because PClass::CreateDerivedClass() did not initialize everything before calling
InsertIntoHash().
- Fixed: Forcefully removing a weapon from a player's inventory would not reset
the player's refire counter.
- Fixed: FreeKeySections() was called before M_SaveDefaults() during shutdown,
so all custom keys would go to a "(null)" section instead of their intended
section.
SVN r248 (trunk)
savegames at this point in time.
- Fixed: R_ClearPlanes() did not clear skybox planes for a full clear.
- Streamlined zip file loading by delaying the processing of the local file
header until it is actually needed.
SVN r247 (trunk)
- Fixed: CCMD(dir) passes FStrings directly to Printf.
For R241:
- Fixed: The defaultbind command still treated the bindings as char pointers and
as a result didn't work.
- Added SpawnSpotFacing ACS function which is the same as SpawnSpot but it uses
the map spot's angle.
- Added ThingCountName ACS function which is the same as ThingCount but it takes
an actor's type name instead of a spawn ID.
SVN r242 (trunk)
ceiling and floor of the front sector, the renderer did not add it to the
solid clip list.
- Blends created with the ACS fade commands now degrade to transparent overlays
when the console is visible, just as they do for the menu.
SVN r240 (trunk)
using duplicate DoomEdNums and will affect all attempts to spawn the
replaced actor. However, because this happens for all spawns and not just
at map load, the replacing actor must be compatible with the replaced
actor, which means that an actor can only serve as a replacement for one
of its baseclasses. For example, if you want to use a modified imp, you can
use this DECORATE:
actor MyImp : DoomImp replaces DoompImp
{
// Put changed properties here
}
- New: The IWAD dialog now remembers the last IWAD you picked and
automatically highlights it the next time you run the game. This also
applies if you check "Don't ask me this again": The IWAD selected will be
the one that gets automatically loaded, not the one located first. (Using
the -iwad parameter will not change the default IWAD.) In addition, you
can now bring the dialog up even if you disable it by holding down SHIFT
during startup.
- Changed ExtractFilePath() and ExtractFileBase() to return FStrings instead
of writing to a provided output buffer. ExtractFileBase() can also
optionally keep the file's extension in the result.
- Removed the -heapsize parameter entirely. The informational message should
no longer be needed.
- Removed -maxdemo parameter. There's no point to having it around since
the demo buffer grows automatically.
SVN r238 (trunk)
for direct descendants of AWeapon and not for every weapon being defined.
- Changed parsing of actor names back to not use C-mode. This change breaks
any definition that contain periods in their name and apparently there's
more than anyone could expect. Also altered the parser to manually check
for colons inside the parsed string so that placing spaces around them
is no longer necessary.
- Fixed: Weapons could be picked up for ammo even if they gave none.
- Fixed: A_Beacon was missing a NULL pointer check for the beacon's owner.
- Fixed: The status bar tried to access CPlayer->camera without checking
its validity. In spy mode there is a possibility that it is NULL.
SVN r237 (trunk)
- Changed the special radius damage handling for the barrel and boss brain
into an actor flag.
- Added A_RadiusThrust code pointer for DECORATE and adjusted the radius
attack functions accordingly.
SVN r236 (trunk)
yanking the mouse around if they keys haven't been read yet to combat the
same situation that causes the keyboard to return DIERR_NOTACQUIRED in
KeyRead(): The window is sort of in focus and sort of not. User.dll
considers it to be focused and it's drawn as such, but another focused
window is on top of it, and DirectInput doesn't see it as focused.
- Fixed: KeyRead() should handle DIERR_NOTACQUIRED errors the same way it
handles DIERR_INPUTLOST errors. This can happen if our window had the
focus stolen away from it before we tried to acquire the keyboard in
DI_Init2(). Strangely, MouseRead_DI() already did this.
- When a stack overflow occurs, report.txt now only includes the first and
last 16KB of the stack to make it more manageable.
- Limited StreamEditBinary() to the first 64KB of the file to keep it from
taking too long on large dumps.
