Re: [MeeGo-dev] non SSSE3 MeeGo
Le vendredi 13 août 2010 à 22:33 +0200, Jean-Christian de Rivaz a écrit : Greg KH a écrit : On Fri, Aug 13, 2010 at 09:20:47PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 05:37:04PM +0200, Jean-Christian de Rivaz wrote: Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? Speed. Seriously, go measure it with it turned off, it is very noticable. And on these tiny netbooks, you need all the speed you can get. There's a reason MeeGo is the fastest booting and running of _all_ distros out there at the moment, and this is one of them. Easy to say, but what real facts can you show to support your claim ? I just got a report from someone today that analized the boot times of all of the currently released distros, and MeeGo was the fastest. Please show, this is interesting. And here it comes : http://blog.crozat.net/2010/08/some-boot-time-comparison-meego-is.html -- Frederic Crozat fcro...@novell.com Novell ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Message: 10 Date: Fri, 13 Aug 2010 22:27:04 +0200 From: Jean-Christian de Rivaz j...@eclis.ch To: Development for the MeeGo Project (discussion list) meego-dev@meego.com Subject: Re: [MeeGo-dev] non SSSE3 MeeGo Message-ID: 4c65aa98.9050...@eclis.ch Content-Type: text/plain; charset=ISO-8859-1; format=flowed SMEEGOL Snip: Did you have an URL ? http://www.wafaa.eu/tag/Smeegol http://www.wafaa.eu/tag/Smeegol JLNH ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Auke Kok a écrit : On 08/12/2010 12:33 PM, Alistair Buxton wrote: no matter how much we lower the baseline, there will always be more people saying but I have this CPU and that one is not supported. At one point you have to say stop, and disappoint people. You can't satisfy everyone. Dear Auke, Given the actual (non existant) position of Meego on the market, you may better considering to satisfy peoples other than just on the edge Intel users if you expect to get a chance to be adopted by a large community. If the expected fantastic build system (OBS) Meego is based on have a real value, then it must be trivial to build a i686 target like every large distribution do. Finally, when a code can get real boost with a set of specific instruction, it's a good practice to detect the instruction set at runtime to select between the generic and to optimized code. I think that every projects working hard on codec have learned this to make everyone happy. Best Regards, Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 05:37:04PM +0200, Jean-Christian de Rivaz wrote: Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? Speed. Seriously, go measure it with it turned off, it is very noticable. And on these tiny netbooks, you need all the speed you can get. There's a reason MeeGo is the fastest booting and running of _all_ distros out there at the moment, and this is one of them. What really block you to do things that all others do easily for years ? Just Start a OBS target with generic compiler flags, end of the story. Exactly, everyone is free to do that, so what's the big deal? Or are you wanting someone else to do this for you? confused as to the whining, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
- Original message - From: Jean-Christian de Rivaz j...@eclis.ch To: Development for the MeeGo Project (discussion list) meego-dev@meego.com Subject: Re: [MeeGo-dev] non SSSE3 MeeGo Date: Fri, 13 Aug 2010 17:37:04 +0200 Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? What really block you to do things that all others do easily for years ? Just Start a OBS target with generic compiler flags, end of the story. Forgive my ignorance if I'm off base, but isn't this a policy topic for the Linux Foundation to address? Should Ibrahim make a statement? Randy -- Ovi Mail: Making email access easy http://mail.ovi.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 13 August 2010 17:53, Greg KH gre...@suse.de wrote: On Fri, Aug 13, 2010 at 05:37:04PM +0200, Jean-Christian de Rivaz wrote: Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? Speed. Seriously, go measure it with it turned off, it is very noticable. And on these tiny netbooks, you need all the speed you can get. There's a reason MeeGo is the fastest booting and running of _all_ distros out there at the moment, and this is one of them. What really block you to do things that all others do easily for years ? Just Start a OBS target with generic compiler flags, end of the story. Exactly, everyone is free to do that, so what's the big deal? Or are you wanting someone else to do this for you? confused as to the whining, I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. The logic is valid, this is a community project, why does the x86 port only support the Intel atom? If it's just the speed then we could just have two targets, atom and generic i686. greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 8/13/2010 9:01 AM, Brendan Le Foll wrote: I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. The logic is valid, this is a community project, why does the x86 port only support the Intel atom? If it's just the speed then we could just have two targets, atom and generic i686. your statement keeps being incorrect. the x86 port draws the minimum bar at the Core2 instruction set, not the Atom one. (Core2 is a 2006/2007 chip, not super recent by any stretch of imagination) what is generic i686? Fedora draws the bar at Pentium IV there (SSE2 or something like that) debian puts it at the Pentium II what do you call generic i686 ? ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
