Announcement

Collapse
No announcement yet.

NVIDIA Says It Has No Plans To Support Wayland

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by smitty3268 View Post
    Like 3d rendering apps that the movie studios use?
    How would a 3d rendering app "like movie studios use" have anything to do with the features in vdpau? FYI, those "3d rendering apps" such as Maya are a form of CAD. Most of the 3d rendering apps out there utilize the CPU for actual rendering. SLi also is not a common use for such types apps. In fact it wasn't until much after SLi was introduced to linux that they finally started making Quadro SLi capable.

    I'm guessing they're the ones who wanted color space stuff.
    Well you would be wrong. That was done at the request of HTPC buffs so that they had a proper YCbCr 16-235 range and was done specifically because of this request for the feature.

    http://www.nvnews.net/vbulletin/show...ghlight=16-235

    Comment


    • #32
      Originally posted by drag View Post
      Yes. ION is not for embedded systems. Atom is not for embedded system?

      Wow.
      LMFAO, look up what embedded is.

      If your not even going to try to think then what is the point in engaging in discussions at all? Stop saying moronic crap like this.
      Seriously, take your own advice.

      So your not confusing having GUI tools that edit your Xorg configuration automatically with not actually having a xorg.conf, right?
      This is what happens on my system:

      Code:
      $ cat /etc/X11/*conf
      cat: /etc/X11/*conf: No such file or directory
      How about your nvidia Linux systems?
      [/quote]

      Gee same as my system with a nvidia card.

      Code:
      dean@nemicron:~> cat /etc/X11/*conf
      cat: /etc/X11/*conf: No such file or directory
      dean@nemicron:~>
      Code:
      It's news to me that Nvidia proprietary driver no longer needs Xorg.conf at all anymore. Somebody should let Nvidia know so they can update their documentation. 
      http://us.download.nvidia.com/XFree8...itxconfig.html
      Why would they when it was X11 that made the changes that did not require an xorg.conf. Those instructions are still perfectly valid for older versions.

      Yes. I am not saying that it does not work, I am saying it's stupid that it still exists.
      Why get rid of something that works?

      Yes. I am totally impressed that you have Powermizer. Oh, wait. I don't care what random trademark'd BS Nvidia claims your stuff supports.

      How about Suspend to RAM? Does that work well for you yet?
      Call it p-states or call it PowerMizer, doesn't matter as they are the same things and they work fine. As far as Suspend to RAM, I already told you it did. "STR works, check"

      Thats the stuff I care about. I know that it works ok for some Nvidia users, but it's still crap for others. Especially for most people that don't have new enough card that is supported by the newest drivers.
      FYI I have 10 year old nvidia cards that are working just fine so you don't need anything to damn new. That's the great thing about Nvidia's cards, they are supported well beyond the point where they are still practical to use.

      Comment


      • #33
        Originally posted by smitty3268 View Post
        Like 3d rendering apps that the movie studios use? I'm guessing they're the ones who wanted color space stuff.
        Yay. We have a winner.

        I'll give you a hint:
        The Shrek movies and other major Hollywood films that use 3D graphics were not modelled, edited, and rendered by people running around with Macbooks with Final Cut Pro installed on them.


        A second hint:
        When people are doing protein unraveling, visual simulations, and are working on rendering the data created by the LHC they are not using clusters of Windows machines.

        Think about it. Would anybody here think that would be a good idea to run this:


        http://codingfreak.blogspot.com/2008...-gnulinux.html






        Linux is still unfortunately unpopular on consumer desktop. But it's very popular in many different places.

        Comment


        • #34


          Also I might add depending on your configuration of system even with the FOSS drivers a xorg.conf maybe needed for anything but basic setups for items such as multiple cards, custom modelines, etc.

          Comment


          • #35
            Originally posted by deanjo View Post
            LMFAO, look up what embedded is.
            I guess I'll have to tell the people I worked for half a decade I when I worked with Intel hardware in their equipment it was not really embedded development. They were all excited about Atom boards and finally able to get rid of the crap VIA graphics, too. I'll have to tell them they have to give all that up because the only processors they are allowed to use is ARM ones.

            I see.

            Wait. I now understand. You can't admit when your wrong about something so you just try using insults to cover up your ignorance and lack of experience.

            Seriously, take your own advice.
            Umm... I do.

            I am interested in having a actual discussion. Not to read people spouting moronic bullshit like somebody is keeping score or gives a shit. I don't live my life in some stupid sitcom were they confuse intelligence and comedy with people sharing insults. If you want to actually have a rational discussion then have one.


