Major code/logic cleanups + it has a multithreaded batch mode now.
The VFS code doesn't like multiple threads at all,
so for now there's a big lock in front of where it matters.
Changed precacher, map/tileset, and worldmap loading to batched mode.
Still TODO:
- fix broken mod preview images
- reloading resources on debug key
- make mod recache entirely unnecessary
- actually drop resources when no longer needed
- does no longer depend on collision mask
- remove global Game::lastCollideMaskIndex
- remove the horribly convoluted and now unused RenderObject::getInvRotPosition()
+ Lua func entity_getBoneLockDelay()
+ Lua func entity_setBoneLockDelay()
This takes out the last of the RenderObject mutables.
This also fixes the long-standing editor bug that map tiles
that are once moved to parallax layers and then back to a
normal layer keep their scroll factor until the tile is deleted
or the map saved and reloaded.
This problem might have been introduced in 06270eaac0 but not sure.
The crash could happen due to a use-after-free memory access:
When deleting an entity with skel in postInit(), deletion would be
delayed by 1 frame because that is how entity_delete() works.
During deletion, RenderObject::children were removed and deleted
but SkeletalSprite::bones still had those pointers.
Additionally an object would delete its children in onUpdate() after
safeKill(), and still proceed to run scripts, which would then possibly
access freed memory too.
This fix causes children to not be deleted until we're out of the update()
cycle, and instead delete children together with the root object.
Also removes Lua functions:
- obj_setOverrideRenderPass()
* entity_setRenderPass() is now the same as obj_setRenderPass(),
without the special case alternate name in Lua
Observations:
- Entity::renderPass was never set to RENDER_ALL -> can simplify some things
- The initial pass check in RenderObject::render() was constant for each pass
-> All logic that is per-pass-constant can be moved to a renderability pre-check
- Core::overrideStartLayer, Core::overrideEndLayer, Core::rlayer were never used
- Should be possible eventually to prepare & render layers in parallel
I am not sure if the changes in this commit are 100% correct, but layer passes
are still working and the hug looks like it should.
Thinking about it, the overrideRenderPass functionality should never have existed.
The game scripts don't actually use entity_setRenderPass
(which in turn calls Entity::setOverrideRenderPass())
so I might remove that function in a future commit,
together with the rest of the "override" functionality.
SkeletalSprite was the only thing using it.
Instead of PM_STATIC it's now added to its Entity as PM_NONE.
PM_NONE is not auto-updated so SkeletalSprite::update() is now called
at the end of Entity::onUpdate()
(There's now a bug that during spirit form the screen stays black,
probably added that in an earlier commit. Will fix.)
(On second tought, maybe ParentManaged should be boolean.
Instead add a noAutoUpdate flag. But i'll save that for another time when i get around
to move all RenderObject bools into a bitmask.)
This untangles some of the gigantic kitchen sink headers
in an attempt to split things into smaller files.
Also don't include gl.h, glext.h, windows.h,
and other such nonsense *everywhere*.
Lots of cleanups on the way too. More dead/unused code removal.
Remove incrFlag(), decrFlag() Lua functions.
Another vector that was always empty, optionally filled when specified
by a skeletal file, but none actually did this. Should be safe to remove.
Also small related optimization in Game::collideSkeletalVsCircle().
This was always (0, 0) and optionally used for bones if defined in skeletal files.
None of the official skeletal files used this, so this should be safe to remove.
Should cause less GL pipeline stalling / driver spinlocking if enabled.
Disable AQUARIA_USE_GLM (default: true) in CMake to use the old GL pipeline version.
Based on the implementation by Matt Bierner:
https://bitbucket.org/mattbierner/ios-aquaria
