The code assumed that it had access to the texture manager but that gets initialized after MAPINFO, which means that MAPINFO can only store the texture names and let the precaching code resolve the actual textures.
- Move MUSINFO change request out of FLevelLocals and into player_t. This
allows the MusicChanger actors to change music for each player
independantly. This is similar to PrBoom+, which switches depending on
the displayplayer. The difference being, we don't actually track the
music other players are listening to. (Which might not be a bad idea to
implement at some point.)
- Moved a few fields in player_t for better packing.
- "If it ain't broke, don't fix it."
- Some of the changes were downright wrong and some were pointless, so undo
everything that doesn't look like an actual improvement.
- extended am_drawmapback so that map backgrounds can only be drawn if used with custom colorsets and Raven's, assuming that these sets were specifically made for the accompanying backgrounds.
once in the level struct and then use that for sectors with a NULL skybox. This fixes zpack's
E2M3 so that when it removes its sector stacks, you will get the default skybox in their
place, since stacked sectors and skyboxes use the same pointers in a sector.
SVN r4224 (trunk)
so when you return to it without leaving the hub, it will be as if you had never set foot in
it. RememberState is provided to turn this flag off.
SVN r3846 (trunk)
Z position by the MapThing's Z, just like for any other MapThing.
- P_SpawnPlayer() now respects a player's SPAWNCEILING and SPAWNFLOAT flags.
SVN r3746 (trunk)
- The 'savebuffer' variable still existed?
- Changed AInventory::Destroy to NULL SendItemUse and SendItemDrop if they point to the destroyed object. Although unlikely it can't be ruled out completely that this can happen with delayed CCMDs.
- fixed: Starting a new game did not clear the hub statistics array.
SVN r3034 (trunk)
It will use negative indices for this. Currently supported strings are level name, level lump name and
skill name.
- Extended skill definitions so that printable name and image lump name are separate fields so that a
printable name can be specified for Doom, too.
SVN r2294 (trunk)
- Added an option to parse lumps named ZMAPINFO in place of MAPINFO. Any MAPINFO lumps in files containing
a ZMAPINFO lump will be completely ignored. This is to allow ZDoom specific definitions which are incompatible
with other engines capable of reading MAPINFO. Any ZMAPINFO lump must be in the new MAPINFO format.
SVN r2208 (trunk)
of this:
* A cluster's flat definition can now be preceded by a $ to do a string table lookup.
* Since the standard flat names are now in the LANGUAGE lump, the normal Dehacked substitution
for these is no longer handled specially and so will not be automatically disabled merely
by providing your own MAPINFO.
SVN r2195 (trunk)
that were changed some time ago.
- fixed: The damage inflictor for a rail attack was the shooter, not the puff.
- Fixed: Floor and ceiling huggers may not change their z-velocity when seeking.
- Fixed: UDMF set the secret sector flag before parsing the sector's properties,
resulting in it always being false.
- Renamed sector's oldspecial variable to secretsector to better reflect its
only use.
- Fixed: A_BrainSpit stored as the SpawnShot's target the intended BossTarget,
not itself contrarily to other projectile spawning functions.
A_SpawnFly then used the target for CopyFriendliness, thinking it'll be the
BossEye when in fact it wasn't.
- Added Gez's submission for a DEHACKED hack introduced by Boom.
(using code pointers of the form 'Pointer 0 (x statenumber)'.
- fixed: Attaching 3DMidtex lines by sector tag did not work because lines
were marked by index in the sector's line list but needed to be marked by
line index in the global array.
- fixed: On Linux ZDoom was creating a directory called "~.zdoom" for
save files because of a missing slash.
- fixed: UDMF was unable to read floating point values in exponential format
because the C Mode scanner was missing a definition for them.
- fixed: The recent changes for removing pointer aliasing got the end sequence
info from an incorrect variable. To make this more robust the sequence index
is now stored as a hexadecimal string to avoid storing binary data in a string.
Also moved end sequence lookup from f_finale.cpp to the calling code so that
the proper end sequences can be retrieved for secret exits, too.
SVN r1777 (trunk)