* do not skip the player_t init when travelling in a hub. The old player may still be needed in some edge cases. This applies only to singleplayer for now. The multiplayer version still needs reviewing. I left it alone because it may shuffle players around which is not wanted when doing hub travelling.
* do not spawn two temp players in G_FinishTravel. Instead handle the case where no player_t::mo can be found gracefully by adding a few nullptr checks. This temp player served no real purpose except for having a valid pointer. The actual start position was retrieved from somewhere else.
After testing with a savegame on ZDCMP2 which is probably the largest map in existence, timing both methods resulted in a speed difference of less than 40 ms (70 vs 110 ms for reading all sectory, linedefs, sidedefs and objects).
This compares to an overall restoration time, including reloading the level, precaching all textures and setting everything up, of approx. 1.2 s, meaning an increase of 3% of the entire reloading time.
That's simply not worth all the negative side effects that may happen with a method that highly depends on proper code construction.
On the other hand, using random access means that a savegame version change is only needed now when the semantics of a field change, but not if some get added or deleted.
- do not I_Error out in the serializer unless caused by a programming error.
It is better to let the serializer finish, collect all the errors and I_Error out when the game is known to be in a stable enough state to allow unwinding.
- always make the top level object randomaccess when opening a JSON file for reading. Some things won't work right if this is opened for sequential access.
- added sanity checks to prevent a savegame from being loaded with an incompatible map
- refactored a few things to simplify serialization.
- started work on main level serializer function.
Since this list is excluded from regular thinker cleaning, anything that may survive through the end of G_FinishTravel will endlessly multiply and severely break the following savegames or just simply crash on broken pointers.
This reverts commit 5ff0abe568.
- use STAT_INVENTORY only for held items.
Seems this was causing some strange issues with hubs, but for items placed in the world it still cannot be allowed to have them in a different statnum.
This was causing issues with sprite sorting. For this to work as intended, all actors in the world that display sprites need to remain in spawn order, including inventory items.
The only thing this statnum was used for were some bot related search actions which are simply not worth breaking actual maps for some very minor performance gain.
Note that even with this change it is still not possible to unarchive any thinker pointers before the thinker list has been loaded as it would create broken lists.
Note that this required splitting P_SerializeWorld, because sector_t and FSectorPortal contain some actor pointers, for which the same rule applies: Portal linking can only be done after all sectors have been read, meaning it cannot be done along with the rest of the data in these structures.
Obviously such a change breaks savegame compatibility so the min. savegame version had to be increased again.
Two issues:
Portal linking requires all skyboxes in the sectors to be present, without them some info is not there when needed.
UpdateWaterLevel was called in AActor::Serialize, which operated on the freshly loaded level where lots of things haven't been set up yet and plane heights may be wrong.
- Converted P_MovePlayer and all associated variables to floating point because this wasn't working well with a mixture between float and fixed.
Like the angle commit this has just been patched up to compile, the bulk of work is yet to be done.
Patched up everything so that it compiles without errors again. This only addresses code related to some compile error. A large portion of the angle code still uses angle_t and converts back and forth.
- First, don't crash when travelling to a map in a hub that doesn't have
any player starts that match the position given to Teleport_NewMap (or
equivalent).
- Seconed, move the player to the location they were at when the left the
level. At least that way, they shouldn't be in random geometry or off
the map entirely.
This was to resolve some circular dependencies with the portal code.
The most notable changees:
* FTextureID was moved from textures.h to doomtype.h because it is frequently needed in files that don't want to do anything with actual textures.
* split off the parts from p_maputl into a separate header.
* consolidated all blockmap related data into p_blockmap.h
* split off the polyobject parts into po_man.h
Note about A_SkelMissile: The direct change of the missile's position was changed to use SetOrigin. If this is later supposed to be portal-aware, such direct coordinate changes are a no-go to ensure that everything is properly maintained.
'ceilingterrain' is needed because the top of 3D-floors refers to the model sector's ceiling, so in order to give a 3D floor a terrain it must be assignable to the sector's ceiling.
Note that although it is basically the same property, its actual function bears no relevance to its use in Eternity.
- Viewheight change was delayed: Viewheight must be copied to player
structure as well as the PlayerPawn. Not sure if should actually use
deltaviewheight to spread the change out over a few tics instead of
being instant.
- Viewheight not preserved when travelling: player->viewheight must be
restored from pawn->ViewHeight, because the temporary player set it back
to the default viewheight.
- 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.
- 'notranslate != NULL' is completely useless, because 'notranslate' is an array, hence removed.
- I interpreted 'SbarInfoScript != NULL' as a typo, since 1)in the next expression inside the condition there's a dereference to 'SBarInfoScript[SCRIPT_CUSTOM]' and 2)'SBarInfoScript[SCRIPT_CUSTOM]' is checked against 'NULL', in line 352, and then dereferenced when introducing 'cstype'.
- When P_OpenMapData() is called by P_CheckMapData(), we don't actually
care if any required lumps are missing. This just means it isn't a valid
map, so don't abort with I_Error().
major change to ACS's workings. However, I had an epiphany yesterday and just had to do this, since it seems like too big a deal to hold off until a later release:
- Dynamically generated strings returned via strparam and get(user)cvar now last as long as they
need to. They do not disappear at the end of each tic. You can now safely store them in
variables and hold on to them indefinitely. In addition, strings from libraries no longer
require you to load the exact same libraries in the exact same order. You can even store a
library's string in a world variable and retrieve it on another map that doesn't load the
library at all, and it will still be the correct string.
- ACS library IDs now only get 12 bits instead of 16 so that each string table can hold up
to about a million strings instead of just 65536. This shouldn't be a problem, although it
means that save games that had strings with the larger IDs stored in variables are no
longer compatible. Since many saves don't involve libraries at all, and even many that do
are not actually affected, I'm not bumping the min save version. The worst that can happen
is that you get no text at all where some was expected.
SVN r4295 (trunk)
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)
for multiplayer: They now always check through the eyes of every player. For players whose
cameras are not players, they also check through the eyes of those cameras.
- Using spynext/spyprev to switch from a non-player to a player now writes a command to the
network stream and lets Net_DoCommand() take care of it later. The logic here is that if
a player is viewing from something that isn't another player, then every player needs to know
about it for sync purposes. Consequently, when they stop viewing from a non-player and switch
to a player, everybody needs to know about that too. But if they are viewing from a
player, it doesn't matter which player it is, so they can spynext/spyprev all they want
without letting the other players know about it (and without potentially breaking demos--due
to the above-mentioned two codepointers--while doing it during demo playback).
- Replaced the instances of checking players[consoleplayer].camera for a valid pointer to
ones that do it for every player.
- Fixed: Upon changing levels, all players but the consoleplayer would have their cameras NULLed.
- Fixed: player_t::FixPointers() needs to bypass the read barriers, or it won't be able to
do substitutions of old objects that are pending deletion.
SVN r3448 (trunk)
- added DECORATE properties for accuracy and stamina.
- Since these changes move properties from player_t to AActor all savegame compatibility code was removed and the min. savegame version bumped.
SVN r3427 (trunk)