- And now I know why gathering crash information in the same process that
crashed can be bad: Stack overflows. You get one spare page to play with
when the stack overflows. MiniDumpWriteDump() needs more than that and
causes an access violation when it runs out of leftover stack, silently
terminating the application. Windows XP x64 offers SetThreadStackGuarantee()
to increase this, but that isn't available on anything older, including
32-bit XP. To get around this, a new thread is created to write the mini
dump when the stack overflows.
- Changed A_Burnination() to be closer to Strife's.
- Fixed: When playing back demos, DoAddBot() can be called without an
associated call to SpawnBot(). So if the bot can't spawn, botnum can
go negative, which will cause problems later in DCajunMaster::Main()
when it sees that wanted_botnum (0) is higher than botnum (-1).
- Fixed: Stopping demo recording in multiplayer games should not abruptly
drop the recorder out of the game without notifying the other players.
In fact, there's no reason why it should drop them out of multiplayer at
all.
- Fixed: Earthquakes were unreliable in multiplayer games because
P_PredictPlayer() did not preserve the player's xviewshift.
- Fixed: PlayerIsGone() needs to stop any scripts that belong to the player
who left, in addition to executing disconnect scripts.
- Fixed: APlayerPawn::AddInventory() should also check for a NULL player->mo
in case the player left but somebody still has a reference to their actor.
- Fixed: DDrawFB::PaintToWindow() should simulate proper unlocking behavior
and set Buffer to NULL.
- Improved feedback for network game initialization with the console ticker.
- Moved i_net.cpp and i_net.h out of sdl/ and win32/ and into the main source
directory. They are identical, so keeping two copies of them is bad.
- Fixed: (At least with Creative's driver's,) EAX settings are global and not
per-application. So if you play a multiplayer ZDoom game on one computer
(or even another EAX-using application), ZDoom needs to restore the
environment when it regains focus.
- Maybe fixed: (See http://forum.zdoom.org/potato.php?t=10689) Apparently,
PacketGet can receive ECONNRESET from nodes that aren't in the game. It
should be safe to just ignore these packets.
- Fixed: PlayerIsGone() should set the gone player's camera to NULL in case
the player who left was player 0. This is because if a remaining player
receives a "recoverable" error, they will become player 0. Once that happens,
they game will try to update sounds through their camera and crash in
FMODSoundRenderer::UpdateListener() because the zones array is now NULL.
G_NewInit() should also clear all the player structures.
SVN r233 (trunk)
the resource script so that Developer Studio won't replace the version macros
in it with their literal values. Now updating version.h is guaranteed to be
enough to bump the version everywhere it's used.
SVN r225 (trunk)
longer used.
- Fixed: strifehelp.acs and the PUMPUPS cheat need to give 10 UpgradeStaminas
because that item is using Inventory::Amount now.
SVN r220 (trunk)
- Went back to a single process model for crash reporting. It appears there
are some machines that don't like having another process collect the
information and report it. I still feel that having another process do
it _should_ be more reliable, because that process can execute in a known
good state. Ah well. I haven't yet seen anything that left the main process
in a totally unrecoverable state, so it's probably okay. Unlike before, this
single-process version is still GCC-compatible.
SVN r217 (trunk)
didn't work properly. It has to be done in the cheat code, not in
APlayerPawn::AddInventory.
- Fixed: The medikit and stimpack used a MaxAmount of 100 so that
stamina upgrades were ineffective.
SVN r214 (trunk)
discards the jumping bits.
- ProcessActor() now sets C mode before retrieving the actor's name.
- Fixed: The new SC_GetString() scanner accepted slashes at the end of tokens
when not in C mode, even if they were the start of a comment. Now if you
want a slash at the end, you must quote it.
SVN r211 (trunk)
contained unused sidedefs.
- Fixed: Color control sequences were written to the log file. Since any
entered console command contains such a sequence it was quite noticable.
SVN r208 (trunk)
while and killing centaurs with the flechette.
- Fixed: Moving to an old level in a hub caused the old player's inventory to
spawn owned by the current player (but still hanging off the old player), so
the game would hang when trying to delete it.
- Modified re2c so that it doesn't add a date to the file it generates. Thus,
if it regenerates a file during a full rebuild, SVN won't see it as a change.