            It's nice that Nvidia has gotten rid of xorg.conf. So you don't need to edit that at all to get twinview working or anything like that, right? At least they got something right then.

            "STR works, check"
            You were just being too lazy to use words and were not spouting off marketing gibberish. Sorry I was mistaken.

            Hrm.
            http://www.google.com/search?sourcei...8&q=Nvidia+STR

            It appears you have a rather unique manner of using those 3 letter combinations. Oh well. Not a big deal.

            FYI I have 10 year old nvidia cards that are working just fine so you don't need anything to damn new. That's the great thing about Nvidia's cards, they are supported well beyond the point where they are still practical to use.

            Ok.... so your telling me that this 10 year old Nvidia hardware using proprietary drivers suspends to RAM perfectly? Because that is _exactly_ what your trying to tell me if your able to actually follow the discussion.

            Because right now I am just thinking your talking about them being able to load a driver and provide some 3D acceleration, which was not what I was talking about all. I have no doubt at all that the the old driver works with only moderate amount of effort given the fact that I have helped several people in the past get decent 3D acceleration out of their old Nvidia hardware in Linux. Just guessing.

            Comment


            • #36
              Originally posted by deanjo View Post

              Also I might add depending on your configuration of system even with the FOSS drivers a xorg.conf maybe needed for anything but basic setups for items such as multiple cards, custom modelines, etc.

              If the drivers suck they do.


              I can add new modelines all day long on my Intel laptop without having to restart X at all. It makes it much easier to get that fabled 'Pixel Perfect' 720p or 1080p on HDTVs... since every HDTVs I run into invariably are shit when it comes to reporting proper mode setting information. It sucks, but there is no amount of effort video card makers can do to deal with shitty television firmwares.

              If your curious how that works I'll explain it to you.


              I even have a little script I use for when I attach monitors, projetors, and other random devices I happen to run across. Just because I am to lazy to click through a GUI to swap monitors and such.

              Code:
              #! /bin/bash
              case ${1} in
                  both)
                  xrandr --output LVDS1 --auto;
                  xrandr --output VGA1 --auto --above LVDS1 ;
                  ;;
                  laptop)
                  xrandr --output LVDS1 --auto;
                  xrandr --output VGA1 --off ;
                  ;;
                  sideways)
                  xrandr --output LVDS1 --auto;
                  xrandr --output VGA1 --auto --right-of LVDS1 ;
                  ;;
                  vga)
                  xrandr --output LVDS1 --off ;
                  xrandr --output VGA1 --auto;
                  ;;
                  *)
                  echo "both, laptop, vga";
                  ;;
              esac

              I just wish my FN-CRT/LCD button worked. :P

              Comment


              • #37
                Originally posted by drag View Post
                I guess I'll have to tell the people I worked for half a decade I when I worked with Intel hardware in their equipment it was not really embedded development. They were all excited about Atom boards and finally able to get rid of the crap VIA graphics, too. I'll have to tell them they have to give all that up because the only processors they are allowed to use is ARM ones.

                I see.

                Wait. I now understand. You can't admit when your wrong about something so you just try using insults to cover up your ignorance and lack of experience.
                Congrats on working in the industry. A few more years and you will catch up to my experience in embedded systems where they are used for a few dedicated functions not general purpose computing. Please, oh please, show me an embedded ION system device.

                Umm... I do.

                I am interested in having a actual discussion. Not to read people spouting moronic bullshit like somebody is keeping score or gives a shit. I don't live my life in some stupid sitcom were they confuse intelligence and comedy with people sharing insults. If you want to actually have a rational discussion then have one.
                Your doing nothing of the sort. Your just spreading bile with no actual fact. You claim that nvidia doesn't give a shit about the regular user and your wrong. If they had not they would not have put the effort in many of their features.

                It's nice that Nvidia has gotten rid of xorg.conf. So you don't need to edit that at all to get twinview working or anything like that, right? At least they got something right then.
                See this is the perfect example of not knowing what you are talking about. Items like twinview are easily loaded off the .nvidia-settings-rc in your home directory which is created from the nvidia control panel.
                You were just being too lazy to use words and were not spouting off marketing gibberish. Sorry I was mistaken.

