeldstraße 5 Processes and Infrastructure Nürnberg SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, Apr 10, 2009 at 03:37, David C. Rankin wrote:

Listmates,

       There are 2 issues that need to be addressed regarding the fglrx driver and the upcoming 11.2 release. Currently in 11.1, there are many laptop cards supposedly "supported" by the ATI driver that crash or hardlock when the driver is installed. Currently my X1200 is one, the 200 is another from the list (I'm not sure what the technical card architecture is for the "200"). Suffice it to say, this is a significant problem for all affected users. The issues that need to be addressed pre-11.2 are:

"Radeon Xpress 200?" -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On 2009-06-24 21:24:41 +0200, Susanne Oberhauser wrote:

Stefan Dirsch writes:

...

On Tue, Jun 23, 2009 at 04:39:49PM +0200, Susanne Oberhauser wrote:

...

Now what we _can_ do is, we start building the different ati driver versions in parallel in an obs project:

project: ati-fglrx packages: ati-fglrx-8.8 ati-fglrx-8.9 ati-fglrx-8.10 ati-fglrx-9.1 ati-fglrx-9.2 ..

Unfortunately we can't do this for fglrx driver and probably any proprietary driver in obs. So either for this we would need a 3rd party buildservice or convince ATI to host such a system of repos.

packman?

imho ... this should be done by amd/ati. not by packman or us. darix -- openSUSE - SUSE Linux is my linux openSUSE is good for you www.opensuse.org -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Wed, Jun 24, 2009 at 09:24:41PM +0200, Susanne Oberhauser wrote:

Stefan Dirsch writes:

...

On Tue, Jun 23, 2009 at 04:39:49PM +0200, Susanne Oberhauser wrote:

...

Now what we _can_ do is, we start building the different ati driver versions in parallel in an obs project:

project: ati-fglrx packages: ati-fglrx-8.8 ati-fglrx-8.9 ati-fglrx-8.10 ati-fglrx-9.1 ati-fglrx-9.2 ..

Unfortunately we can't do this for fglrx driver and probably any proprietary driver in obs. So either for this we would need a 3rd party buildservice or convince ATI to host such a system of repos.

packman?

How many years do I know provide the RPMs "sources" in X11:Drivers:Video for building the fglrx locally via osc or a 3rd party buildservice (same for nvidia, btw)? Nobody ever showed interest into doing this. I doubt this changed meanwhile and is going to change ever. :-( Best regards, Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE LINUX Products GmbH Tel: 0911-740 53 0 Maxfeldstraße 5 FAX: 0911-740 53 479 D-90409 Nürnberg http://www.suse.de Germany ----------------------------------------------------------------- SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) ----------------------------------------------------------------- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

"David C. Rankin" writes:

Now for openSuSE, currently there is no upgrade path beyond 11.0 for users effected by the 8-10 to 9-3 driver issues, because the 8-9 driver does NOT support xorg 7.4. The only fglrx driver offering that opensuse has is for the 2400+ Series cards. (Guess how many laptops have 2400+ Series GPUs -- very, very few)

Did we ever discuss to build the current user space for the older X and kernel? I'm asking because on the older systems there is no big value in current kernel or just X --- usually, what worked with 10.0 or 10.1 or 10.2 works today and what didn't work doesn't, but there is a huge desire for current applications. I understand that if kernel system management APIs have changed (like /proc or /sys or ioctls), then tools that rely on that (udev/hal/dbus stack, YaST, alsa stack) need tweaks to work with theolder X and/or kernel. But I think it's just unrealistic to really change such component vendors. The only thing that we can realistically change is ourselves and how we deal with these realities. So what if we pick a kernel and X that will be around for a while (like the ones that are coincidential with sle release, hint, hint) and give this current-userspace-on-dated-systems a spin? Q then would be: what is the delineation line from user space to sys mgmgt and hardware abstraction... S. -- Susanne Oberhauser +49-911-74053-574 SUSE -- a Novell Business OPS Engineering Maxfeldstraße 5 Processes and Infrastructure Nürnberg SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

...