your statement keeps being incorrect. the x86 port draws the minimum bar at the Core2 instruction set, not the Atom one. Don't they use the same instruction set? What is the difference? (Core2 is a 2006/2007 chip, not super recent by any stretch of imagination) what is generic i686? Fedora draws the bar at Pentium IV there (SSE2 or something like that) debian puts it at the Pentium II what do you call generic i686 ? What wikipedia calls it. P6 microarchitecture. I'm not all that great with my intel lingo sorry ;-) I didn't realise i686 was probably not the right term for it. If it's because of the lack of support then why is the answer not something like - if there are people from the community willing to do it then we'll add the target on the internal OBS - Sorry about the address mixup Arjan. ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, 13 Aug 2010, Brendan Le Foll wrote: If it's because of the lack of support then why is the answer not something like - if there are people from the community willing to do it then we'll add the target on the internal OBS One reason: because people from the community is hard to depend on (at times)... and when they are MIA, the the Linux Foundation will get stuck with the responsibility. If the community is serious about supporting a non-Core-2 arch, then I would suggest that they set up their own OBS and demonstrate just how serious they are about supporting it. As far as I can tell, even OBS is in Gitorious... so they've already given you the tools. -gabriel ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 13 August 2010 18:26, Gabriel M. Beddingfield gabrb...@gmail.com wrote: If the community is serious about supporting a non-Core-2 arch, then I would suggest that they set up their own OBS and demonstrate just how serious they are about supporting it. As far as I can tell, even OBS is in Gitorious... so they've already given you the tools. I have set up my own OBS server. How do I clone the Meego project into it without a login on build.meego.com? After I do that, how do I bootstrap it, given that OBS relies on existing RPM packages that are built for SSSE3? Last time we had this discussion, 6 months ago, nobody was able to answer either one. -- Alistair Buxton a.j.bux...@gmail.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 5:16 PM, Arjan van de Ven ar...@linux.intel.com wrote: On 8/13/2010 9:01 AM, Brendan Le Foll wrote: I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. The logic is valid, this is a community project, why does the x86 port only support the Intel atom? If it's just the speed then we could just have two targets, atom and generic i686. your statement keeps being incorrect. the x86 port draws the minimum bar at the Core2 instruction set, not the Atom one. (Core2 is a 2006/2007 chip, not super recent by any stretch of imagination) what is generic i686? Fedora draws the bar at Pentium IV there (SSE2 or something like that) debian puts it at the Pentium II Actually no it doesn't it draws the bar at i686. That includes the AMD Geode processor that runs on the XO-1. Peter ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 06:01:36PM +0200, Brendan Le Foll wrote: On 13 August 2010 17:53, Greg KH gre...@suse.de wrote: [snap] I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. Are you willing to support it? If so, great, but note that supported means a lot here... I want to understand what you means by a lot here. As far as I known, Meego, currently, almost exclusively use the same upstream projects as any major distribution does. All that majors distributions seem to have no issue at all in building a generic enough x86 (ex: i686) target. So why do you think that this will be a lot of work for Meego ? Please explain. Regards, Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 13 August 2010 19:13, Gabriel M. Beddingfield gabrb...@gmail.com wrote: Maybe something like this: http://meego.gitorious.org/meego-developer-tools/obs-project-config/blobs/master/MeeGo:1.0:Core You show a prjconf for Meego. I already have this, and have loaded it into my OBS server. According to the documentation on Meego wiki I still need to use obs_mirror_project to mirror the Meego files. (http://wiki.meego.com/Build_Infrastructure/Sysadmin_Distro/OBS1.8_setup_openSUSE112#Cloning_Repositories) This process is not documented except to say that the file must be edited and process requires a login on build.meego.com which I do not possess. Or maybe asking nicely... I think you'll find I did ask nicely. Why so defensive? Or maybe do it the hard way... like /they/ did. So you admit it's not as easy as you and others have previously claimed? In addition the package repos are mirrored at mirror.kernel.org, including the daily builds (if that's any help). bootstrap it, given that OBS relies on existing RPM packages that are built for SSSE3? You either host it on a Core 2 server, or rebuild the packages. Or is this a trick question? :-) I don't have a Core 2 server so I need to rebuild the packages - otherwise known as bootstrapping. Which is why I asked how to bootstrap. You are telling me that I must rebuild the packages by rebuilding the packages. Can you see why I find this a little bit annoying? If you don't want to help that is fine, but why do you insist on derailing the topic with attackes any time anyone even suggests it? -- Alistair Buxton a.j.bux...@gmail.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 8/13/2010 11:32 AM, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 06:01:36PM +0200, Brendan Le Foll wrote: On 13 August 2010 17:53, Greg KHgre...@suse.de wrote: [snap] I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. Are you willing to support it? If so, great, but note that supported means a lot here... I want to understand what you means by a lot here. As far as I known, Meego, currently, almost exclusively use the same upstream projects as any major distribution does. All that majors distributions seem to have no issue at all in building a generic enough x86 (ex: i686) target. So why do you think that this will be a lot of work for Meego ? Please explain. it's more than zero work. Auke ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 08:32:21PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 06:01:36PM +0200, Brendan Le Foll wrote: On 13 August 2010 17:53, Greg KH gre...