Also updated it to 0.10.5.
- Fixed: SC_GetString() did not properly terminate sc_String when the last
token in the file had no white space after it. Since I could not actually
find the problem (it works fine in debug mode and I saw no logic errors),
I decided to take this opportunity to reimplement it using an re2c-generated
scanner. Now it's 1.6x faster than before and correctness is easier to
verify.
- Fixed: FMODSoundRenderer::Shutdown() also needs to reset NumChannels.
- Added back the Manifest to zdoom.rc for non-VC8 Windows compilers.
- Fixed MinGW compilation again. Now it uses the same method as Makefile.linux
to find all the source files so that it doesn't need to be manually updated
each time source files are added or removed.
- Added the SVN revision number to the version string. A new tool is used to
obtain this information from the svnversion command and write it into a
header file. If you don't have the svn command line tools installed or didn't
check it out from the repository, you can still build. I added some rules for
this to Makefile.linux, and I assume they work because they do for
Makefile.mingw.
- Fixed: MIDISong2 did not delete MusHeader in its destructor.
SVN r200 (trunk)
- Fixed map name checks in idclev, hxvisit, for +map and the titlemap.
- Changed handling of Zips so that the patches/, graphics/, sounds/ and
music/ subdirectories no longer are placed in the global namespace. Instead
new namespaces are defined. These namespaces aren't merged, however and
searching in them either returns a lump inside it or one from the global
namespace when it doesn't come from a Zip file. Proper order of files is
still observed though. As a result proper use of the directories inside Zips
is strictly enforced now so that for example anything used as a patch must be
in the patches/ directory and won't be found anywhere else.
SVN r199 (trunk)
and aren't limited to the script's activator.
- Added GetSectorLightLevel(tag), GetActorCeilingZ(tid) and
SetActorPosition(tid, x, y, z, fog) ACS functions.
- Fixed: First initialization of camera textures should not mark the rendered
lines as mapped.
SVN r198 (trunk)
respawning in coop. Now the new inventory code should finally be complete. :-)
- Fixed: PROP_Inventory_PickupMessage was improperly defined for non-VC++
compilation.
SVN r197 (trunk)
- Added a PickupMessage property to the internal actor parser, replaced
most of the virtual PickupMessages with it and placed the code that
reads the metadata into AInventory::PickupMessage. Now the
PickupMessage method is truly virtual and I can do:
Added a Health.LowMessage property to define double message items like
Doom's medikit in DECORATE.
- Since defining Mana3 as an ammo type and then overriding the TryPickup
method means that this item defeats all ammo checks in the game it might
as well be defined as a CustomInventory item. At least this fixes the
amount given in easy and very hard skills.
- Converted all ammo items to DECORATE.
- Changed internal property setting of ammo types and sister weapons
to use fuglyname as for DECORATE definitions. This allows to export
the ammo definitions into DECORATE definitions without doing it for
the weapons themselves.
- Replaced obituary methods with actor properties.
- Fixed: The secret map check didn't work for maps inside Zips.
SVN r196 (trunk)
in all the public versions where monster speed is not fixed point, you
couldn't modify the monster's speed due to a bug in P_Move() anyway. So
there's nothing to be backward compatible with.
SVN r190 (trunk)
- Fixed: The Zip loader loaded all WADs inside a Zip into the lump directory.
This is only supposed to be done for WADs in the root directory.
- Complete restructuring of the map loading code. Previously the only way
to put maps into Zips was to load them as embedded WADs which caused
some problems, most importantly that the map's file name was irrelevant
and the internal map label was used instead. With the new code there
is now a properly defined way to add maps to Zips:
* Maps are placed in a subdirectory called 'maps'.
* Maps are stored as WADs that contain all map related lumps.
* The first lump in the map's WAD directory must be the map label.
* All lumps not belonging to the first map are ignored.
* The map's file name determines the name the map is identified with.
For maps stored this way the internal map label is ignored so with this
method renaming maps is as easy as renaming a file and it is no longer
necessary to manipulate the map label.