...

On 6/25/2009 at 14:27, Susanne Oberhauser wrote: Finding a project to eventually host the resulting binaries will then be the next step, and we'll need to explain why this is a good thing for the adoption of Linux, and how this actually supports the open source community.

Hmm. how much data transfer is there to be expected? I could host the binaries on my own server (until legal notice appears ;) ). I'm running an OBS and as such could just link to the source project which is maintained by Stefan. It would merely be to have the binareis hosted 'off' the opensuse.org domain. If we can arrange for some DNS entries on opensuse-community.org for example, this would not cause a problem at all for me (my data transfer limit is 50GB/month at the moment of which I hardly use 1GB now). Dominique -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Stefan Dirsch writes:

...

For quite a while and it's a real cool thing, an excellent starting point.

Currently youu have the fglrx, fglrxG01 and fglrxG02 packages.

And nvidia-gfx, nvidia-gfxG01, nvidia-gfxG02.

right.

...

In _theory_ each of them is covering a larger number of chipsets. In practice, the current versions break some older card.

Now the question would be if it would be hard to 'save' the package when a new driver version comes out, e.g. in some X11:Drivers:Video:Previous?

Finding a project to eventually host the resulting binaries will then be the next step,

No, this would be the first step. Second step would be to find someone, who does the packaging work.

I don't understand: when the fglrx- package is updated to a new fglrx driver version, make a copy of the curent package: osc copypac X11:Drivers:Video fglrxG02 X11:Drivers:Video:Previous fglrx-9.2 That's all I'm talking about. S. -- Susanne Oberhauser +49-911-74053-574 SUSE -- a Novell Business OPS Engineering Maxfeldstraße 5 Processes and Infrastructure Nürnberg SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Am Donnerstag, 25. Juni 2009 14:37:04 schrieb Dominique Leuenberger:

Hmm. how much data transfer is there to be expected? I could host the binaries on my own server (until legal notice appears ;) ).

I'm running an OBS and as such could just link to the source project which is maintained by Stefan. It would merely be to have the binareis hosted 'off' the opensuse.org domain. If we can arrange for some DNS entries on opensuse-community.org for example, this would not cause a problem at all for me (my data transfer limit is 50GB/month at the moment of which I hardly use 1GB now).

Dominique

I really like that idea, but I am somewhat the opinion we can't just release amd totally out of it's duty here. Their current repos are somewhat messy, an updated kmp after a kernel update always lags behind quite some time, they say you should only use one and only that version with this and that card, etc. pp. Why not suggest them to sponsor a community maintained obs machine, they could even use this for hosting mandriva and fedora in the same go, with some messing around propably also debian and u-h the other ones. Karsten -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Thu, Jun 25, 2009 at 03:08:40PM +0200, Susanne Oberhauser wrote:

Stefan Dirsch writes:

...

Well, get rid of proprietary drivers is what the open source community wants ...

Sigh.

Who doesn't?

After a few years of trying ot enforce this, how about winning by numbers? for a change, first we gain the market share that we deserve. _then_ we can enforce things.

So ignore legal problems before we become big enough? Hah, no such chance, sorry. Remember, it's not just a legal issue: http://www.linuxfoundation.org/en/Kernel_Driver_Statement thanks, greg k-h -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Thu, Jun 25, 2009 at 04:25:38PM -0400, Ken Schneider - openSUSE wrote:

Greg KH pecked at the keyboard and wrote:

...

On Thu, Jun 25, 2009 at 03:08:40PM +0200, Susanne Oberhauser wrote:

...

Stefan Dirsch writes:

...

Well, get rid of proprietary drivers is what the open source community wants ... Sigh.

Who doesn't?

After a few years of trying ot enforce this, how about winning by numbers? for a change, first we gain the market share that we deserve. _then_ we can enforce things.

So ignore legal problems before we become big enough?

Hah, no such chance, sorry.