@suse.de wrote: [snap] I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. Are you willing to support it? If so, great, but note that supported means a lot here... I want to understand what you means by a lot here. As far as I known, Meego, currently, almost exclusively use the same upstream projects as any major distribution does. All that majors distributions seem to have no issue at all in building a generic enough x86 (ex: i686) target. So why do you think that this will be a lot of work for Meego ? It's the line supported that I'm concerned about. Is the community going to be ready to accept all bug reports and other issues reported by these different arch packages? That includes the noticable slow-down that happens when building for a different arch. People will notice... thanks, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, 13 Aug 2010, Alistair Buxton wrote: You show a prjconf for Meego. I already have this, and have loaded it into my OBS server. According to the documentation on Meego wiki I still need to use obs_mirror_project to mirror the Meego files. (http://wiki.meego.com/Build_Infrastructure/Sysadmin_Distro/OBS1.8_setup_openSUSE112#Cloning_Repositories) This process is not documented except to say that the file must be edited and process requires a login on build.meego.com which I do not possess. What is it mirroring? The package repos? Those are mirrored at kernel.org. (I honestly don't know what's being mirrored... but the pkg repo makes the most sense... just a guess) build.meego.com doesn't have a public rsync which you need for efficient mirroring. So, to mirror from there you would need a log in. However, kernel.org DOES have a public rsync.) Or maybe asking nicely... I think you'll find I did ask nicely. Why so defensive? You mean today? I think in the context of the discussion... there are many who would not consider that message a polite request. Or maybe do it the hard way... like /they/ did. So you admit it's not as easy as you and others have previously claimed? I've never done it. I don't know if it's easy or hard. Looks like it's probably a lot easier than setting up the FTP server system that Debian uses. (Not buildd... the one that they /really/ use.) I don't have a Core 2 server so I need to rebuild the packages - I rent a virtual one for about US$25/mo. Hosted in a data center. otherwise known as bootstrapping. Which is why I asked how to bootstrap. You are telling me that I must rebuild the packages by rebuilding the packages. Can you see why I find this a little bit I still don't understand your problem. If you're rebuilding packages, why do you need Core 2? Are the packages using assembly code or something? Or are you trying to use a pre-built binary of OBS or something? -gabriel ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 05:37:04PM +0200, Jean-Christian de Rivaz wrote: Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? Speed. Seriously, go measure it with it turned off, it is very noticable. And on these tiny netbooks, you need all the speed you can get. There's a reason MeeGo is the fastest booting and running of _all_ distros out there at the moment, and this is one of them. Easy to say, but what real facts can you show to support your claim ? The booting process has been analyzed by many peoples the last couple of years. I have not read all of there conclusions, but from what I remember, reducing the I/O activity (in number and in latency) was a far biggest point than the SSSE3 set instruction. If I compare the boot time of for example Debian squeeze i386 and Debian squeeze amd64 on the same machine (with a AMD or a Intel CPU), I did not see a noticeable difference. The SSSE3 instructions will probably only play a noticeable difference in a codec or stream processing code. In those case, the good practice is to select at runtime between a generic code and a instruction specific code. If fact I highly doubt that any upstream project Meego uses will run only on SSSE3 machine. Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
'MeeGo is the community' and is helping support Meego though Forum Triage, the MeeGo Greeters program, new platforms, raising bug reports, solving big reports etc. We are all in this together paid/unpaid, Nokia/Intel affiliated or not. vgrade Greg KH wrote: On Fri, Aug 13, 2010 at 08:32:21PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 06:01:36PM +0200, Brendan Le Foll wrote: On 13 August 2010 17:53, Greg KH gre...@suse.de wrote: [snap] I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. Are you willing to support it? If so, great, but note that supported means a lot here... I want to understand what you means by a lot here. As far as I known, Meego, currently, almost exclusively use the same upstream projects as any major distribution does. All that majors distributions seem to have no issue at all in building a generic enough x86 (ex: i686) target. So why do you think that this will be a lot of work for Meego ? It's the line supported that I'm concerned about. Is the community going to be ready to accept all bug reports and other issues reported by these different arch packages? That includes the noticable slow-down that happens when building for a different arch. People will notice... thanks, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 13 August 2010 20:12, Gabriel M. Beddingfield gabrb...@gmail.com wrote: On Fri, 13 Aug 2010, Alistair Buxton wrote: You show a prjconf for Meego. I already have this, and have loaded it into my OBS server. According to the documentation on Meego wiki I still need to use obs_mirror_project to mirror the Meego files. (http://wiki.meego.com/Build_Infrastructure/Sysadmin_Distro/OBS1.8_setup_openSUSE112#Cloning_Repositories) This process is not documented except to say that the file must be edited and process requires a login on build.meego.com which I do not possess. What is it mirroring? The package repos? Those are mirrored at kernel.org. (I honestly don't know what's being mirrored... but the pkg repo makes the most sense... just a guess) build.meego.com doesn't have a public rsync which you need for efficient mirroring. So, to mirror from there you would need a log in. However, kernel.org DOES have a public rsync.) I do not know what it is mirroring... I cannot find any documentation, and I cannot look into build.meego.com to see what it is trying to fetch, so I don't know if any of these mirrors actually contain the correct files or not. Or maybe asking nicely... I think you'll find I did ask nicely. Why so defensive? You mean today? I think in the context