Announcement

Collapse
No announcement yet.

Direct3D 10/11 Is Now Natively Implemented On Linux!

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

  • Originally posted by BlackStar View Post
    - you can use OpenGL on Aero just fine
    False, only through special implementations. Aero 'shuts off' for half the legacy applications out there, and in some cases, OpenGL.

    Don't see that in Compiz.

    Comment


    • Originally posted by BlackStar View Post
      ICD stands for Installable Client Driver. Which makes your statement an oxymoron.
      So you don't know the difference between what an ICD provides and what DDI provides. It originally tied back into how microsoft wanted to make it more difficult to implement OpenGL on windows.
      Microsoft are allowed to do this of course, but it in no way shows which API is better.

      Comment


      • Originally posted by yotambien View Post
        HOW INTERESTING

        If you go to detailed statistics for that add-on, you'll find:

        a) The average daily users currently amounts to about 400,000
        b) The breakdown by OS for Sep. 2010 shows:

        WINNT: 6,945,962
        Darwin: 1,209,132
        Linux: 1,182,127


        This further convinces me that the effect of changing the user-agent string in browser statistics is absolutely minimal.
        The problem has long been fixed on Microsofts behalf and thus the use agent switcher is no longer needed, so indeed no surprise

        Comment


        • Originally posted by mirv View Post
          So you don't know the difference between what an ICD provides and what DDI provides. It originally tied back into how microsoft wanted to make it more difficult to implement OpenGL on windows.
          Microsoft are allowed to do this of course, but it in no way shows which API is better.
          COME ON, man, libGL works the same way on Linux. Must you always see conspiracy theories everywhere?

          Comment


          • Originally posted by kazetsukai View Post
            They reached it through underhanded deals and by selling software they didn't even have at the time. And they strongarmed their way into the 90%+ desktop share.
            That's what "outmaneuvered" and "outsmarted" refers to. It was a dog-eat-dog situation and they managed to come out on top.

            Also, their software did work (or people wouldn't have used it).

            They're terrified of Linux in alot of ways- its the only software stack enroaching on all of their markets simultaneously. Linux competition is so fierce that they have to fight not just for a companies as a whole- they have to undertake the impossible task of trying to force it out of the different departments of their own customers. Even the majority of Windows-only IT shops have a Linux box sitting around doing a task of one nature or another.

            To stop this competition they have not only given away product, they've subsidized projects that would otherwise involve free software like Linux, and forced their biggest OEMs by offering heavy price reductions on their OS to manufacturers who sell their products exclusively. They have a team dedicated to 'outbreaks' of non-Microsoft products in the OS space, and have specific escalation guidelines for Free Software. At the point of these escalations, they're no longer selling software, they're forcing it on organizations by making deals impossible for other companies to make. This isn't a normal, fair, or sustainable business model, and it will eventually fail.
            While I agree in essense, Microsoft won't be toppled by desktop linux anytime soon. The change is coming from different avenues (Google, Intel/Nokia, Samsung, etc from their mobile appliances).

            Comment


            • Originally posted by kazetsukai View Post
              False, only through special implementations. Aero 'shuts off' for half the legacy applications out there, and in some cases, OpenGL.

              Don't see that in Compiz.
              I happen to be an OpenGL developer and I can ascertain that this is completely wrong. Aero doesn't "shut off" unless (a) you draw to the frontbuffer or (b) you explicitly instruct it so.

              Regurgitating FUD much?

              Comment


              • Originally posted by BlackStar View Post
                COME ON, man, libGL works the same way on Linux. Must you always see conspiracy theories everywhere?
                People have shown you documented areas where microsoft have used various tricks to lock people in to D3D, and reduce the effectiveness of OpenGL on windows, and you're refusing to believe it.
                This was a business choice by microsoft to further their own graphics stack, and get more people locked in to using windows. It didn't make D3D better than OpenGL (not at the start, though by continually pushing it, D3D was made better for a little while). And at the end of the day, the only people really using D3D are game developers; the scientific community still favours OpenGL. Get used to it. Man.

                Comment


                • Originally posted by V!NCENT View Post
                  The problem has long been fixed on Microsofts behalf and thus the use agent switcher is no longer needed, so indeed no surprise
                  So do you agree or not that the number of people playing with the user-agent string is not a valid criticism of web browsing statistics to estimate OS' usage?

                  Comment


                  • Originally posted by BlackStar View Post
                    I happen to be an OpenGL developer and I can ascertain that this is completely wrong. Aero doesn't "shut off" unless (a) you draw to the frontbuffer or (b) you explicitly instruct it so.

                    Regurgitating FUD much?
                    No; a sucking nVidia driver.

                    We've been through this before, remember?

                    Comment


                    • Originally posted by yotambien View Post
                      So do you agree or not that the number of people playing with the user-agent string is not a valid criticism of web browsing statistics to estimate OS' usage?
                      I both agree and do not agree.

                      The web statistics for the years people needed the user agent switch trick do not count. The other years do count. 2010 counts.

                      Comment


                      • Originally posted by V!NCENT View Post
                        I both agree and do not agree.