Remember, it's not just a legal issue: http://www.linuxfoundation.org/en/Kernel_Driver_Statement

thanks,

greg k-h

Then provide a working alternative!

Um, we are trying the best that we can, with our limited resources and access to limited specifications. As always, patches gladly accepted. Whines directed to /dev/null. greg k-h -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Luis Medinas pecked at the keyboard and wrote:

On Thu, 2009-06-25 at 16:25 -0400, Ken Schneider - openSUSE wrote:

...

Greg KH pecked at the keyboard and wrote:

...

On Thu, Jun 25, 2009 at 03:08:40PM +0200, Susanne Oberhauser wrote:

...

Stefan Dirsch writes:

...

Well, get rid of proprietary drivers is what the open source community wants ... Sigh.

Who doesn't?

After a few years of trying ot enforce this, how about winning by numbers? for a change, first we gain the market share that we deserve. _then_ we can enforce things. So ignore legal problems before we become big enough?

Hah, no such chance, sorry.

Remember, it's not just a legal issue: http://www.linuxfoundation.org/en/Kernel_Driver_Statement

thanks,

greg k-h Then provide a working alternative!

The alternative is using OSS drivers or just use graphic card that does have OSS drivers. In case of ATI, novell is working on a OSS driver. If you have an nvidia you can use neauveau, intel provides OSS drivers too. There are plenty of options!

Luis

That's not an alternative for people with laptops STUCK with LOUSY ATI cards. I'm not a fan of closed source either but when there is no alternative there is no choice. Personally I have (and only will use) NVidia cards and haven't heard of the neauveau driver. What repo is it available from? zypper wp neauveau Loading repository data... Reading installed packages... No providers of 'neauveau' found. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Greg KH writes:

On Thu, Jun 25, 2009 at 03:08:40PM +0200, Susanne Oberhauser wrote:

...

Stefan Dirsch writes:

...

Well, get rid of proprietary drivers is what the open source community wants ...

Sigh.

Who doesn't?

After a few years of trying ot enforce this, how about winning by numbers? for a change, first we gain the market share that we deserve. _then_ we can enforce things.

So ignore legal problems before we become big enough?

Hah, no such chance, sorry.

Remember, it's not just a legal issue: http://www.linuxfoundation.org/en/Kernel_Driver_Statement

I share this. I relaocated my family and left my friends to work on Open Source. I've just helped Serna, a proprietary xml editor to become open source, after five years of again and again explaining the benefits. So why am I approaching it different from you? two reasons: 1. There is tons of users who happen to have one of these fine cards that need that driver for The Joy of Linux. Our ethically correct hesitance whether it's right or wrong to have this proprietary piece'cr*p in the stack not solving their problem. 2. In one sentence, the graphics driver is the dongle locking high volume high performance gaming cards from the low volume gfx workstations market. With their hat on, would I canibalize this market just to make some open source freaks like Susanne Oberhauser or Greg KH happy? Certainly not. All I say is: let's please solve that end user problem first. ATI and nVidia will, I'm sure, come up with reasonable ideas if the Linux desktop share is growing to 5%, 10%, or more. S. -- Susanne Oberhauser +49-911-74053-574 SUSE -- a Novell Business OPS Engineering Maxfeldstraße 5 Processes and Infrastructure Nürnberg SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, Jun 26, 2009 at 11:37:40AM +0200, Dominique Leuenberger wrote:

...

...

...

On 6/25/2009 at 14:37, "Dominique Leuenberger" wrote: I'm running an OBS and as such could just link to the source project which is maintained by Stefan. It would merely be to have the binareis hosted 'off' the opensuse.org domain. If we can arrange for some DNS entries on opensuse-community.org for example, this would not cause a problem at all for me (my data transfer limit is 50GB/month at the moment of which I hardly use 1GB now).

I went down the ally and tried to link the sources from X11:Drivers:Video to my local OBS. Not surprisingly of course, this does not (yet) work, as the 'real' blob is missing in the Source repo on OBS (The user is supposed to download it before the build command). (I tried nvidia only... simply because I have those cards in all my systems and it would allow me to test the resulting RPMs).

