OGL can take advantage of this if they want but it won't break
anything if they don't. also state block code has been implemented,
that's totaly internal to D3D no high level code is involved.
Set laser function to only render the effect one sided.
Added some stuff to the credits.
Set D3D fogging to fall back to vertex fog if table fog not supported.
OK, these changes effect a lot of file, I suggest everyone updates ASAP:
Removal of many files from project.
Removal of meanless Gr_bitmap_poly variable.
Removal of glide, directdraw, software modules all links to them, and all code specific to those paths.
Removal of redundant Fred paths that arent needed for Fred OGL.
Have seriously tidied the graphics initialisation code and added generic non standard mode functionality.
Fixed many D3D non standard mode bugs and brought OGL up to the same level.
Removed texture section support for D3D8, voodoo 2 and 3 cards will no longer run under fs2_open in D3D, same goes for any card with a maximum texture size less than 1024.
Simulated speech doesnt say tags marked by $ now.
The following are fixes to issues that came up testing TBP in fs2_open and fred2_open:
Changed vm_vec_mag and parse_tmap to fail gracefully on bad data.
Error now given on missing briefing icon and bad ship normal data.
Solved more species divide by zero error.
Fixed neb cube crash.
Its optional and off by default, use -d3d_particle to activiate.
Also bumped up D3D ambient light setting, it was way too dark.
Its now set to something similar to the original game.
started work on fixing the node model not rendering,
but most of that got commented out so I wouldn't have to deal with it
while mucking about with the polylist
and fixed a small error in the matrix code,
I told you it was indeed suposed to be gr_start_instance_matrix
in g3_done_instance,
g3_start_instance_angles needs to have an gr_ API abstraction version of it made
It doesnt use the backgrounds anymore but its still an improvement.
Currently it fogs to a brighter colour than it should because of Bob specular code.
I will fix this after discussing it with Bob.
Also tided up some D3D stuff, a cmdline variable name and changed a small bit of
the htl code to use the existing D3D engine instead of work around it.
And added extra information in version number on bottom left of frontend screen.
still needs a lot of work, ubt the frame rates were getting with it are incredable
the biggest problem it still has is a bad lightmanegment system, and the zbuffer
doesn't work well with things still getting rendered useing the sofware pipeline, like thrusters,
and weapons, I think these should be modifyed to be sent through hardware,
it would be slightly faster and it would likely fix the problem
also the thruster glow/particle stuff I did is now in.
suport for seperate specular levels on lights and
optional strings for the stars table
code has been made more organised,
though there seems to be a bug in the state selecting code
resulting in the HUD being rendered incorectly
and specmapping failing ocasionaly
useing a phong lighting model I have made specular highlights
that are mapped to the model,
rendering them is still slow, but they look real purdy
also 4 new (untested) comand lines, the XX is a floating point value
-spec_exp XX
the n value, you can set this from 0 to 200 (actualy more than that, but this is the recomended range), it will make the highlights bigger or smaller, defalt is 16.0 so start playing around there
-spec_point XX
-spec_static XX
-spec_tube XX
these are factors for the three diferent types of lights that FS uses, defalt is 1.0,
static is the local stars,
point is weapons/explosions/warp in/outs,
tube is beam weapons,
for thouse of you who think any of these lights are too bright you can configure them you're self for personal satisfaction
Goober5000