This refactors the ship registry and the OSWPT struct to store data as indexes rather than pointers. This is a prerequisite for making these collections dynamic in the future. See also #5800.
This advances the versions for Clang to 16 and GCC to 13 along with
fixing any warnings these new versions are detecting now.
The new image versions also include Vulkan so the additional workflow
step for installing that is not necessary anymore.
Previously (since 2012!) a waypoint would store its parent list as a pointer. However, this pointer can become invalidated if the waypoint list vector is resized. To prevent this, store the parent list as an index.
This also renames the waypoint's `objnum` field to `m_objnum` to match the naming of the other fields.
In many calls to `ship_get_subsystem_strength`, the caller doesn't care about the actual number, just whether the subsystem type is destroyed. So this adds an optimization to check specifically for that. In the case of `SUBSYSTEM_ENGINE`, this avoids an iteration over every subsystem on the ship which, for a ship with many turrets, can be significant.
Add `const` to the `mc_info` struct so its values aren't inadvertently overwritten. Tweak some of the collision code to prevent overwrites.
Also remove the `flak_muzzle_flash` declaration since that function no longer exists.
For certain calculations involving armor, the `GetDamage` function needs to know if the damage is being applied by a beam so it can apply the calculations with the proper frametime. The parameter was not provided in all cases, so this fixes that.
Follow-up to #1609.
When a non-looping persistent sound expires, `obj_snd_stop()` actually does need to be called. Even though the sound is no longer playing, the `Num_obj_sounds_playing` tally needs to be decremented, otherwise it will eventually exceed its limit and prevent any additional persistent sounds from playing. And it turns out `obj_snd_stop()` returns gracefully for sounds that aren't playing.
Follow-up to #3510. Fixes a bug that appears to have been around since retail.
* fast version of setting position and orientation
* preliminary stale collide time
* add global collision cache stale flag instead
* clean up
* more clean up
* no periods >.>
* bring back single object version
* rename and bring back stale object list
* use a set
1. Add a function to retime *only* the collision pairs involving a specific object. This is useful when only one object has moved, since only collisions involving that object are affected.
2. Use this function when objects are moved for autopilot, via sexp, and via script.
Previously collision pairs were not retimed when objects were moved via script, leading to bugs. This fixes that.
Take into account the collision flag and collision groups when spawning a ship. Should fix#5357.
Also clean up the code a bit, including setting a few `int`s to `bool`s and adding a missing axis case.
This initializes `mc_info` to the correct values when it is constructed, removing the need for `mc_info_init`.
This PR also fixes a group of incorrectly inverted assignments in beam.cpp.
Follow-up to #5040. This extends the checks to most iterations over `obj_used_list` in FSO. There are a few exceptions such as `waypoint_add` and `waypoint_remove`, plus a few others. And none of the object iterations in FRED have the guard because FRED does not use the same object life cycle as FSO.
When iterating through various object collections, check for the should-be-dead flag to avoid returning objects that are on the verge of being deleted.
Separately, this adds the `isExpiring` function to check for should-be-dead objects.
This PR also makes object iteration consistent across all object types, both to reduce code duplication and to ensure that the length and iteration are done using the same algorithm. Furthermore, if the number of should-be-dead objects is nonzero, the Lua array count for that frame will be different from the internal array count for that frame.
Fixes#5034.
* Fixes CTD with decals on dying ships
If a ship was getting shot with weapons that used decals, the decal code would hit an assertation, because the invalid decals were not being all removed.
This PR updates the decal cleanup loop, to ensure that no decals are missed and all invalid ones are removed. Tests confirm this fixes the issue.
Also cleaned up some comments and moved another assertation before setting of variables.
* Exposes modder useful options to decals
Now that decals are very stable, this PR exposes two options that are very useful for using decals, a random range for the radius along with the option to have a random rotation when the decal is created.
Tested and everything works as expected.
* add script ability to toggle decals
* properly use set arg not func
* fix two minor things in decal shader
* one more cleanup line
* fixes
Goober5000