Stefan, any good ideas on how we can achieve this to be 'more automatic' than it is now? Maybe integrating the 'fetch.sh' logic inside the spec file?

Where is the issue? If you link to openSUSE.org:X11:Drivers:Video, just add the 'real' blobs to the package of your local OBS.

I'll gladly offer my server for building the RPMs and I can host them on my server too, so we can get our users on track with a good experience... I would even build for Factory, but this would most likely cause some 'issues' on my side, as inter-obs does not link changing base repos that much (it typically end's up in a scheduler look).

Better forget about factory. The kABI changes all the time. Or do you want to host the kernel RPMs as well, so these get updated at the same time? Providing these for 10.3, 11.0 and 11.1 would be sufficient. Best regards, Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE LINUX Products GmbH Tel: 0911-740 53 0 Maxfeldstraße 5 FAX: 0911-740 53 479 D-90409 Nürnberg http://www.suse.de Germany ----------------------------------------------------------------- SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) ----------------------------------------------------------------- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, Jun 26, 2009 at 02:15:01PM +0200, Dominique Leuenberger wrote:

...

...

...

On 6/26/2009 at 12:19, Stefan Dirsch wrote: On Fri, Jun 26, 2009 at 11:37:40AM +0200, Dominique Leuenberger wrote: Stefan, any good ideas on how we can achieve this to be 'more automatic' than it is now? Maybe integrating the 'fetch.sh' logic inside the spec file?

Where is the issue? If you link to openSUSE.org:X11:Drivers:Video, just add the 'real' blobs to the package of your local OBS.

I'd have preferred a 'single point of change' for the entire thing. Like this I have to grab the pkg file whenever there is an update for example. but do-able.. I'll give it a try.

Will have to analyze the spec file how much it likes having the two different pkg.run for 64bit and 32bit.

Sure. This is implemented in the specfiles. Otherwise they would be completely useless.

...

...

I'll gladly offer my server for building the RPMs and I can host them on my Better forget about factory. The kABI changes all the time. Or do you want to host the kernel RPMs as well, so these get updated at the same time? Providing these for 10.3, 11.0 and 11.1 would be sufficient.

As I'm following Factory myself closely (have it running on 2 systems.. compared to 11.1 only one 1) I was thinking of having the KMPs too on the server. (Greg, please don't shoot!)

You mean the kernel RPMs, too, right?

It depends a bit on how useful it is, as the 'external' obs will always lag a little with the KMPs compared to the factory distribution. It might or might not work (especially I do not have an automatic 'publish' running.

See my comment above. Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE LINUX Products GmbH Tel: 0911-740 53 0 Maxfeldstraße 5 FAX: 0911-740 53 479 D-90409 Nürnberg http://www.suse.de Germany ----------------------------------------------------------------- SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) ----------------------------------------------------------------- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

...

...

On 6/26/2009 at 12:19, Stefan Dirsch wrote: Better forget about factory. The kABI changes all the time. Or do you want to host the kernel RPMs as well, so these get updated at the same time? Providing these for 10.3, 11.0 and 11.1 would be sufficient.

Building started for the nvidia modules (6 packages, the nvidia-* and the x11-nvidia*). Surprisingly, for openSUSE_11.1 the build of the kmps fails: (.o vs. .ko...) Dominique RPM build errors: cannot open Pubkeys index using db3 - No such file or directory (2) File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates/nvidia.ko File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates/nv-kernel.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates/nv-linux.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates/nvidia.ko File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates/nv-kernel.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates/nv-linux.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates/nvidia.ko File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates/nv-kernel.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates/nv-linux.o Installed (but unpackaged) file(s) found: /lib/modules/2.6.27.7-9-default/updates/nv-kernel.o /lib/modules/2.6.27.7-9-default/updates/nv-linux.o /lib/modules/2.6.27.7-9-default/updates/nvidia.ko /lib/modules/2.6.27.7-9-trace/updates/nv-kernel.o /lib/modules/2.6.27.7-9-trace/updates/nv-linux.o /lib/modules/2.6.27.7-9-trace/updates/nvidia.ko /lib/modules/2.6.27.7-9-xen/updates/nv-kernel.o /lib/modules/2.6.27.7-9-xen/updates/nv-linux.o /lib/modules/2.6.27.7-9-xen/updates/nvidia.ko -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, Jun 26, 2009 at 03:22:41PM +0200, Dominique Leuenberger wrote:

