What changed in EDuke32, build by build

Patch from bioman to fix an issue with static linking GTK and to correct the alignment of the fullscreen checkbox in the startup window when building with POLYMER=0

Run engine source through new version of astyle, add printf() call to OSD_Printf() to facilitate console access through stdin/stdout for dedicated server mode, minor changes to sdlayer to allow operating without actually initializing SDL

don't link build utils to all the libs needed for the actual engine itself

Remove detection of sprites outside of sectors at map load and fix green sprites in Mapster32 with Polymer

Polymer performance improvement and some game related bug fixes. Improves Polymer framerate by about 25% in most cases and fixes problems causing Duke Caribbean doors to malfunction.

fix screwed up movement in Mapster32

Emit RPG_EXPLODE, PIPEBOMB_EXPODE and LASERTRIP_EXPLODE from the actual EXPLOSION2 sprite instead of whatever spawned EXPLOSION2, to prevent the sound from being owned by whatever sprite happens to spawn with the ID the RPG/pipebomb/whatever had before it was removed

Final round (I hope!) of sound fixes. Should also fix the crash on exit on Windows and the REACTOR sprite size bug.

This is mostly just build fixes for USE_OPENGL=0, POLYMOST=0 and POLYMER=0. Also updated the precompiled libogg/libvorbis/libvorbisfile that are supplied for MinGW.

Fix a couple of stupid typos and remove some dead code from Mapster32

Updated synthesis script, moved some non-essential files around, updated license headers on all non-BUILDLIC files, added memory cache on top of texcache to improve load times, moved some duplicated keyboard handling stuff from winlayer/sdlayer to baselayer, fixed keypad / in Mapster32 tile selector, fixed bug where p->rotscrnang and p->look_ang never reset to 0, added support for reloading maphack based polymer lights after mode changes/savegame loads, other minor changes

Fix building engine utils w/ nedmalloc

force r_vbos to 0 for Polymost on ATI R520

More nedmalloc updates, possibly fix win32 mouse death at ~10 fps or less, fix dummytile crash, more untested multiplayer changes... different enet channels for different types of traffic in order to eliminate latency due to the occasional reliable packet getting dropped, therefore causing a hitch in the flow of unreliable movement packets on the same channel, remove mmulti API from engine in favor of having everything in game land

A bunch of crap I meant to commit days ago... gtk startup window patch from bioman, nedmalloc REPLACE_SYSTEM_ALLOCATOR fix (related code uses #ifdef REPLACE_SYSTEM_ALLOCATOR so defining it to 0 was still satisfying the condition), slightly lower CPU usage in Mapster 2D mode, fix an ancient bug causing the player to not make vent footstep sounds on sprite based vents, fix map save state cvars, add "skill" cvar to console, fix saving the game while CON_SHOWVIEW is in use

clean up "renderer setup" menu a bit so we can put polymer shit in it later

Memory cleanups, use Bm/c/realloc/free everywhere (configured as nedmalloc version in compat.h), add large page support for Vista/7, tweak nedmalloc settings, add support for projecting textures from SE50 spotlights placed in maps via sprite owner member (manipulated through CONs via sprite htpicnum member), fix building Mapster32 with MSVC, increase PR_MAXLIGHTS to 1024