With the new code it is also possible to load external maps without
adding them to the WAD list. Type 'open mapfile.wad' in the console
to start such a map.
The new code also performs stricter lump name checks to prevent accidental
loading of non-map data.
SVN r188 (trunk)
monster speeds. Now it uses fixed point so that an actor's speed property
can always be considered is always fixed point. So DoSetActorProperty()
should scale very slow speeds, just like dehacked's PatchThing() has done
for some time now.
SVN r187 (trunk)
format. Now it calls FTexture::CreateTexture to do proper checks.
- Removed the PickupSound method from FakeInventory and changed it so that
it uses AInventory::PickupSound to store its custom pickup sound.
- Removed the PickupMessage method from FakeInventory. This can be handled
by the standard pickup message code now that it uses the meta data for the
message.
- Fixed: The maximum indices for StrifeTypes were inconsistent. Now the
allowed range is 0-1000 in all situations.
- Fixed: Setting a local SNDINFO for a map deleted all skin based sounds.
- Added a crouchsprite property to the skin info.
- Fixed: Crouching sprites must be checked each frame, not just each tic.
- Added an srand call to D_DoomMain in order to randomize the values returned
by rand which is being used to shuffle the playlist.
SVN r185 (trunk)
- Fixed: A classic decorate FakeInventory's PickupText was never freed.
- Fixed: Colored lights were never freed.
- Fixed: When a dehacked patch was applied, the dehacked StateMap was never
freed.
- Removed termdone checks around atterm(S_ClearSoundData) and
atterm(S_Shutdown) because atterm() already checks for duplicates.
- Fixed: S_ClearSoundData() should unload all sounds before it clears S_sfx.
- Fixed: AltSoundRenderer::LoadSound() didn't check if the sound had already
been loaded and lost the old sound data if it had been.
- Fixed: FinishDehPatch() needlessly duplicated the new DehackedPickup's name.
- Fixed: PatchStrings() allocated a private string and never freed it.
SVN r183 (trunk)
full path of a file in a Zip.
- Fixed: intermusic in MAPINFO was limited to WAD lumps and couldn't handle
external data.
ACS:
- Fixed: Global and World array symbols didn't initialize their array information.
SVN r182 (trunk)
or not SSE2 is available at runtime. Since most of the time is spent in
ClassifyLine, using SSE2 in just this one function helps the most.
- Nodebuilding is a little faster if we inline PointOnSide.
- Changed FEventTree into a regular binary tree, since there just aren't enough
nodes inserted into it to make a red-black tree worthwhile.
- Added more checks at the start of ClassifyLine so that it has a better chance
of avoiding the more complicated checking, and it seems to have paid off with
a reasonably modest performance boost.
- Added a "vertex map" for ZDBSP's vertex selection. (Think BLOCKMAP for
vertices instead of lines.) On large maps, this can result in a very
significant speed up. (In one particular map, ZDBSP had previously
spent 40% of its time just scanning through all the vertices in the
map. Now the time it spends finding vertices is immeasurable.) On small maps,
this won't make much of a difference, because the number of vertices to search
was so small to begin with.
SVN r173 (trunk)
dialogs.
- Added Thing_Raise special that allows Arch-Vile like resurrections from scripts
or DECORATE states.
- Added a RadiusDamageFactor property for actors. This replaces the hard coded
factor of 0.25 for Hexen's players.
- Added new SpawnProjectile function to ACS. It's the same as Thing_Projectile2
but the projectile is specified by name, not spawn ID.
- Added MAPINFO option to set the compatibility flags. If this is done these
explicit settings will take precedence over the compatflags CVAR.
SVN r164 (trunk)
- Added code to explicitly handle outputting overlapping segs when
building GL nodes with ZDBSP, removing the check that discarded
them early on.
- AddIntersection() should convert to doubles before subtracting the vertex
from the node, not after, to avoid integer overflow. (See cah.wad, MAP12
and MAP13.) A simpler dot product will also suffice for distance calculation.
- Splitters that come too close to a vertex should be avoided. (See cata.wad.)
- Red-Black Tree implementation was broken and colored every node red.
- Moved most of the code for outputting degenerate GL subsectors into another
function.