...

...

...

On 6/26/2009 at 12:19, Stefan Dirsch wrote: Better forget about factory. The kABI changes all the time. Or do you want to host the kernel RPMs as well, so these get updated at the same time? Providing these for 10.3, 11.0 and 11.1 would be sufficient.

Building started for the nvidia modules (6 packages, the nvidia-* and the x11-nvidia*).

Surprisingly, for openSUSE_11.1 the build of the kmps fails: (.o vs. .ko...)

Dominique

RPM build errors: cannot open Pubkeys index using db3 - No such file or directory (2) File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates/nvidia.ko File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates/nv-kernel.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-default-default/updates/nv-linux.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates/nvidia.ko File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates/nv-kernel.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-trace-trace/updates/nv-linux.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates/nvidia.ko File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates/nv-kernel.o File not found: /var/tmp/nvidia-gfxG01-173.14.18-build/lib/modules/2.6.27.7-9-xen-xen/updates/nv-linux.o Installed (but unpackaged) file(s) found: /lib/modules/2.6.27.7-9-default/updates/nv-kernel.o /lib/modules/2.6.27.7-9-default/updates/nv-linux.o /lib/modules/2.6.27.7-9-default/updates/nvidia.ko /lib/modules/2.6.27.7-9-trace/updates/nv-kernel.o /lib/modules/2.6.27.7-9-trace/updates/nv-linux.o /lib/modules/2.6.27.7-9-trace/updates/nvidia.ko /lib/modules/2.6.27.7-9-xen/updates/nv-kernel.o /lib/modules/2.6.27.7-9-xen/updates/nv-linux.o /lib/modules/2.6.27.7-9-xen/updates/nvidia.ko

I don't see these issues in our internal obs. Looks like you need to use kmp-filelist-old for your openSUSE_11.1 target. No idea why. See nvidia-gfxG01.spec: [...] # patch the kmp template %if %suse_version > 1100 %define kmp_template -t %define kmp_filelist kmp-filelist %define kmp_post kmp-post.sh %else %define kmp_template -s %define kmp_filelist kmp-filelist-old %define kmp_post kmp-post-old.sh %endif [...] Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE LINUX Products GmbH Tel: 0911-740 53 0 Maxfeldstraße 5 FAX: 0911-740 53 479 D-90409 Nürnberg http://www.suse.de Germany ----------------------------------------------------------------- SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) ----------------------------------------------------------------- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, Jun 26, 2009 at 09:19:33AM -0700, Greg KH wrote:

On Fri, Jun 26, 2009 at 02:15:01PM +0200, Dominique Leuenberger wrote:

...

...

...

...

On 6/26/2009 at 12:19, Stefan Dirsch wrote: On Fri, Jun 26, 2009 at 11:37:40AM +0200, Dominique Leuenberger wrote: Stefan, any good ideas on how we can achieve this to be 'more automatic' than it is now? Maybe integrating the 'fetch.sh' logic inside the spec file?

Where is the issue? If you link to openSUSE.org:X11:Drivers:Video, just add the 'real' blobs to the package of your local OBS.

I'd have preferred a 'single point of change' for the entire thing. Like this I have to grab the pkg file whenever there is an update for example. but do-able.. I'll give it a try. Will have to analyze the spec file how much it likes having the two different pkg.run for 64bit and 32bit.

...

...

