* initial OrbitalShuttleTimeMsg packet and tests; new objects to support HART shuttle transport system
* master was stale
* grouped scheduling for timing orbital shuttle activity
* door lock controls for HART shuttle lifecycle, and specifically for the doors that lead into the shuttle boarding hallway
* separation of the door from the door unlocking logic, which now has to be provided if performed by an outside source; a door that is locked either by bolt, HART routine, or other reason, can now be shut immediately; message when HART is not docked with a corresponding entry hallway door
* better degree of door logic control; all shuttle-related messages have been moved to LocalService; careful managing of 'original state' for the shuttle's cycle
* modification of seat mounting and cargo mounting support entities to expand functionality
* absolutely very little to do with the feature of this branch and a lot to do with yak-shaving; long story short, class inheritance is greatly modified and mountable seats can now accept multiple players if initialized properly
* a lot has changed: distribution of MountableBehavior, mount point information is more complex, vehicles convert differently, the routine of the shuttle timer is initialized differently; you can now successfully utilize the HART shuttle to drop into a zone
* swap of shutle from pad to pad control; tests and comments
* eject players from HART gantry hallway as if passengers dismounting from seat when not boarding through the use of environmental geometry; HART system uses duration from config settings to set scheduler
* rebase to curious master; repairs to vector rotation calculations; regression of mountable changes involving seats with occupancy greater than 1; orbital shuttle as a unique vehicle and amenity; corrected dismount offsets and offset calculations; weird angle of nc hart a building has been properly accommodated; hart events have prerequisite animation states
* rebase with master; looks like rebase with merged_master, which is also a commit
* lots of tests (though not nearly enough); checking the permission group of a shuttle seat no longer creates that seat
* fixing explosions
* fixed the persistence monitor service potentially using non-printable unicode in actor names
* can not use a droppod to gain access to one's own sanctuary
* removed hart facility update that causing open bay doors and beeping
* PR review changes
* fix for aggravation issues
* removed suspicious shadowing and clarified failure message; customized skipped bundling
* smp history is now a no-less-efficient circular array
* adjustment to bundle dispatch timing; adjustment to inbound sequence reorder queue
* adjustments to handling inbound packets with missing subslots
* unused PacketCoding features
* comments; delayed start of the queue processor task; turned sequence reorder task and subslot missing task into function literals
* optimizations to the inbound re-order by sequence routines by controlling execution flow
* the subslot request timer has been separated from the standard bundling processor; config values for bundling, sequence resolution, and subslot requests in the middleware actor have been included
* replacing func-array with conditional logic
* difficulty modifiers for repair rate and drain in config
* changed autorepair request-repair pattern to wait until a repair is fulfilled before issuing the next request; moved integration tests for efficiency
* auto-repair accounts for delay between request-reply when schduling next request; base ntu drain is halved; auto-repair mending values adjusted based on wiki times
* moving config call into auto-repair mixin
* deleting the old integration testing code
* obligatory test fixes
* more of the same
* correcting exceptions with calculating subsequent auto-repair time and transfer messaging behavior with WarpGate objects; transfer animation and transfer delivery now have slightly different timing; wrote bunch of tests that don't work in the standard manner, but do pass
* planar classes to describe levels of water and other fluid parallel to the ground
* corrected purpose of field in OxygenStateMessage and adjusted the structure of the packet; the environment is now 'regions filled with stuff'; messaging pathways to facilitate drowning and drown recovery in SessionActor, WorldSession, and PlayerControl, as well as the avatar event system
* drowning height is now a featur - recommend going through GlobalDefinitions; fixed lava pool collision to work on pool entry rather than drown level; lava now burns; painbox damage now is directed towards players control agency first
* drowning timer works correctly for both player and vehicle targets; timing and dive depth information for targets defined, but currently originates from a generic location (ObjectDefinition); packet OSM has been modified for efficiency; classes for environment features previously exclusive to drowning mechanics have been pushed towards generic naming conventions
* added sea and pools for z4, z5, z8, and z10
* vehicles now take damage (to the point of destruction) when exposed to lava due the expansion of environmental damage reasons and environmental damage modifiers; modification of the environment exposure lingo; streamlining of vital activity record system
* added basic drown params to flying vehicle definitions; object trait and control mixin for environment interaction, code moved from SessionActor and WorldSession
* separated environmental classes; handled waterlogged flying vehicles, in properties and code; wrote comments and tests
* players mounting vehicles and players subjected to the vehicle transfer process should receive updated drown-state status of the vehicle; drowning should suspend while in the middle of vehicle transfer, in the case the process is long
* increased damage performed to vehicles by lava
* created base damage interaction classes and replaced various projectile-based damage that utilized ResolvedProjectile; not refined, maintains redundancy and overloads, but should work
* continuing to reduce the exposure of ResolvedProjectile and replacing it with applications of DamageInteraction, DamageResult, and DamageReason
* removed ResolvedProjectile from the project; adjusted remaining code paths to work around it
* vitals.test became vital.base; no one liked this
* lots of inheritance, polymorphism, and other chicanery; moved around files, so it also looks like more files have changed when they have not (even if they did)
* codecov file correction
* master rebase; vital directory structure changed, so file imports have been modified in several other files; ResolutionSelection has been removed, requiring direct function literal assignment; tests repaired, where necessary; no actual functional change
* code comments
* DamageResult is its own case class now, wrapping around a before/after target and the interaction used in its calaculations; tests have been corrected
* adjusted Player.Die() to demonstrate a damage-based suicide approach
* resolved circular inheritance in projectile damage modifiers; better employed explosion reason, damages players around exploding vehicle as example
* expanded explosions to other object types; exploding is now a flag and the damage is an innate property of the object type; removed advanced references to properties on the damage source, since the damage source is easily accessible; wrote comments; fixed tests
* overhaul to painbox damage to align with normal player damage handling, thus assimilating it properly into the damage system
* future development; normal vector from euler angles; custom proximity test
* where 'innateDamage' should have not replaced 'explosion'
* moved the hitPos for the generator test; attempting to imrpove the reliability of the auto-repair integration tests (didn't ...)
* spelling and private val
* numeric session ids now restored; LoginActor knows about connection address; corpses are barren unless searched
* session and login id/counter moved under server; function literal definition changed
* corpse channel
* trying to fix docker as per set-env requirement changes; I don't know what I'm doing
Originally I made list_players return the full avatar
object rather than just the name string. Which was kind of
a bad idea because certain contents were unserializable
which caused the command to break (sometimes).
Weeks later, I attempted to fix it by returning to the
original behaviour of only returning names. By that point
I had forgotten that PSFPortal was changed to expect an
object.
So here's the final fix. player_list now returns a very
simple object made up of the name and faction id.