SVN r160 (trunk)
of its segs thrown away, so the map could not start. This was because the
nodebuilder assumed all subsectors would be 2D and could not handle the
case where a degenerate 1D subsector is created. In this case, that happens
because that map has three polyobjects in the middle of the void, so the only
way to assign them to a subsector is to use a 1D subsector.
SVN r153 (trunk)
were supposed to hate them.
- Since I was editing the file anyway I added checks for Heretic's and Strife's
damaging floor types to DCajunMaster::IsDangerous.
- Added a NULL pointer check to DCajunMaster::TurnToAng because a crash log
indicated that this can happen.
- Fixed: Strife's energy pod contains 20 units when dropped by monsters.
To achieve this I added an Ammo.DropAmount property because there are
no other means to control this from inside a conversation script.
SVN r151 (trunk)
data enough (2 bits instead of 8), so it was super loud and aliased.
- Fixes for GCC 4.1: Several type-punned pointer warnings, but more
importantly, declaring a friend function inside a class body is no longer
enough to declare that function globally; you must declare it again outside
the class.
- Upgraded FArchive::SerializePointer so that it can store 32-bit indices.
- ACS printing pcodes now build their string in an FSttring instead of a fixed
sized buffer on the stack.
SVN r145 (trunk)
re2c instead of "Linking X".
- Updated lemon and re2c to the latest versions and ported dehsupp to use them
for code generation. (Xlatcc is next.)
- Added function level linking for Makefile.mingw.
SVN r144 (trunk)
values to calculate the spawn position.
- Fixed: A_SpawnItem tried to set the spawned actor's angle before checking
the validity of the returned pointer.
SVN r141 (trunk)
- Did some very preliminary support for x64 compiling.
- Now I see why makewad is using the __fastcall calling convention by default:
Because the release zlib is built with it as well.
- Fixed: Conversion from 'const char *' to 'char *' in DEM_SAVEGAME case in
Net_DoCommand().
- Updated the project files for Visual C++ 2005.
SVN r138 (trunk)
be changed is scheduled for destruction.
- Fixed: G_FinishTravel added the player to the hash chain twice.
- Fixed: The animations for Doom's E3 intermission used incorrect patch names.
- Added a check for MF_NOLIFTDROP to PIT_CeilingRaise because the overlapping
bridges in 007LTSD got moved by this function.
SVN r136 (trunk)
- Changed f_finale.cpp/atkstates[] into a static variable, since its
anonymous type prevents it from being accessed from other files anyway.
- Fixed: The behavior of the eventtail advancement in d_net.cpp/CheckAbort()
was compiler-dependant.
- Fixed warnings GCC 4 threw up while compiling re2c and lemon.
- Removed __cdecl from makewad.c again. This is already defined as a builtin
for MinGW, and redefining it produces a warning. (Why is main explicitly
declared __cdecl anyway?)
- Fixed building ccdv-win32 with GCC 4. GCC 4 creates a memcpy call, which
won't work because it doesn't get linked with the standard C library.
SVN r135 (trunk)
- New: Pausing the game (through any means, not just the pause key) now pauses
sound effects as well as music. "PauseMusicInMenus" has been added as a
MAPINFO flag to also pause the music when a menu or the console are open.
SVN r134 (trunk)
- Changed type PClass::FreeIndices to TArray<unsigned int> because that's
the type of the indices.
- Fixed: makewad.c needs __cdecl for non-Windows builds.
- Fixed: FinishThingdef didn't check whether the WeaponClass pointer in
AWeaponPiece was a valid name.
SVN r133 (trunk)
used ZatPointDist instead of PointToDist to calculate the destination
distance for the plane equation.
- Fixed: Sky scroll positions could "jump" once they wrapped past position
32767. An fmod now keeps them within range of the sky textures' real widths.
SVN r132 (trunk)
This prevents the front ceiling from being moved up if there has been no
texture defined to draw on the wall. This makes various parts of The
Darkest Hour render properly again while still allowing the crosses in
heroes2.wad and lamasery.wad to render correctly.
