Test build 2014-07-20

[CrystalH]

Right. So both solutions work and I see gl3 used.
Nothing crashes, I still got all settings at max, just AA at 0.
The only difference between vdrift1 is only seen in log.
So, using #version 400, vdrift.exe

Code:

INFO: Profiling summary:
      ai: 0.762368 %
      car: 0.171276 %
      physics: 3.45316 %
      render draw: 50.6365 %
      render setup: 0.891035 %
      render sync: 39.265 %
      scenegraph: 0.107861 %
      sound: 0.0135081 %

Using that #ifdef block, vdrift.exe

Code:

INFO: Profiling summary:
      ai: 1.26186 %
      car: 0.260083 %
      physics: 5.25692 %
      render draw: 32.9256 %
      render setup: 1.18507 %
      render sync: 54.6134 %
      scenegraph: 0.133541 %
      sound: 0.0166932 %

Using #version 400, vdrift1.exe

Code:

INFO: Profiling summary:
      ai: 1.2202 %
      car: 0.236257 %
      physics: 4.83391 %
      render draw: 86.2108 %
      render setup: 0.984074 %
      render sync: 0.103944 %
      scenegraph: 0.447483 %
      sound: 0.0168334 %

Using #ifdef block, vdrift1.exe

Code:

INFO: Profiling summary:
      ai: 1.01938 %
      car: 0.231233 %
      physics: 4.75602 %
      render draw: 85.6062 %
      render setup: 0.939689 %
      render sync: 0.105508 %
      scenegraph: 0.432001 %
      sound: 0.016031 %

I see nothing in profiling txt on screen, since all are microseconds.
Only difference is that vdrift1 spends time in draw and not in sync, if I get this right.
In game looks so: http://imgur.com/yfQHPED
is that ok ? For me contrast seems to be too high, track too bright, shadows too dark and cars somehow dull/dark too.