[PyOpenGL-Users] FW: Missing functions on MacOSX?
Brought to you by:
mcfletch
Menu
▾
▴
[PyOpenGL-Users] FW: Missing functions on MacOSX?
|
From: Robert K. <r.e...@du...> - 2013-12-12 11:52:32
|
Hi Guys, Further to my previous email, I have now re-installed PyOpenGL and can call the glInitDrawInstanced* functions for ARB and EXT without them seg faulting but they both return False. Is it a case of importing the correct modules to get ARB/EXT methods initialised and if so, which modules? Any thoughts or advice would be greatly appreciated. Thanks very much for your forbearance and patience, Rob On 12/12/2013 10:02, "Robert Kent" <r.e...@du...> wrote: >Hi Mike, > >Thanks (again) for your help. > >I'm glad to know that instanced drawing is the correct method and may >actually be available to me. What is the best way to go about testing the >availability of the ArB/EXT extensions? Calling glIntDrawInstancedARB() >caused a segmentation fault when I tried from the terminalŠ > >Thanks, Rob > >On 11/12/2013 21:36, "Mike C. Fletcher" <mcf...@vr...> wrote: > >>On 13-12-11 12:39 PM, Robert Kent wrote: >>> Hi Mike, >>> >>> Thanks again for your prompt and helpful answers. >>> >>> I'm using PyQt4 as my GUI library and it turns out that, even though >>> you can request a version 3.2 context, it doesn't actually support >>> anything above version 2.1 (a problem that is fixed in Qt5) without a >>> hack and a recompile, which I would rather like to avoid. >>> >>> So, with that in mind, it seems I'm stuck with OpenGL 2.1 for now >>> (until I port our application Qt5 next year). In the meantime could >>> anyone offer some suggestions on how to optimise the following: >>You are only talking to me here (you removed the PyOpenGL-users list >>from the replies a couple emails back :) ). >> >>Before you give up on instanced drawing, can you actually confirm that >>you *don't* have the ARB/EXT extensions available? Even an OpenGL 2.1 >>context should still provide those extensions from the nVidia driver, >>you just have to use them from the alternate namespaces. >> >>Instanced drawing *is* the right way to do this, IMO, >>Mike >> >>> >>> On 10/12/2013 17:53, "Mike C. Fletcher" <mcf...@vr... >>> <mailto:mcf...@vr...>> wrote: >>> >>> On 13-12-10 12:27 PM, Robert Kent wrote: >>> >>> Hi Mike, >>> >>> Thanks very much for your reply. >>> >>> I've discovered that, for some reason, PyOpenGL is using >>> OpenGL 2.1 >>> (acquired from glGetString(GL_VERSION) which results in "2.1 >>> NVIDIA-8.12.47 310.40.00.05f01"), even though my graphics card >>> should >>> support version 3.2 (according to Apple), so that is why the >>> functionality >>> I'm after isn't there. My next question therefore is *why* is >>> PyOpenGL >>> only using version 2.1? Is there some configuration or >>> attributes I need >>> to set or is there something wrong with my core OpenGL >>> installation? >>> >>> Thanks very much again for your help and quick responses, >>> Rob >>> >>> >>> You may need to explicitly request a more advanced context from >>> your GUI >>> library. It is a bit strange that you're seeing the low number on >>>an >>> nVidia driver, however (I'd thought they always provided the >>>highest >>> core driver possible). Are you doing off-screen rendering or >>> something >>> of that nature in your back-end that would force the use of a >>> non-hardware-accelerated path? >>> >>> If you have FreeGLUT installed (I'm guessing it is available for >>> OS-X), >>> you could run the source-package tests/test_glut_fc.py script[1] >>> to see >>> if you get a 3.1 compatible context in a regular rendering >>> window. The >>> strange thing is, OS-X users are normally complaining about >>> *unintentionally* getting core/forward-compatible-only contexts, >>> having >>> the reverse happen is weird. >>> >>> Hope that helps, >>> Mike >>> >>> [1] >>> >>>http://bazaar.launchpad.net/~mcfletch/pyopengl/xml-generation/view/head: >>>/ >>>tests/test_glut_fc.py >>> >>><http://bazaar.launchpad.net/%7Emcfletch/pyopengl/xml-generation/view/he >>>a >>>d:/tests/test_glut_fc.py> >>> >>> >>><http://bazaar.launchpad.net/%7Emcfletch/pyopengl/xml-generation/view/he >>>a >>>d:/tests/test_glut_fc.py> >>> >>> >>> On 10/12/2013 16:27, "Mike C. Fletcher" >>> <mcf...