I'll gladly offer my server for building the RPMs and I can host them on my Better forget about factory. The kABI changes all the time. Or do you want to host the kernel RPMs as well, so these get updated at the same time? Providing these for 10.3, 11.0 and 11.1 would be sufficient.

As I'm following Factory myself closely (have it running on 2 systems.. compared to 11.1 only one 1) I was thinking of having the KMPs too on the server. (Greg, please don't shoot!)

Remember, you can not redistribute these packages, so you better not make them public for anyone else to get :)

What you do on your own systems is up to you.

So you're going to sue Dominique if he does? CU, Stefan Public Key available ------------------------------------------------------ Stefan Dirsch (Res. & Dev.) SUSE LINUX Products GmbH Tel: 0911-740 53 0 Maxfeldstraße 5 FAX: 0911-740 53 479 D-90409 Nürnberg http://www.suse.de Germany ----------------------------------------------------------------- SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) ----------------------------------------------------------------- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

...

...

On 6/26/2009 at 19:25, Greg KH wrote: So you're going to sue Dominique if he does?

If he redistributes the prebuilt nvidia driver, I will take the same action that I have taken against other individuals and companies that distribute such a thing in the past.

Knowingly ignoring someone who violates your copyright doesn't look good if you wish to do future enforcement of your copyright. Or so says my lawyers...

Dear Greg, Wow, that's a nice attitude from a direct member of the project I was intending to help. Greg: just out of curiosity: did your lawyers finally contact NVidia and ATI? You might be aware that their servers are distributing binary blobs. Just in case you were not aware: Take this mail in it's informative way to let you know about the fact. Should you now actually consider NOT pursuing against them, I can more or less be sure to be able to use exactly THIS fact against any steps you might raise against me. It might not put you in the best shed by having to admit that you *knowingly* accepted somebody doing it and for somebody else you do not tolerate it. And as of course nobody could say something if you're not officially made aware and nobody could verify your knowledge of such an infringement, this mail serves exactly THAT purpose. Best wishes, Dominique -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, Jun 26, 2009 at 07:37:08PM +0200, Dominique Leuenberger wrote:

...

...

...

On 6/26/2009 at 19:25, Greg KH wrote: So you're going to sue Dominique if he does?

If he redistributes the prebuilt nvidia driver, I will take the same action that I have taken against other individuals and companies that distribute such a thing in the past.

Knowingly ignoring someone who violates your copyright doesn't look good if you wish to do future enforcement of your copyright. Or so says my lawyers...

Dear Greg,

Wow, that's a nice attitude from a direct member of the project I was intending to help.

{sigh} What would you do if someone violates your copyright? Would you expect a company to ignore it? Would you expect an individual?

Greg: just out of curiosity: did your lawyers finally contact NVidia and ATI? You might be aware that their servers are distributing binary blobs. Just in case you were not aware: Take this mail in it's informative way to let you know about the fact.

My lawyers have contacted them, and many other companies/individuals who have done the same in the past.

Should you now actually consider NOT pursuing against them, I can more or less be sure to be able to use exactly THIS fact against any steps you might raise against me. It might not put you in the best shed by having to admit that you *knowingly* accepted somebody doing it and for somebody else you do not tolerate it.

This is not true, see above. Also, please note that there are some very large companies who hold the copyright on the kernel, and they have some very large legal departments, so don't think it is just individuals who you would be infringing apon. best of luck, greg k-h -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Fri, 26 Jun 2009 20:33:16 +0200, you wrote:

X-Mailer: Novell GroupWise Internet Agent 8.0.0 Date: Fri, 26 Jun 2009 20:33:16 +0200

Dominique, please use a different MUA. As GW can't do proper replies (i.e. generate In-Reply-To) , you're disrupting the reply threading and make it extra hard to follow the discussion. Philipp -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Sat, 27 Jun 2009 11:00:19 +0200, you wrote:

Greg's and Dominique's mails are threaded correctly both in KMail and in the archives,

Yes, they are. I've now discoverwed the reason why I was iritated. I have three different threads all starting with a mail by Dominique and all with the same subject but with different In-Reply-To: In-Reply-To: <20090626101936.GA18628@suse.de> In-Reply-To: <20090626125131.GA20365@suse.de> In-Reply-To: <20090626182125.GA17846@suse.de> So I'm missing the start of the thread and thus the irritating threading. Sorry for the noise. Philipp -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

On Thursday 25 June 2009 12:29:11 am Susanne Oberhauser wrote:

"David C. Rankin" writes:

...