                Hrm.
                http://www.google.com/search?sourcei...8&q=Nvidia+STR

                It appears you have a rather unique manner of using those 3 letter combinations. Oh well. Not a big deal.
                STR is a common acronym for Suspend to Ram. Surprising with your embedded experience that you completely missed that.

                http://www.acronymfinder.com/Suspend...%28STR%29.html
                http://www.webopedia.com/TERM/S/Suspend_to_RAM.html
                http://www.auditmypc.com/acronym/str.asp
                http://www.answers.com/topic/str
                http://www.acronymdb.com/acronym/STR

                Ok.... so your telling me that this 10 year old Nvidia hardware using proprietary drivers suspends to RAM perfectly? Because that is _exactly_ what your trying to tell me if your able to actually follow the discussion.
                That is exactly the case. That old Geforce 2 works fine with STR and also even supports xserver 1.9.

                Comment


                • #38
                  Originally posted by drag View Post
                  If the drivers suck they do.


                  I can add new modelines all day long on my Intel laptop without having to restart X at all. It makes it much easier to get that fabled 'Pixel Perfect' 720p or 1080p on HDTVs... since every HDTVs I run into invariably are shit when it comes to reporting proper mode setting information. It sucks, but there is no amount of effort video card makers can do to deal with shitty television firmwares.

                  If your curious how that works I'll explain it to you.
                  No need to explain, I am extremely aware of the shit firmwares on some TV's (some monitors too, I'm looking at you Toshiba and Samsung)

                  I even have a little script I use for when I attach monitors, projetors, and other random devices I happen to run across. Just because I am to lazy to click through a GUI to swap monitors and such.
                  I'm lazy too that's why I have a little batch script to double click on as well using nvidia-settings.

                  Comment


                  • #39
                    Originally posted by deanjo View Post
                    How would a 3d rendering app "like movie studios use" have anything to do with the features in vdpau? FYI, those "3d rendering apps" such as Maya are a form of CAD. Most of the 3d rendering apps out there utilize the CPU for actual rendering.

                    Not rendering with clusters. Actually modelling.. using Linux workstations. They've been doing it for years.

                    And not just modelling. Compositing also.

                    Don't be confused by people that think stuff Final cut pro or Photosho are high end applications... They are mickey mouse stuff compared to a lot of other stuff.


                    And don't try to cover up by saying that 3D modelling is the same as CAD. It's obviously not. Your just making yourself look silly.

                    SLi also is not a common use for such types apps. In fact it wasn't until much after SLi was introduced to linux that they finally started making Quadro SLi capable.

                    Huh. Think about what you just said there.

                    Well you would be wrong. That was done at the request of HTPC buffs so that they had a proper YCbCr 16-235 range and was done specifically because of this request for the feature.

                    http://www.nvnews.net/vbulletin/show...ghlight=16-235

                    Are you sure you have the right thread?

                    All I see there is people developing patches for Mytthv and mplayer to take advantage of features in VDPAU.

                    Were is the feature request for Nvidia? Were is the Nvidia engineer adding support to the driver on the request of folks at the NvNews forums? Is there a bug request or tracking ticket or anything?

                    In your post you made it sound like some Nvidia employees were reading the forums and decided to add it to the hardware drivers. That thread was about improving the VDPAU supprot in a couple popular linux applications.

                    Comment


                    • #40
                      Originally posted by drag View Post
                      Not rendering with clusters. Actually modelling.. using Linux workstations. They've been doing it for years.

                      And not just modelling. Compositing also.
                      I am aware of the types of applications you are talking about.
                      Don't be confused by people that think stuff Final cut pro or Photosho are high end applications... They are mickey mouse stuff compared to a lot of other stuff.
                      Never did make that conclusion

                      And don't try to cover up by saying that 3D modelling is the same as CAD. It's obviously not. Your just making yourself look silly.
                      Not trying to cover up at all. Like it or not 3D modelling is CAD in nature. To what levels varies with the application.

                      Huh. Think about what you just said there.
                      So because a feature is found later in Quadro products you assume it was all done for them? SLi was derived from their 3dfx acquirement and was on consumer boards years before quadros became capable. In fact most Quadro software official certified drivers do not give any benefit to the apps that the drivers are optimized for.

                      Are you sure you have the right thread?

                      All I see there is people developing patches for Mytthv and mplayer to take advantage of features in VDPAU.

                      Were is the feature request for Nvidia? Were is the Nvidia engineer adding support to the driver on the request of folks at the NvNews forums? Is there a bug request or tracking ticket or anything?

                      In your post you made it sound like some Nvidia employees were reading the forums and decided to add it to the hardware drivers. That thread was about improving the VDPAU supprot in a couple popular linux applications.
                      Guess you missed this:

                      http://www.nvnews.net/vbulletin/show...53&postcount=7
                      http://www.nvnews.net/vbulletin/show...1&postcount=10

                      Which then appeared in the 260 series drivers

                      Added ColorSpace and ColorRange features for HDMI. These give the ability to output YUV over HDMI and select full/reduced color range on RGB over HDMI. ColorSpace and ColorRange are X Configuration options and can be changed dynamically through nvidia-settings.

