File di log originale · 20190302-7379.txt
------------------------------------------------------------------------
r7379 | terminx | 2019-03-02 15:22:12 -0800 (Sat, 02 Mar 2019) | 3 lines
A curious Polymost optimization from Nuke.YKT
This bails out of drawing any domost spans that seem to be entirely outside of the range of the screen coordinates. Since this seems so obvious, I have to wonder if Polymost is supposed to be doing this elsewhere already...
------------------------------------------------------------------------
r7378 | terminx | 2019-03-02 15:22:08 -0800 (Sat, 02 Mar 2019) | 1 line
Fix CON issues introduced in r7357 and r7358
------------------------------------------------------------------------
r7377 | terminx | 2019-03-02 15:22:04 -0800 (Sat, 02 Mar 2019) | 1 line
Am I missing something here?
------------------------------------------------------------------------
r7376 | terminx | 2019-03-02 15:22:00 -0800 (Sat, 02 Mar 2019) | 1 line
Remove reference pointing to vm.pPlayer because it seems incompatible with the new dispatch method
------------------------------------------------------------------------
r7375 | terminx | 2019-03-02 15:21:56 -0800 (Sat, 02 Mar 2019) | 1 line
This might not be needed anymore after the previous commit
------------------------------------------------------------------------
r7374 | terminx | 2019-03-02 15:21:52 -0800 (Sat, 02 Mar 2019) | 1 line
Work around floating point precision issues in Polymost causing problems when rendering certain geometry from particular angles
------------------------------------------------------------------------
r7373 | terminx | 2019-03-02 15:21:48 -0800 (Sat, 02 Mar 2019) | 1 line
Lower MAXWALLSB
------------------------------------------------------------------------
r7372 | terminx | 2019-03-02 15:21:44 -0800 (Sat, 02 Mar 2019) | 1 line
Don't include the system SDL headers on Windows
------------------------------------------------------------------------
r7371 | terminx | 2019-03-02 15:21:41 -0800 (Sat, 02 Mar 2019) | 1 line
Free inthash tables used by the CON compiler using the contents of inttables[] instead of listing them explicitly in C_Compile()
------------------------------------------------------------------------
r7370 | terminx | 2019-03-02 15:21:37 -0800 (Sat, 02 Mar 2019) | 3 lines
Patch from Striker to allow selective tile redefinition in .def files based on the crc of the original tile
This only works for "tilefromtexture" for now.
------------------------------------------------------------------------
r7369 | terminx | 2019-03-02 15:21:32 -0800 (Sat, 02 Mar 2019) | 1 line
Rework previous commit to use a little less floating point math
------------------------------------------------------------------------
r7368 | terminx | 2019-03-02 15:21:29 -0800 (Sat, 02 Mar 2019) | 1 line
Patch from Nuke.YKT to fix voxel rotation in Polymost
------------------------------------------------------------------------
r7367 | terminx | 2019-03-02 15:21:25 -0800 (Sat, 02 Mar 2019) | 1 line
Marginally less sucky version of a function that completely sucks
------------------------------------------------------------------------
r7366 | terminx | 2019-03-02 15:21:21 -0800 (Sat, 02 Mar 2019) | 1 line
Clean up a few uses of the hash table functions to account for hash_findcase() working properly now
------------------------------------------------------------------------
r7365 | terminx | 2019-03-02 15:21:16 -0800 (Sat, 02 Mar 2019) | 1 line
Fix a crash when trying to access cvars through typing the name into the console with the wrong case
------------------------------------------------------------------------
r7364 | terminx | 2019-03-02 15:21:13 -0800 (Sat, 02 Mar 2019) | 3 lines
Use lowercase when computing hashes in hash_getcode()
It turns out hash_findcase() has never worked properly because the correct bucket would never have been searched.
------------------------------------------------------------------------
r7363 | terminx | 2019-03-02 15:21:09 -0800 (Sat, 02 Mar 2019) | 1 line
This was slightly faster here
------------------------------------------------------------------------
r7362 | terminx | 2019-03-02 15:21:05 -0800 (Sat, 02 Mar 2019) | 3 lines
Optimize nonpow2_mhline and nonpow2_thline by replacing divide operation by cheaper multiplication
Patch from Nuke.YKT.
------------------------------------------------------------------------
r7361 | terminx | 2019-03-02 15:21:01 -0800 (Sat, 02 Mar 2019) | 3 lines
Clean up DirectSound driver in audiolib
This includes a small patch from Nuke.YKT as well.
------------------------------------------------------------------------
r7360 | terminx | 2019-03-02 15:20:56 -0800 (Sat, 02 Mar 2019) | 1 line
Remove a "cheap fix" from multivoc that is no longer needed and should never have been needed in the first place
------------------------------------------------------------------------
See http://svn.eduke32.com/listing.php?repname=eduke32 for more details.
Engine
A curious Polymost optimization from Nuke.YKT
This bails out of drawing any domost spans that seem to be entirely outside of the range of the screen coordinates. Since this seems so obvious, I have to wonder if Polymost is supposed to be doing this elsewhere already...
FixCON
Fix CON issues introduced in r7357 and r7358
Am I missing something here?
Remove reference pointing to vm.pPlayer because it seems incompatible with the new dispatch method
This might not be needed anymore after the previous commit
Engine
Work around floating point precision issues in Polymost causing problems when rendering certain geometry from particular angles
Lower MAXWALLSB
Don't include the system SDL headers on Windows
CON
Free inthash tables used by the CON compiler using the contents of inttables[] instead of listing them explicitly in C_Compile()
Patch from Striker to allow selective tile redefinition in .def files based on the crc of the original tile
This only works for "tilefromtexture" for now.
Rework previous commit to use a little less floating point math
FixEngine
Patch from Nuke.YKT to fix voxel rotation in Polymost
Marginally less sucky version of a function that completely sucks
Clean up a few uses of the hash table functions to account for hash_findcase() working properly now
Fix
Fix a crash when trying to access cvars through typing the name into the console with the wrong case
Use lowercase when computing hashes in hash_getcode()
It turns out hash_findcase() has never worked properly because the correct bucket would never have been searched.
This was slightly faster here
Optimize nonpow2_mhline and nonpow2_thline by replacing divide operation by cheaper multiplication
Patch from Nuke.YKT.
Clean up DirectSound driver in audiolib
This includes a small patch from Nuke.YKT as well.
Fix
Remove a "cheap fix" from multivoc that is no longer needed and should never have been needed in the first place