of the discussion... there are many who would not consider that message a polite request. You should see the messages I decided *not* to send :) Or maybe do it the hard way... like /they/ did. So you admit it's not as easy as you and others have previously claimed? I've never done it. I don't know if it's easy or hard. Looks like it's probably a lot easier than setting up the FTP server system that Debian uses. (Not buildd... the one that they /really/ use.) But how does it compare with, say, building an android image? Or building gentoo from stage1 using crosstools? From my point of view the answer is not favorably. I don't have a Core 2 server so I need to rebuild the packages - I rent a virtual one for about US$25/mo. Hosted in a data center. I have a perfectly good quad core server, it just doesn't have SSSE3. If it did, I wouldn't need to go to all this trouble in the first place. otherwise known as bootstrapping. Which is why I asked how to bootstrap. You are telling me that I must rebuild the packages by rebuilding the packages. Can you see why I find this a little bit I still don't understand your problem. If you're rebuilding packages, why do you need Core 2? Are the packages using assembly code or something? Or are you trying to use a pre-built binary of OBS or something? The problem is I don't know how to tell OBS to rebuild all the packages without using prebuilt Meego RPMs to do it - all the existing documentation is geared towards someone who wants to build a single package against an already bootstrapped distro. -- Alistair Buxton a.j.bux...@gmail.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 08:32:21PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 06:01:36PM +0200, Brendan Le Foll wrote: On 13 August 2010 17:53, Greg KH gre...@suse.de wrote: [snap] I guess people want the official OBS to have the generic i686 target so it becomes a supported Architecture. Are you willing to support it? If so, great, but note that supported means a lot here... I want to understand what you means by a lot here. As far as I known, Meego, currently, almost exclusively use the same upstream projects as any major distribution does. All that majors distributions seem to have no issue at all in building a generic enough x86 (ex: i686) target. So why do you think that this will be a lot of work for Meego ? It's the line supported that I'm concerned about. Is the community going to be ready to accept all bug reports and other issues reported by these different arch packages? That includes the noticable slow-down that happens when building for a different arch. People will notice... My question was about technical fact. Not politic. politic on This was the Meego decision, without the community, to not be based on a upstream distribution to solve this kind of issue. Several of us have warned at that time that this will be a big wast of time. The argument was severely rejected by claim that Meego will be something incredible. It's now maybe the time to lower the expectation and to go back to the reality: without a generic enough build, Meego will never reach any audience like Ubuntu for example. politic off Now, I am ok the let my politic thinking above all in a trash and to pragmatically take the actual situation as a fact. So please do the same and show your real facts without escaping into a flameware. If some nice Meego OBS specialist will be kind enough to help the interested peoples to setup a generic OBS on a other non-ssse3 machine, this will show the reality to everyone without slowing down your so busy severs. Regards, Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 09:53:09PM +0200, Jean-Christian de Rivaz wrote: politic on This was the Meego decision, without the community, to not be based on a upstream distribution to solve this kind of issue. Several of us have warned at that time that this will be a big wast of time. The argument was severely rejected by claim that Meego will be something incredible. It's now maybe the time to lower the expectation and to go back to the reality: without a generic enough build, Meego will never reach any audience like Ubuntu for example. Since when was that _ever_ the goal of MeeGo? politic off Now, I am ok the let my politic thinking above all in a trash and to pragmatically take the actual situation as a fact. So please do the same and show your real facts without escaping into a flameware. I, and others, have pointed out the fact about slowdowns for the system overall (the user's system), and the lack of ability so support such a rebuild system. Are those not good enough facts? If some nice Meego OBS specialist will be kind enough to help the interested peoples to setup a generic OBS on a other non-ssse3 machine, this will show the reality to everyone without slowing down your so busy severs. It's not a server issue at all. You can do this on a sse3 machine just fine. Heck, you could do this today, in the openSUSE build service if you want to. Hey look, someone already is, look at the Smeegol project there, that sounds like what you want. good luck, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 09:20:47PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 05:37:04PM +0200, Jean-Christian de Rivaz wrote: Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? Speed. Seriously, go measure it with it turned off, it is very noticable. And on these tiny netbooks, you need all the speed you can get. There's a reason MeeGo is the fastest booting and running of _all_ distros out there at the moment, and this is one of them. Easy to say, but what real facts can you show to support your claim ? I just got a report from someone today that analized the boot times of all of the currently released distros, and MeeGo was the fastest. The booting process has been analyzed by many peoples the last couple of years. I have not read all of there conclusions, but from what I remember, reducing the I/O activity (in number and in latency) was a far biggest point than the SSSE3 set instruction. Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. thanks, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 09:53:09PM +0200, Jean-Christian de Rivaz wrote: politic on This was the Meego decision, without the community, to not be based on a upstream distribution to solve this kind of issue. Several of us have warned at that time that this will be a big wast of time. The argument was severely rejected by claim that Meego will be something incredible. It's now maybe the time to lower the expectation and to go back to the reality: without a generic enough build, Meego will never reach any audience like Ubuntu for example. Since when was that _ever_ the goal of MeeGo? http://meego.com/about MeeGo currently targets platforms such as netbooks/entry-level desktops, [...] If some nice Meego OBS specialist will be kind enough to help the interested peoples to setup a generic OBS on a other non-ssse3 machine, this will show the reality to everyone without slowing down your so busy severs. It's not a server issue at all. You can do this on a sse3 machine just fine. Heck, you could do this today, in the openSUSE build service if you want to. Hey look, someone already is, look at the Smeegol project there, that sounds like what you want. Did you have an URL ? I cannot find this project on Google nor in Meego search. Aside of that, I notice that Alistair Buxton is currently trying and face some difficulties. Jean-Christia ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 09:20:47PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 05:37:04PM +0200, Jean-Christian de Rivaz wrote: Foster, Dawn M a écrit : On Aug 12, 2010, at 3:42 AM, David Greaves wrote: This is the way open source projects are supposed to work. The people who start the project pick a manageable set of hardware to get us started just like when Linus only supported 386 with AT drives in the first version of the Linux kernel because that's what he was using at the time[1]. [1] http://www.linux.org/people/linus_post.html Sorry, but this argument is completely wrong: At the time Linus started Linux, porting to other hardware was a hug task that involved years of work. Today, every large distribution routinely build generic i686 build. Why not Meego ? Speed. Seriously, go measure it with it turned off, it is very noticable. And on these tiny netbooks, you need all the speed you can get. There's a reason MeeGo is the fastest booting and running of _all_ distros out there at the moment, and this is one of them. Easy to say, but what real facts can you show to support your claim ? I just got a report from someone today that analized the boot times of all of the currently released distros, and MeeGo was the fastest. Please show, this is interesting. The booting process has been analyzed by many peoples the last couple of years. I have not read all of there conclusions, but from what I remember, reducing the I/O activity (in number and in latency) was a far biggest point than the SSSE3 set instruction. Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. Again, you failed to list a single Meego package that will not support a non SSSE3 machine by design. Others kind of accelerated graphics optimization exists aside of the SSSE3 set. Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 10:33:18PM +0200, Jean-Christian de Rivaz wrote: The booting process has been analyzed by many peoples the last couple of years. I have not read all of there conclusions, but from what I remember, reducing the I/O activity (in number and in latency) was a far biggest point than the SSSE3 set instruction. Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. Again, you failed to list a single Meego package that will not support a non SSSE3 machine by design. Others kind of accelerated graphics optimization exists aside of the SSSE3 set. Again, it's not an issue of not supporting, it's an issue of it actually works better. Go rebuild everything without SSSE3 (like the kernel and xorg and xorg libraries) and compare the performance differences. They are very noticeable. thanks, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 10:27:04PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 09:53:09PM +0200, Jean-Christian de Rivaz wrote: politic on This was the Meego decision, without the community, to not be based on a upstream distribution to solve this kind of issue. Several of us have warned at that time that this will be a big wast of time. The argument was severely rejected by claim that Meego will be something incredible. It's now maybe the time to lower the expectation and to go back to the reality: without a generic enough build, Meego will never reach any audience like Ubuntu for example. Since when was that _ever_ the goal of MeeGo? http://meego.com/about MeeGo currently targets platforms such as netbooks/entry-level desktops, [...] Heh, if you want to look a bit above that line, you forgot to quote: - Performance optimizations and features which enable rich computational and graphically oriented applications and connected services development Netbooks today support this processor. Heck, netbooks from 3 years ago support it just fine as well. I wouldn't recommend anyone run MeeGo on a netbook that doesn't, it's as simple as that. It comes down to what works well, and right now, the current build does for the hardware it was designed for. If you rebuild it for older platforms, it will not perform as well, and you are behind a huge curve to try to fix it up to do so. And still I fail to see why you are trying to compare MeeGo to Ubuntu. If you like Ubuntu, it's there for you to use, no one is forcing you to use MeeGo, right? If some nice Meego OBS specialist will be kind enough to help the interested peoples to setup a generic OBS on a other non-ssse3 machine, this will show the reality to everyone without slowing down your so busy severs. It's not a server issue at all. You can do this on a sse3 machine just fine. Heck, you could do this today, in the openSUSE build service if you want to. Hey look, someone already is, look at the Smeegol project there, that sounds like what you want. Did you have an URL ? I cannot find this project on Google nor in Meego search. It's somewhere on build.opensuse.org. Aside of that, I notice that Alistair Buxton is currently trying and face some