                        The web statistics for the years people needed the user agent switch trick do not count. The other years do count. 2010 counts.
                        You are arguing that at some point people needed to change the user-agent string because of hotmail, assuming that a sizeable amount of linux users actually used hotmail and made that change (*). But the actual figures (no assumptions here) for that add-on are so low that they couldn't have any major effect at any time. For reference, NetMarketshare deals with 160 million individual users per month. Sorry, it just doesn't hold.

                        Furthermore, if you accept the current figures one wonders where all those users went. Perhaps they got tired of changing their browsers and moved to Windows? Spontaneous combustion? Or they simply didn't exist in the first place?

                        (*) And seriously, what percentage of users would be in that situation? Even taking into consideration that it's not just hotmail but other websites I'm not aware of, is it reasonable to assume that the amount of linux users with browsers reporting a Windows machine is actually higher than those who don't? And not just slightly higher, but 5-10 times higher (to make it up to 5-10% market share).

                        Comment


                        • Hahah, and I just realised that the purpose of the user agent switcher add-on is to easily change on-the-fly your browser identification, i.e. a common scenario would be to browse normally and only activate it when you hit a website that gives you problems (I can't remember any). Even more, the user agent is reset to its default value every time you close the browser.

                          Comment


                          • I know you're trying to be funny, but most people keep their browser open, first thing they do is check their mail and then don't bother anymore. Furthermore 1-something-million hotmail users alone is a massive impact considdering there are only a million-something-Linux users.

                            And because you insist on debating this stupid tiny issue: one particular website != representing the world wide web demograph. So there you have it.

                            Comment


                            • Originally posted by VINCENT
                              I know you're trying to be funny, but most people keep their browser open, first thing they do is check their mail and then don't bother anymore. Furthermore 1-something-million hotmail users alone is a massive impact considdering there are only a million-something-Linux users.

                              And because you insist on debating this stupid tiny issue: one particular website != representing the world wide web demograph. So there you have it.
                              I'm not trying to be funny in any way. I laughed because I had just noticed yet one more basic objection to your argument.

                              Look, you are doing all of the following: assuming stuff with absolutely no basis to do so; making figures up; failing to have a minimum grasp of geography; failing to do basic research before posting; dodging the points you can't deal with.

                              I insist because this is not an stupid issue. I'm amazed at how often lunatic conspiracy theories are used with regards to everything related to Linux, when the simple reality of numbers suffices to explain most of it. This is the difference between logical reasoning and faith, or between honesty and fraud (not related, I don't want to discuss god in Phoronix). And before the off-topic police starts shooting, I invite its officers to read this paragraph again.

                              So after assuming that hotmail is such an important website for linux users, you now assume that "most people keep their browser open, first thing they do is check their mail [which is hotmail, or so the argument goes] and don't bother anymore". Who told you so? Did you personally ask "most people" what do they do? Did you ask "linux users" whether they change their user-agent string? How many of them said yes? See, I could make unfounded assumptions like you do, and say that most people I know who are past their n00b days actually use IMAP/POP3. But I don't, because I simply don't know what "most people" do, and I'm not going to assume anything out of the blue.

                              Secondly, where did you take the "1 million-something-Linux users" from? Have you seriously considered how ridiculous this figure is? At 1% market share and 1M linux users the total number of computer users would be 100M. It doesn't sound right, does it? 822M PCs in use in 2004 and a predicted billion for 2008 say that it doesn't. More stuff: 1,661,000 unique IP addresses collected via Yum for Fedora 13 as of last week; 2,107,000 Fedora hosts voluntary registered to smolt; Ubuntu claiming 12M users some months ago, and so on and so forth. Still think 1M means anything?

                              Lastly, I don't understand your last sentence. Are you talking about the NetMarketshare example I gave?. If you are, I remind you that their collected statistics come from 40,000 different websites across the globe, that they count unique visitors per day to avoid distortions, and that the data is weighted by country based on the international internet traffic as reported, apparently, in the CIA Internet Traffic Count table.

                              Comment


                              • Originally posted by imamdxl View Post
                                Is there any possibility that Direct3D 9 is also implemented by the same way. Because there are thousands of existing Direct3D 9 games that can be played by WINE.
                                I feel like replying to all you D3D9 wishing people. I was the one working on it for fun and I scrapped all the code (because it sucked). Luca inspired me to pick it back up again and I just committed a working trivial/clear case to my personal repo.
                                It doesn't do anything besides clear, but it does work in WINE provided you compile the specific target:
                                Code:
                                scons statetrackers=nine drivers=r300 winsys=radeon targets=libd3d9-wine
                                and pass it the right env vars
                                Code:
                                LD_LIBRARY_PATH=~/Programming/mesa/build/linux-x86-debug/lib/ D3D9_DLL_DRIVERS_DIR=~/Programming/mesa/build/linux-x86-debug/lib/ D3D9_DLL_DEBUG=all NINE_DEBUG=all wine Direct3DDevice.exe
                                Please feel free to try it out or better yet, write some patches

                                Comment

                                Working...
                                X