Hi GEM maintainers,
i have had some trouble compiling GEM from cvs for Windows (with VC7): - i had to change all occurences of GL_BGRA or GL_BGR to GL_BGRA_EXT and GL_BGR_EXT, respectively (i hope that's exactly the same....) - in text2d.h the GLTTPixmapFont is defined in the class, but the header file "GLTTPixmapFont.h" is only included for linux and OSX. There doesn't seem to be a necessity for that. (and why is FTGL defined for other platforms but for Windows it is GLTT?) - the particle and pspray projects have absolute file positions which makes them practically unusable - in pspray.cpp, line 139 the pow function is neither float, nor double precision... it's necessary to make the constant 0.3333f - there's an interesting warning for glm.c, line 1485 that model->texcoords is of type GLfloat *, which can't easily be printed as an integer.... (should that be numtexcoords?)
the most important problem is that some of the FT*.c of the GLTT 2.5.2 distribution don't seem to match to their corresponding FT*.h files. They also have different dates. Which GLTT version is typically used with GEM?
many thanks, T
On Mon, 21 Jul 2003, Thomas Grill wrote:
Hi GEM maintainers,
i have had some trouble compiling GEM from cvs for Windows (with VC7):
- i had to change all occurences of GL_BGRA or GL_BGR to GL_BGRA_EXT and
GL_BGR_EXT, respectively (i hope that's exactly the same....)
this is really strange. Seems to be a OpenGL thing, but i can not imagine that why this change was done on windows OpenGL implementation.
- in text2d.h the GLTTPixmapFont is defined in the class, but the header
file "GLTTPixmapFont.h" is only included for linux and OSX. There doesn't seem to be a necessity for that. (and why is FTGL defined for other platforms but for Windows it is GLTT?)
Because gem moved to FTGL recently, the windows part has not been worked on yet.
- the particle and pspray projects have absolute file positions which makes
them practically unusable
this is in gemlibs ? the particle library is now included in gem/src/Particle. No need to compile that anymore.
- in pspray.cpp, line 139 the pow function is neither float, nor double
precision... it's necessary to make the constant 0.3333f
I think pspray is a demo for partlib and has nothing to do with gem. So you can ignore it.
- there's an interesting warning for glm.c, line 1485 that model->texcoords
is of type GLfloat *, which can't easily be printed as an integer.... (should that be numtexcoords?)
the most important problem is that some of the FT*.c of the GLTT 2.5.2 distribution don't seem to match to their corresponding FT*.h files. They also have different dates. Which GLTT version is typically used with GEM?
On my system it says 2.5.
Greetings,
Guenter
From: "guenter geiger" geiger@xdv.org
On Mon, 21 Jul 2003, Thomas Grill wrote:
Hi GEM maintainers,
i have had some trouble compiling GEM from cvs for Windows (with VC7):
- i had to change all occurences of GL_BGRA or GL_BGR to GL_BGRA_EXT and
GL_BGR_EXT, respectively (i hope that's exactly the same....)
this is really strange. Seems to be a OpenGL thing, but i can not imagine that why this change was done on windows OpenGL implementation.
Microsoft have basically left the default openGL implementation at version 1.1. This means that even if you're using a card with a driver that offers opengl > 1.1, you're limited to using 1.1 features at compile time unless you include a newer glext.h and to the appropriate runtime checks.
As the windows maintainer of GEM I find myself making that particular change all the time. It should probably be solved with the inclusion of an up-to-date glext.h. Since all Windows openGL implementations supprt GL_BGRA_EXT and GL_BGR_EXT it would not be necessary to check support for these particular extensions.
Thomas: Thanks for pointing out the other issues... most of these are the result of recent changes that I've not yet had time to get going on Windows.
If you want a recent build (just prior to the font system being changed over) please find a binary here: http://www.bogusfront.org
Daniel