@vr... <mailto:mcf...@vr...>> wrote: >>> >>> On 13-12-10 10:09 AM, Robert Kent wrote: >>> >>> Hi Guys, >>> >>> I'm currently in the process of creating an OpenGL >>> backend for >>> matplotlib as the application I develop has a >>> requirement to plot >>> millions of data points whilst remaining responsive. >>> This was all >>> going very well until I tried to use a couple of >>> functions to allow me >>> to do instanced rendering and I find they are >>> 'missing' from PyOpenGL >>> (or at least my version). When I say 'missing', the >>> reference is there >>> but it is a null method, for example: >>> >>> from OpenGL.GL import * >>> print glDrawArraysInstanced, >>> bool(glDrawArraysInstanced) >>> >>> <OpenGL.platform.baseplatform.glDrawArraysInstanced >>> object at >>> 0x1078fee50> False >>> >>> If I attempt to call this method I get a >>> NullFunctionError. Is there >>> something wrong with my setup/installation or perhaps >>> I am missing >>> something? I would really appreciate any help with >>> this. My >>> configuration is as follows: >>> >>> MacBook Pro (2.6GHz i7, 8GB RAM) >>> MacOSX 10.8 Mountain Lion >>> Python 2.7.2 >>> PyOpenGL 3.1.0a3 >>> >>> Afraid I don't have any OS-X machines on which to test, >>> but the most >>> likely cause of this would be having the function actually >>>not >>> supported, or only supported via an extension. >>> glDrawArraysInstanced was >>> added to core in OpenGL 3.1, and exists in the ARB, EXT >>> and NV specific >>> extensions as well. That is, you should be able to do >>> this (once the >>> context is created, as the entry points are always null >>> until the >>> context exists): >>> >>> glDrawArraysInstanced = ( >>> OpenGL.GL.glDrawArraysInstanced or >>> OpenGL.ARB.draw_instanced.glDrawArraysInstanced >>>or >>> OpenGL.EXT.draw_instanced.glDrawArraysInstanced >>>or >>> OpenGL.NV.draw_instanced.glDrawArraysInstanced >>> ) >>> >>> or use OpenGL.extensions.alternate() to wrap the functions >>> to use the >>> first available one and not have to wait for context >>> creation. If your >>> context *is* OpenGL 3.1+ and doesn't have >>> glDrawArraysInstanced then >>> something is broken. I test glDrawArraysInstanced in an >>> OpenGLContext >>> test, so the function itself should be working, at least >>> on Linux/GLX. >>> >>> HTH, >>> Mike >>> >>> -- >>> ________________________________________________ >>> Mike C. Fletcher >>> Designer, VR Plumber, Coder >>> http://www.vrplumber.com >>> http://blog.vrplumber.com >>> >>> >>> >>>------------------------------------------------------------------------ >>>- >>>- >>> ---- >>> Rapidly troubleshoot problems before they affect your >>> business. Most IT >>> organizations don't have a clear picture of how >>> application performance >>> affects their revenue. With AppDynamics, you get 100% >>> visibility into >>> your >>> Java,.NET, & PHP application. Start your 15-day FREE TRIAL >>> of AppDynamics >>> Pro! >>> >>>http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clk >>>t >>>r >>> k >>> _______________________________________________ >>> PyOpenGL Homepage >>> http://pyopengl.sourceforge.net >>> _______________________________________________ >>> PyOpenGL-Users mailing list >>> PyO...@li... >>> <mailto:PyO...@li...> >>> https://lists.sourceforge.net/lists/listinfo/pyopengl-users >>> >>> >>> The University of Dundee is a registered Scottish Charity, No: >>> SC015096 >>> >>> >>> >>> -- >>> ________________________________________________ >>> Mike C. Fletcher >>> Designer, VR Plumber, Coder >>> http://www.vrplumber.com >>> http://blog.vrplumber.com >>> >>> >>> >>> The University of Dundee is a registered Scottish Charity, No: SC015096 >> >> >>-- >>________________________________________________ >> Mike C. Fletcher >> Designer, VR Plumber, Coder >> http://www.vrplumber.com >> http://blog.vrplumber.com >> > The University of Dundee is a registered Scottish Charity, No: SC015096 |