difficulties. Yes, that are outside of the MeeGo issue, it is where he is trying to put MeeGo on top of openSUSE-compatible pieces, like NetworkManager. I'm sure he could use help if you are offering to do so. thanks, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 10:27:04PM +0200, Jean-Christian de Rivaz wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 09:53:09PM +0200, Jean-Christian de Rivaz wrote: politic on This was the Meego decision, without the community, to not be based on a upstream distribution to solve this kind of issue. Several of us have warned at that time that this will be a big wast of time. The argument was severely rejected by claim that Meego will be something incredible. It's now maybe the time to lower the expectation and to go back to the reality: without a generic enough build, Meego will never reach any audience like Ubuntu for example. Since when was that _ever_ the goal of MeeGo? http://meego.com/about MeeGo currently targets platforms such as netbooks/entry-level desktops, [...] Heh, if you want to look a bit above that line, you forgot to quote: - Performance optimizations and features which enable rich computational and graphically oriented applications and connected services development Netbooks today support this processor. Heck, netbooks from 3 years ago support it just fine as well. I wouldn't recommend anyone run MeeGo on a netbook that doesn't, it's as simple as that. You failed to understand that there exists recent and advanced CPU that use other optimization set than the Intel SSSE3 specific one. It comes down to what works well, and right now, the current build does for the hardware it was designed for. If you rebuild it for older platforms, it will not perform as well, and you are behind a huge curve to try to fix it up to do so. And still I fail to see why you are trying to compare MeeGo to Ubuntu. If you like Ubuntu, it's there for you to use, no one is forcing you to use MeeGo, right? Yes I was a bit forced since Maemo is replaced by it. With Maemo dev process I have no problem at all running my own applications on my desktop machines. Now with Meego dev process I still don't see a way to get my application running on my machines. How I will do development and testing ? There nothing like scratchbox2 in the Meego way: a running Meego image is alway needed. That's the problem. If some nice Meego OBS specialist will be kind enough to help the interested peoples to setup a generic OBS on a other non-ssse3 machine, this will show the reality to everyone without slowing down your so busy severs. It's not a server issue at all. You can do this on a sse3 machine just fine. Heck, you could do this today, in the openSUSE build service if you want to. Hey look, someone already is, look at the Smeegol project there, that sounds like what you want. Did you have an URL ? I cannot find this project on Google nor in Meego search. It's somewhere on build.opensuse.org. Call me stupid, but I still failed to find it on the build.opensuse.org search and projects list. Aside of that, I notice that Alistair Buxton is currently trying and face some difficulties. Yes, that are outside of the MeeGo issue, it is where he is trying to put MeeGo on top of openSUSE-compatible pieces, like NetworkManager. This is not my understanding of the situation. Please read the end of his last mail. Regards, Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Hi, Excerpts from Jean-Christian de Rivaz's message of Fri Aug 13 21:33:18 +0100 2010: Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. Again, you failed to list a single Meego package that will not support a non SSSE3 machine by design. Others kind of accelerated graphics optimization exists aside of the SSSE3 set. I think all of this is on a bit of a tangent. Nobody in this discussion (as far as I know?) has advocated that doing away with an ssse3-optimised build would be a step forward. Going around in circles over what benefits ssse3 does and doesn't provide isn't helping this debate I think, because at the end of the day, if Intel want this, then they're more than capable of making it happen. As I understood it, The argument is that an additional, non-optimised build should be introduced, so that those using AMD processors (and yes, slightly older Intel processors) can also benefit and use/test/develop MeeGo. reasoning This is especially important in an application developer context: if an application developer can't use MeeGo, and the competition Just Works(tm), then they aren't going to shell out £300 or whatever other insignificant amounts might be required to get new hardware: they'll use the competition. Given that MeeGo is arriving a bit late to this game, I think this is something that can be ill-afforded, personally: issues like this aren't going to help us gain developer mindshare. At all. /reasoning Jean-Christian -- Robin Burchell http://rburchell.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Excerpts from Greg KH's message of Fri Aug 13 22:14:19 +0100 2010: And still I fail to see why you are trying to compare MeeGo to Ubuntu. If you like Ubuntu, it's there for you to use, no one is forcing you to use MeeGo, right? Ah, but there's the rub: if developers can't use it, they can't work on it (even more of a problem with application developers). And I'd say it isn't unfair to theorise that not all of them have the ability (or desire) to buy new hardware to get a new OS working that doesn't (yet) have much of an install base - and not much mindshare as a result. At the end of the day: Do we want developers on MeeGo? Because if we do, we need to minimise the barriers for them. That's the core issue here IMO. thanks, greg k-h -- Robin Burchell http://rburchell.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Greg KH a écrit : On Fri, Aug 13, 2010 at 10:33:18PM +0200, Jean-Christian de Rivaz wrote: The booting process has been analyzed by many peoples the last couple of years. I have not read all of there conclusions, but from what I remember, reducing the I/O activity (in number and in latency) was a far biggest point than the SSSE3 set instruction. Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. Again, you failed to list a single Meego package that will not support a non SSSE3 machine by design. Others kind of accelerated graphics optimization exists aside of the SSSE3 set. Again, it's not an issue of not supporting, it's an issue of it actually works better. Go rebuild everything without SSSE3 (like the kernel and xorg and xorg libraries) and compare the performance differences. They are very noticeable. Sorry but I perfectly live for years with machines that have kernel and xorg stuff without any SSSE3 instruction. And I am pretty sure that I am not the only one. Your position is so strong that this start feeling like Meego is now a Intel marketing tools. I hope I am wrong on this, because I am here from the Maemo history. Regards, Jean-Christian ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 8/13/10 2:53 PM, Robin Burchell virot...@viroteck.net wrote: Hi, Excerpts from Jean-Christian de Rivaz's message of Fri Aug 13 21:33:18 +0100 2010: Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. Again, you failed to list a single Meego package that will not support a non SSSE3 machine by design. Others kind of accelerated graphics optimization exists aside of the SSSE3 set. I think all of this is on a bit of a tangent. Nobody in this discussion (as far as I know?) has advocated that doing away with an ssse3-optimised build would be a step forward. Going around in circles over what benefits ssse3 does and doesn't provide isn't helping this debate I think, because at the end of the day, if Intel want this, then they're more than capable of making it happen. As I understood it, The argument is that an additional, non-optimised build should be introduced, so that those using AMD processors (and yes, slightly older Intel processors) can also benefit and use/test/develop MeeGo. reasoning This is especially important in an application developer context: if an application developer can't use MeeGo, and the competition Just Works(tm), then they aren't going to shell out £300 or whatever other insignificant amounts might be required to get new hardware: they'll use the competition. Given that MeeGo is arriving a bit late to this game, I think this is something that can be ill-afforded, personally: issues like this aren't going to help us gain developer mindshare. At all. /reasoning There's nothing wrong with the reasoning, but the question comes (as Greg KH brought up) is who supports this. Adding a new build for older hardware is far from free. Who's infrastructure does the release build? Who's infrastructure does daily builds? (Building QT alone right now takes 4+ hours in OBS.) Who is going to put forth the validation resources for that build? Who is going to triage the bugs that come in for that build? Who is going to fix the bugs that come in for that build? Those are only the questions that came to mind before I had to pause and think my next sentence. There are many issues with adding another build; the vast majority of which are currently un-scoped and only the least significant (changing the build configuration) has been spoken for. Getting an additional build out in the field is far from a small effort. Ryan smime.p7s Description: S/MIME cryptographic signature ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Fri, Aug 13, 2010 at 10:58:49PM +0100, Robin Burchell wrote: Excerpts from Greg KH's message of Fri Aug 13 22:14:19 +0100 2010: And still I fail to see why you are trying to compare MeeGo to Ubuntu. If you like Ubuntu, it's there for you to use, no one is forcing you to use MeeGo, right? Ah, but there's the rub: if developers can't use it, they can't work on it (even more of a problem with application developers). If you are a developer for MeeGo, and you don't have the minimum requirement for the platform, how can you be a developer for it? You can't test your program out, even if you are an application developer. If you are concerned about the mobile stuff, just run it in a kvm/quemu window, as people have explained. That way you don't need any specific Linux distro or hardware to get up and running. good luck, greg k-h ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 8/13/10 3:04 PM, Jean-Christian de Rivaz j...@eclis.ch wrote: Greg KH a écrit : On Fri, Aug 13, 2010 at 10:33:18PM +0200, Jean-Christian de Rivaz wrote: The booting process has been analyzed by many peoples the last couple of years. I have not read all of there conclusions, but from what I remember, reducing the I/O activity (in number and in latency) was a far biggest point than the SSSE3 set instruction. Agreed, it didn't have much to do about SSE3, but when the accelerated graphics portion takes over, _that_ is the part it shows up. Again, you failed to list a single Meego package that will not support a non SSSE3 machine by design. Others kind of accelerated graphics optimization exists aside of the SSSE3 set. Again, it's not an issue of not supporting, it's an issue of it actually works better. Go rebuild everything without SSSE3 (like the kernel and xorg and xorg libraries) and compare the performance differences. They are very noticeable. Sorry but I perfectly live for years with machines that have kernel and xorg stuff without any SSSE3 instruction. And I am pretty sure that I am not the only one. True, but the kernels you use now undoubtedly use features of the instruction set that are not available on older systems. Would you like us to eliminate those as well? Where is that line of, Which subset of the modern IA32 processor instruction set is considered to be good enough? There is no such thing as generic x86 support. All of the distributions make assumptions on what hardware they support. All distributions have an arbitrary line. Given the above referenced goals for MeeGo, it made sense to require support for more modern processors. As others have stated, there is nothing precluding anyone from using the MeeGo source code and building it with different options. Your position is so strong that this start feeling like Meego is now a Intel marketing tools. I hope I am wrong on this, because I am here from the Maemo history. You mean strong Intel employees like Greg K-H with his suse.de email address? Ryan smime.p7s Description: S/MIME cryptographic signature ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Saturday 14. August 2010 00.04.48 Jean-Christian de Rivaz wrote: Sorry but I perfectly live for years with machines that have kernel and xorg stuff without any SSSE3 instruction. And I am pretty sure that I am not the only one. Including low-powered processors like the Atom, without out-of-order execution? Really? -- Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org Senior Product Manager - Nokia, Qt Development Frameworks PGP/GPG: 0x6EF45358; fingerprint: E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358 signature.asc Description: This is a digitally signed message part. ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 8/13/2010 8:28 PM, Alistair Buxton wrote: The problem is I don't know how to tell OBS to rebuild all the packages without using prebuilt Meego RPMs to do it - all the existing documentation is geared towards someone who wants to build a single package against an already bootstrapped distro. Not wearing my Collabora, Freedesktop, Ubuntu, or Debian hats here, but I've given some thought in the past (and more, as I've seen this discussion rage on) to doing a baseline 686 (or even 586, but gcc's 586 targets aren't as well-tested these days as they used to be) MeeGo port just to help get more community developers on board and the like. Yes, I understand the arguments of others that it will lack certainly whizz-bangery, and be less optimal on my Atom-based netbook, and I don't much care. I can still run it, and I can develop for it, and I can run the optimized version, should that tickle my fancy. I'd need to scare up some resources (perhaps with work, perhaps within the community at large) to make it happen, but from the rampant me-too-ing going on here, it looks like it would be a project that would actually see some use, even by people I work with (hi Robin). Just so people who don't know who I am realize I'm not jumping into a thread to blow smoke up various backsides, I did porting to several architectures in Debian for years, I bootstrapped Ubuntu's LPIA (y'know, before Atom had a name) and ARM ports when I worked at Canonical, and this sort of work is more or less in my blood. And hey, if it would stop my INBOX from being flooded by list mail, I'm all for it. ... Adam (Seriously, though, can I wake up to this thread not having expanded by another 200 mails? My list-reading OCD is going to kill my weekend) ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 12/08/10 00:06, martin brook wrote: Hi, After a lively discussion on #meego (http://mg.pov.lt/meego-irclog/%23meego.2010-08-11.log.html from 21:17) I have created a wiki page to further community developmet of a non SSSE3 Meego build. http://wiki.meego.com/Devices/nonSSSE3 If you can help in any way or have any comments please add your meego nick to the wiki with details or catch up with us in #meego And I've been collecting comments - not all of which are mine and not all of which stand up to scrutiny - but which I think *do* reflect a lot of opinions and general 'buzz'. http://mer-l-in.blogspot.com/2010/08/are-intel-subverting-meegocom.html David -- Don't worry, you'll be fine; I saw it work in a cartoon once... ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 08/12/2010 06:42 AM, David Greaves wrote: On 12/08/10 00:06, martin brook wrote: Hi, After a lively discussion on #meego (http://mg.pov.lt/meego-irclog/%23meego.2010-08-11.log.html from 21:17) I have created a wiki page to further community developmet of a non SSSE3 Meego build. http://wiki.meego.com/Devices/nonSSSE3 If you can help in any way or have any comments please add your meego nick to the wiki with details or catch up with us in #meego And I've been collecting comments - not all of which are mine and not all of which stand up to scrutiny - but which I think *do* reflect a lot of opinions and general 'buzz'. http://mer-l-in.blogspot.com/2010/08/are-intel-subverting-meegocom.html David, Seeing your post is a bit disconcerting. I would completely disagree that there are problems on non-Atom hardware. The specific problem is that a Core 2 or newer is needed. Yes, that still does leave a lot of older hardware out in the cold since they do not support SSE3. It does however, include a lot of hardware that is out there. It did start selling in July of 2006. I understand that you say some of the points on the blog post don't stand up to scrutiny, but it would be really helpful if you pointed out which ones didn't. BTW, this is being sent from a Lenovo T400 running MeeGo which definitely doesn't use an Atom. Ryan ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On Thu, 12 Aug 2010, martin brook wrote: After a lively discussion on #meego (http://mg.pov.lt/meego-irclog/%23meego.2010-08-11.log.html from 21:17) I have created a wiki page to further community developmet of a non SSSE3 Meego build. So, what happens when a vendor ships SSSE3 binaries for MeeGo and the hapless users can't run it because they don't have the right processor? -gabriel ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
On 12 August 2010 20:33, Alistair Buxton a.j.bux...@gmail.com wrote: $ cat /proc/cpuinfo | grep SSSE3 Of course, I meant cat /proc/cpuinfo | grep ssse3 (lower case) -- Alistair Buxton a.j.bux...@gmail.com ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
Re: [MeeGo-dev] non SSSE3 MeeGo
Isn't it possible to build non-ssse3 images with OBS? I don't think they require too much code change, or if any is needed, just different gcc settings. - Fernando ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev
[MeeGo-dev] non SSSE3 MeeGo
Hi, After a lively discussion on #meego (http://mg.pov.lt/meego-irclog/%23meego.2010-08-11.log.html from 21:17) I have created a wiki page to further community developmet of a non SSSE3 Meego build. http://wiki.meego.com/Devices/nonSSSE3 If you can help in any way or have any comments please add your meego nick to the wiki with details or catch up with us in #meego vgrade ___ MeeGo-dev mailing list MeeGo-dev@meego.com http://lists.meego.com/listinfo/meego-dev