Conference decwet::visual

513.0. "OpenGL, debugger, and Intel/Alpha questions" by NNTPD::"[email protected]" (Ken Miller) Wed May 07 1997 08:57

Hi,

I have a customer I'm supporting who's porting an OpenGL application
to Intel & Alpha NT using Visual C++.  He's presented a few questions which
I hope I can get help answering here.  He's running NT 4.0 and Visual C++ 5.0.

1.  glGetString(GL_Version) returns "1.0 Rel-4.1", yet the compiler macro
    GL_VERSION_1_1 is set (to 1) and the release notes clearly indicate the
    OpenGL that comes with Visual C++ 5.0 is Version 1.1 .  This implies 1.1
    header files but a 1.0 renderer, no?  What is the correct version of
OpenGL
    that ships with Visual C++ 5.0 under NT 4.0?

2.  In his application, he's seeing some pointer variables represented in
    the debugger without a corresponding "+" sign to allow him to expand (e.g.
    display) the contents of these pointers.  Why would that be?  He's not yet
    seen a pattern to why some pointer variables contents are viewable while
    others are not.

3.  When this customer copies his projects back to an Intel box running 4.0
and
    Visual C++, the files do not appear as "compileable" (e.g. there is no 
    option available to compile.  What would cause this?

Thanks in advance!

Regards,

           == ken miller ==
----------------------------------------------------------
 [email protected]  http://www.mindspring.com/~kenmiller
  Digital (Metro DC Area)   FGR Engineering, COP Facility
  PGP  '96 1F 82 38 7A 38 6C ED  E5 9C 81 D2 D2 74 87 04'

[Posted by WWW Notes gateway]

    Miller -- 
    
    You are correct -- glGetString(GL_Version)
    returns the version supported by the renderer (graphics card),
    not the version compiled against.
    
    The PowerStorm 4D40T, 4D50T and 4D60T support OpenGL V1.1.
    Older cards (ZLXp series and the PowerStorm 4D20) only support
    OpenGL V1.0.

Hi,

Thanks CHALTAS for clarifying the graphics card functionality in .1 .  I've
also
learned that the problem in (3) (e.g. alpha --> intel) was the lack of a
config
specifically for Intel.  This only leaves (2):

2.  In his application, he's seeing some pointer variables represented in
    the debugger without a corresponding "+" sign to allow him to expand (e.g.
    display) the contents of these pointers.  Why would that be?  He's not yet
    seen a pattern to why some pointer variables contents are viewable while
    others are not.

Any further ideas?

Thanks in advance!

Regards,

           == ken miller ==
----------------------------------------------------------
 [email protected]  http://www.mindspring.com/~kenmiller
  Digital (Metro DC Area)   FGR Engineering, COP Facility
  PGP  '96 1F 82 38 7A 38 6C ED  E5 9C 81 D2 D2 74 87 04'
[Posted by WWW Notes gateway]

    Here's what was written to me:
    
    It'may be because these pointers are typedef'd as PVOID, however there
    is also a compiler bug with typedefs on Alpha that may be affecting
    this (it should be fixed in the SP). We need an example to tell exactly
    what's going on.