Now for openSuSE, currently there is no upgrade path beyond 11.0 for users effected by the 8-10 to 9-3 driver issues, because the 8-9 driver does NOT support xorg 7.4. The only fglrx driver offering that opensuse has is for the 2400+ Series cards. (Guess how many laptops have 2400+ Series GPUs -- very, very few)

Did we ever discuss to build the current user space for the older X and kernel?

Susanne, I don't understand the technical issues involved enough to be of much use extracting the 8-9 driver and patching it to run on 11.1, but I can tell you it is something that should be seriously looked at or there are going to be a whole lot of laptop users with ATI hardware that are given the impression that Linux just isn't a serious desktop yet when 11.2 is released. I swapped drives tonight to do some additional testing of kde 4.2.95-1 on Archlinux working with the radeonhd driver. I pulled built and installed the latest git versions of drm/mesa and the radeonhd driver. Desktop effect performance is miserable -- to the point that the compositing manager shuts down when doing cube rotation because the compositing is too slow (or whatever the exact text of the message you get in the lower right of the desktop says) Earlier today (for the better part of the day) I was working on this same laptop with my drive that holds openSuSE 11.0/kde4.3 Beta 2 where I can still use the ATI 8-9 fglrx driver. Everything worked fine in kde 4.3 beta 2 with the fglrx driver. Performance was excellent, desktop effects worked perfectly, the only difference -- the fglrx driver instead of the radeonhd driver. The difference is Night-and-day. If I were a new user that loaded 11.1 or the upcoming 11.2 without having the benefit of the fglrx driver -- I would get the impression that Linux was just to bloated to run well on my hardware and probably ditch it. It makes that big of a difference. If Novell/openSuSE has any resources to look into what it would take to move the 8-9 driver forward to the current xorg for 11.2, or what options there are to negotiate with ATI to fix the issues introduced with the 8-10 through 9-3 drivers, it would probably be one of the best uses of time and manpower to prevent against 11.2 being a dud on a whole lot of laptop hardware. Keep the faith... -- David C. Rankin, J.D.,P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Peter Nikolic wrote:

On Thursday 25 Jun 2009 21:29:58 Greg KH wrote:

...

Whines directed to /dev/null. now that it has to be said is just flippin typical criticise a duff decission and you get dev/nulled nuff said know what i mean ?

Pete .

In the days when I was involved in providing customers with solutions, cooperation between and within teams was key to delivery of world-class results and high customer satisfaction. If I have a problem I seek a solution and will work with all and sundry to achieve that, egos are disruptive of the process, so let's focus on delivery. As I have opined in private email to other concerned users, I hope that the techies get a fair hearing and are not given arbitrary deadlines for 11.2 - I've been there and I know how ugly it gets, a bad product followed by a brilliant follow-up or clean-up still equals a disaster and loss of customers -- Remember Vista! The one that was supposed to be the King of OS's. The marketing guys will happily move on to selling dog food, making the numbers and commission, unfortunately no such easy out for the techies. I support the efforts here, even when some of my bugs do not seem to be getting the attention I expect - I understand the picture from your side, no bones to pick with anyone, just focussed on helping you to help us. Regards Sid. -- Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support Specialist, Cricket Coach Microsoft Windows Free Zone - Linux used for all Computing Tasks -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Greg KH writes:

...

...

...

As I'm following Factory myself closely (have it running on 2 systems.. compared to 11.1 only one 1) I was thinking of having the KMPs too on the server. (Greg, please don't shoot!)