SVN r130 (trunk)
a crash report indicating that an actor being pushed up by a moving floor
had a NULL sector. Since this field should be valid for every actor, the
debug build gets an assert here, and the release build just returns without
doing anything.
- Fixed: Camera textures were not rendered properly when the underlying
canvas's pitch and width were different (which, really, only happens if
you use ridiculously large camera textures).
- Fixed: FCanvasTextureInfo's were never freed.
- Fixed: MAPINFO special action structures were not freed.
- Fixed: FSingleLumpFont::LoadFON2() never freed its widths2 array.
SVN r129 (trunk)
incorrect path was created for voices.wad.
- Fixed: The code building the path for Strife's voices.wad called the
wrong FString constructor.
- Fixed: LS_Thing_SetGoal set MF5_CHASEGOAL in flags, not flags5.
SVN r128 (trunk)
- Fixed: Any touching_sectorlists for actors unlinked in G_StartTravel() were
lost forever.
- Fixed: DLightningThinker::Serialize() did not delete the old
LightningLightLevels array when loading from an archive.
- Fixed: Although I moved the correct polyobject freeing code into
P_FreeLevelData(), I left the old wrong code there too, which just deleted
the array without deleting anything hanging off of it.
SVN r127 (trunk)
maps). To use it, compile the ACS files as ordinary libraries placed
between A_START/A_END markers. Then outside the markers, create a lump
called LOADACS. This is just a plain text lump that lists all the libraries
you want to autoload with every map. You can do this with as many libraries
as you want, and LOADACS lumps are also cummulative.
SVN r123 (trunk)
calculating a non-NULL state. When compiling with Visual C++, states are
stored in the defaults list as byte values, but when compiling with GCC,
they are passed as 32-bit arguments to the function directly. So in VC++,
using ~0 to specify a NULL state appears as 255, but in GCC, it appears as
-1.
SVN r121 (trunk)
memcpy to copy the player structures.
- Fixed compilation with MinGW again and removed most of the new warnings.
And following is the log that I forgot to paste in for the previous commit:
- Changed the memory management for FString. Instead of using a garbage
collected heap, it now uses normal heap calls and reference counting to
implement lazy copying. You may now use bitwise operators to move
(but not copy!) FStrings around in memory. This means that the
CopyForTArray template function is gone, since TArrays can now freely
move their contents around without bothering with their specifics.
There is one important caveat, however. It is not acceptable to blindly 0
an FString's contents. This necessitated the creation of a proper
constructor for player_s so that it can be reset without using memset. I
did a quick scan of all memsets in the source and didn't see anything else
with a similar problem, but it's possible I missed something.
- Fixed: Build tiles were never deallocated.
- Fixed: Using Build's palette.dat only got half the palette right.
SVN r117 (trunk)
other more important message types which have lower priority for some
reason.
- Added a menu option for show_messages after accidentally switching it
off far too often and wondering why no messages appear.
- Added crouching DoomPlayer sprites submitted by Enjay.
- Fixed DF_NO_CROUCH was not checked.
- Fixed: The intermission script parser had some of its cases in the
keyword parser incorrectly sorted.
- Fixed: atterm was still defined to take a STACK_ARGS function argument.
- Added an AltFlash state for weapons.
- Turned FloatSpeed into an actor property and changed the value to 5
for all floating Strife actors, as a comment in the source indicated.
(r114 below):
- Added GZDoom's code for Vavoom slope things because I wanted to test
something with a map from Silent Steel.
- Added nocrouch and allowcrouch MAPINFO commands and a DF_NO_CROUCH
dmflag.
- Added GZDoom's crouching code after cleaning it up so that adding
crouching sprites will be easier.
SVN r115 (trunk)
message. When trying to puke a script outside a map it crashed.
- Fixed: The random number generator for large numbers must mask out the
sign bit before performing a modulo.
- Now that the conversation states are pointers there is no need to make
AActor::ConversationAnimation virtual. No class overrides this method
anymore.
- Replaced AMacil1::TakeSpecialDamage with MF5_NODAMAGE.
- Fixed: AMacil2::TakeSpecialDamage and AOracle::TakeSpecialDamage didn't
check whether inflictor was NULL and crashed when used with 'kill monsters'.