                      Comment


                      • #41
                        Your doing nothing of the sort. Your just spreading bile with no actual fact. You claim that nvidia doesn't give a shit about the regular user and your wrong. If they had not they would not have put the effort in many of their features.

                        No, not regular users. I am claiming they don't care much about _linux_desktop_ users. That is Far Far Far from regular users.

                        Say Linux desktop has about 1% market penetration. How many of those doe you think that care about gaming and home theater?

                        50%? 40%? We are talking about fractions of fractions.

                        What possible reason does Nvidia give a shit about that? What major OEMs support Linux with Nvidia consumer hardware?

                        Nvidia cares because it's cheaper to use Linux gamers to debug and Q/A their drivers then it is to pay employees to do it.

                        STR is a common acronym for Suspend to Ram. Surprising with your embedded experience that you completely missed that.
                        Whatever dude. If you used the term "S3" I would known exactly what your talking about and, hell, you would of even saved yourself a keypress.

                        Compare and contrast:

                        http://www.google.com/search?q=s3+site:nvidia.com

                        http://www.google.com/search?q=STR+site:nvidia.com


                        That is exactly the case. That old Geforce 2 works fine with STR and also even supports xserver 1.9.
                        Thats a hell of a lot better then any Geforce 2 running system that I've ran across. Not that I don't believe you or anything.


                        See this is the perfect example of not knowing what you are talking about. Items like twinview are easily loaded off the .nvidia-settings-rc in your home directory which is created from the nvidia control panel
                        I was _asking_ a question. Did you not think I was not looking for a answer?

                        Comment


                        • #42
                          So because a feature is found later in Quadro products you assume it was all done for them? SLi was derived from their 3dfx acquirement and was on consumer boards years before quadros became capable. In fact most Quadro software official certified drivers do not give any benefit to the apps that the drivers are optimized for.

                          No you have it backwards. Sorta.

                          I'll paraphrase (and make myself look really smart, too):

                          ME> Nvidia cares about Linux support because of high-end customers that refuse to use Windows, but are willing to drop huge amounts of cash on their hardware. VDPAU and things like that exist for consumer electronics since Linux is amazingly popular in that market, too.

                          YOU> No your wrong. Nvidia loves us and Linux desktop users generate lots of money for them and is a huge market.


                          <bunch of silly words and everybody being stupid>


                          YOU> SLI is not even originally used on high-end hardware. Hell, It took so long for professional cards that it was not until shortly after Linux had SLI support that Quadro cards gained SLI support.

                          ME> LOL.

                          Comment


                          • #43
                            Originally posted by drag View Post
                            No, not regular users. I am claiming they don't care much about _linux_desktop_ users. That is Far Far Far from regular users.

                            Say Linux desktop has about 1% market penetration. How many of those doe you think that care about gaming and home theater?

                            50%? 40%? We are talking about fractions of fractions.

                            What possible reason does Nvidia give a shit about that? What major OEMs support Linux with Nvidia consumer hardware?

                            Nvidia cares because it's cheaper to use Linux gamers to debug and Q/A their drivers then it is to pay employees to do it.
                            And as I said before features such as vdpau and SLi prove that your wrong. Those are completely "desktop" features.

                            Whatever dude. If you used the term "S3" I would known exactly what your talking about and, hell, you would of even saved yourself a keypress.

                            Compare and contrast:

                            http://www.google.com/search?q=s3+site:nvidia.com

                            http://www.google.com/search?q=STR+site:nvidia.com
                            The reason I didn't say S3 is for the same reason why it pulls up the first results in your search. When talking S3 it is easily confused with S3 graphics when talking about video.

                            Thats a hell of a lot better then any Geforce 2 running system that I've ran across. Not that I don't believe you or anything.
                            Well my extensive experience with nvidia cards in linux shows otherwise. Just because your having bad luck does not mean that it effects everyone. Having setup computer labs in the past with cards starting from Geforce 2+ str has never been a big concern unlike items like PulseAudio that cause issues with str.

                            Comment


                            • #44
                              Originally posted by drag View Post

                              YOU> SLI is not even originally used on high-end hardware. Hell, It took so long for professional cards that it was not until shortly after Linux had SLI support that Quadro cards gained SLI support.

                              ME> LOL.
                              Shortly? Try years after.

                              Comment


                              • #45
                                BTW, Google "S3 (STR)" (with quotes)

                                Comment

                                Working...
                                X