Remember, you can not redistribute these packages, so you better not make them public for anyone else to get :)

What you do on your own systems is up to you.

So you're going to sue Dominique if he does?

If he redistributes the prebuilt nvidia driver, I will take the same action that I have taken against other individuals and companies that distribute such a thing in the past.

Knowingly ignoring someone who violates your copyright doesn't look good if you wish to do future enforcement of your copyright. Or so says my lawyers...

You could grant this one individual the License to use your personal fraction of the overall kernel copyright in liaison with the ATI and nvidia driver. As the copyright owner you are allowed to do such a dual License. Supposed the primary objective was to make openSUSE users happy and make openSUSE the insanely easy solution to "I want to use Linux on this machine". And solve the license problem when we have 20, 30, 40 % desktop and workstation market share, growing. Which the current tactics is extremely effectively blocking. btw, how's the status sueing ubuntu/canonical on their villainous copyright infringements with their fglrx binary packages? They are serving their users first, and then, with the help of a large number of users, they'll change rules to the better of the GPL. S. -- Susanne Oberhauser +49-911-74053-574 SUSE -- a Novell Business OPS Engineering Maxfeldstraße 5 Processes and Infrastructure Nürnberg SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

* Susanne Oberhauser [07-01-09 06:47]:

You could grant this one individual the License to use your personal fraction of the overall kernel copyright in liaison with the ATI and nvidia driver. As the copyright owner you are allowed to do such a dual License.

?? Maybe the "you" is a problem.. Does he, "you", control the copyright or is it controlled by a very large group of individuals? -- Patrick Shanahan Plainfield, Indiana, USA HOG # US1244711 http://wahoo.no-ip.org Photo Album: http://wahoo.no-ip.org/gallery2 Registered Linux User #207535 @ http://counter.li.org -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

Patrick Shanahan writes:

* Susanne Oberhauser [07-01-09 06:47]:

...

You could grant this one individual the License to use your personal fraction of the overall kernel copyright in liaison with the ATI and nvidia driver. As the copyright owner you are allowed to do such a dual License.

?? Maybe the "you" is a problem.. Does he, "you", control the copyright or is it controlled by a very large group of individuals?

IANAL, yet afaict you control the copyright to subsystems you have contributed. If that is core you basically control the kernel. That's how the usb subsystem got pure "export gpl". A friend of mine now won't use linux as he just wanted to play with it and needs to buy a new usb wlan stick to play this game. Ain't gonna happen. S. -- Susanne Oberhauser +49-911-74053-574 SUSE -- a Novell Business OPS Engineering Maxfeldstraße 5 Processes and Infrastructure Nürnberg SUSE LINUX Products GmbH, GF: Markus Rex, HRB 16746 (AG Nürnberg) -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On Friday, 2009-07-03 at 15:32 -0500, Rajko M. wrote:

On Thursday 02 July 2009 12:32:33 am Susanne Oberhauser wrote:

...

IANAL, yet afaict you control the copyright to subsystems you have contributed. If that is core you basically control the kernel. That's how the usb subsystem got pure "export gpl". A friend of mine now won't use linux as he just wanted to play with it and needs to buy a new usb wlan stick to play this game. Ain't gonna happen.

Same here, I can't get Linux on any box where user rather pay and keeps running bunch of malware protection applications then to give up on fast graphics, multimedia support, or some hardware.

You are absolutely right that when Linux makes 20% of the market, then discussion with vendors will have quite different base, but keeping principles, or "principles", right now, is efficient way to prevent that this ever happen.

IMO, that's true. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.9 (GNU/Linux) iEYEARECAAYFAkpPGLkACgkQtTMYHG2NR9XdRQCggDRJm8lCU9AyeNpBi/lBwaSy O/UAn3/cQSUYawbB1VYvsCvtXiddKdDQ =TPTv -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org For additional commands, e-mail: opensuse-factory+help@opensuse.org