- Fixed: Some Strife decorations didn't loop their animation
SVN r113 (trunk)
longer determined by game. Instead there's a new flag, MF5_BLOODSPLATTER
which is deciding what to do. To keep backwards compatibility this flag
is unset for projectiles in Doom and Strife and set for them in Heretic
and Hexen. The same applies to DECORATE but of course the flag can be
manipulated here.
- BLODxx sprites are now globally renamed to BLUDxx when not playing Doom.
This allows using the same states in every game, including the
Raven-specific blood actors.
- Gave the bullet puff and the axe blood masses of 5 so that the make small
splashes.
- Added A_Light(value) code pointer for DECORATE to generalize the weapon
light effect.
- Added 'noskillmenu' option to MAPINFO episode definitions. This is for
WADs that want to implement a skill selection level.
- Added APROP_ChaseGoal and APROP_Frightened actor properties for ACS.
- Added MF5_CHASEGOAL flag that makes monsters to go after their goal even
if they have a valid target.
- Fixed some issues with the changes to P_NewChaseDir I made to include
MBF's dropoff logic.
- Added a PowerFrightener powerup class. It seemed like such a waste to
have this cool feature but no means to use it in a decent fashion.
- Fixed: S_Init and S_ParseSndInfo should call atterm only once but not
each time they are called.
SVN r112 (trunk)
functions added to the exit chain with atterm so that they can be called
in a deterministic order and not whatever order the linker decides to put
them in.
- Fixed: DCajunMaster did not free its getspawned.
- Fixed: P_FreeLevelData() did not free ACS scripts.
- Fixed: Level snapshots were not freed at exit.
- Fixed: The save/load menu list was not freed at exit.
- Fixed: FCompressedMemFile needs a destructor to free the m_ImplodedBuffer.
- Fixed: G_DoLoadGame() did not free the engine string.
- Fixed: M_ReadSaveStrings() did not free the engine string.
- Fixed: Processing DEM_SAVEGAME did not free the pathname string.
- Added a check for truncated flats to FFlatTexture::MakeTexture() because
Heretic's F_SKY1 is only four bytes long.
- Added a dump of the offending state to the "Cannot find state..." diagnostic.
- Fixed: FCompressedFile did not initialize m_Mode in its default constructor.
- Fixed: Heretic and Hexen status bars did not initialize ArtiRefresh.
- Fixed: PNGHandle destructor should use delete[] to free TextChunks.
SVN r111 (trunk)
- Added string replacement option to obituary strings.
- Changed lock and pickup message handling so that it is either a real
string or a LANGUAGE identifier. The only reason I did the mixed format
in the first place were Hexen's lock messages but they have been
replaced by unique strings in the mean time so the feature isn't needed
any more.
- Added a flags parameter to TranslucentLine.
- Extended the second arg of Line_SetIdentification to set not only
ML_ZONEBOUNDARY but all 8 bits of the third byte in the flag word.
This allows a relatively simple means of setting all the new flags
directly.
- Moved ML_ZONEBOUNDARY to doomdata.h so that it is in the same place as the
other line flags.
- Fixed: Strife's teleport swirl didn't loop its animation.
- Fixed: Strife's rat is not supposed to be shootable.
SVN r110 (trunk)
height if the floor had moved while they were there before. This was because
the player was spawned on the original copy of the map before the changes to
it were dearchived, so they didn't know about the new floor height.
- Fixed: Calling BaseFileSearch() and letting it fill in the file's extension
didn't work because the space for the path was deallocated before it
returned.
- Guess we're not leak-free yet. Try travelling around in a hub and see that
it leaks. I don't have time to track it down right now.
SVN r107 (trunk)
merge a lot of these static destructor-only structs into regular
functions added to the exit chain with atterm so that they can be called
in a deterministic order and not whatever order the linker decides to put
them in. (Interestingly, the amount of memory used when repeatedly
executing the same map command at the console varies up and down, but it
now stays relatively stable rather than increasing unbounded.)
- Fixed: The list of resolutions in the video modes menu was not freed
at exit.
- Fixed: mus_playing.name was not freed at exit.
- Fixed: SN_StopAllSequences() should be called at the start of
P_FreeLevelData(), not just before the call to P_SetupLevel() in
G_DoLoadLevel(), so it can run even at exit. And C_FullConsole() can
call P_FreeLevelData() to free more memory too.
- Fixed: StatusBar was not freed at exit.
- Fixed: spritesorter was not freed at exit.
- Fixed: Bad things happened if FString's data pool was destroyed before
all C_RemoveTabCommand() calls were made.
- Added an overload for FArchive << FString.
- Fixed: The players' log text was not freed at exit.
- Fixed: Bot information was not freed at exit.
- Fixed: doomcom was not freed at exit. But since it's always created,
there's no reason why it needs to be allocated from the heap. My guess
is that in the DOS days, the external packet driver was responsible for
allocating doomcom and passed its location with the -net parameter.
- Fixed: FBlockNodes were not freed at exit.
- Fixed: Openings were not freed at exit.
- Fixed: Drawsegs were not freed at exit.
- Fixed: Vissprites were not freed at exit.
- Fixed: Console command history was not freed at exit.
- Fixed: Visplanes were not freed at exit.
- Fixed: Call P_FreeLevelData() at exit.
- Fixed: Channel, SoundCurve, and PlayList in s_sound.cpp were not freed at
exit.
- Fixed: Sound sequences were not freed at exit.
- Fixed: DSeqNode::Serialize() did not resize the m_SequenceChoices array
when loading.
SVN r106 (trunk)
taller than 256 pixels. There was a very slight performance hit for this,
but I was able to tweak mvlineasm4 to make it approximately as fast as
before. Interestingly, maskwallscan manages to be nearly as fast as
wallscan despite having to check every pixel for transparency. I'm
tempted to dump all the old masked rendering code and use
(trans)maskwallscan for everything for the sake of simplicity: Only
two functions to maintain for each render style, and much less
complicated supporting code. Currently, I need five different functions
for each rendering style: One traditional column-at-a-time style like
Doom did it originally, two for rt_draw4cols, and two for transmaskwallscan.
(Right now, I have cheated, and just done the ones that can be used
by walls for transmaskwallscan, so the actual number of different functions
isn't quite so high.) For small textures, such as font characters and
far-away sprites, I'm sure maskwallscan is faster than the current code.
For large textures, it's probably still competitive even if it isn't faster.
But considering how similar wallscan and maskwallscan perform, the
difference is probably pretty minimal, and maskwallscan still might come
out ahead due to its simpler overhead.
SVN r105 (trunk)
string. With strings being passed as names this is unsafe.
- Removed unused parameter types from the function parameter parser for
DECORATE.
- Changed: All actor name parameters in DECORATE are now passed as
FNames, not as strings.
- Fixed: The MAPINFO parser stored the RedirectType as a type pointer.
But at this point DECORATE hasn't been read yet so this was limited to
the internal classes.
- Fixed: TXT_NEED_IDCARD wasn't terminated with a ';'.
- Fixed: Strife's DeadRebel was missing its DoomEdNum.
- With names as type identifiers it is no longer necessary to remap
the monster types to internal constants in A_BossDeath.
- Fixed: A_BossDeath got the string from a name - just to get a name from
the string. Using the name directly is sufficient.
SVN r103 (trunk)
Valgrind seems to say so.
- Fixed: The FWadCollection destructor needs to use free to free the LumpInfo
and Wads arrays.
- Fixed: The ColorMapKiller needs to use the delete[] form of delete.
- Fixed: FConfigFile::ClearCurrentSection() should be calling the delete[] form
of delete to free the entry.
- Fixed: FPatchTexture::MakeTexture() does not need to blindly recreate the
Spans if they already exist.
- Fixed: The FMultiPatchTexture destructor did not call its Unload() method.
- Restored the original padding calculation to FMultiPatchTexture::MakeTexture().
I believe the Valgrind errors were caused by accessing off the end of the
screen buffer, not from accessing off the end of a texture.
SVN r98 (trunk)
