userGriglercancel

1,517 Commits over 730 Days - 0.09cph!

2 Months Ago
merge from main
2 Months Ago
merge from wakeaiz_optim
2 Months Ago
added additional filter to Any grid query, lets you set an ignore filter for contents that you never care about and can consider the region "empty" for - as well as the usual filter based on gameplay spec (finer distance, player state, etc.) - lets us sleep the WakeAIZ properly while still allowing early-exit on the first found filter-matching player
2 Months Ago
corrected empty grid logic
2 Months Ago
switch to AnyPlayersInSphereFast to inline non-npc + trigger volume checks against players, avoids gathering all players in a huge region to then get a passing check on the 1st one, still provides information on whether the grid regions checked were all empty to allow the trigger to stop ticking
2 Months Ago
additional boat types, missing files
3 Months Ago
DDraw.Arrow passes distanceFade and zTest params to the Line() call correctly (was just repsected by the arrow head)
3 Months Ago
basis for bridge generation, just finding bridge segments on road paths and breaking down into entry/middle/exit sections
3 Months Ago
made ProceduralMapRoadSimple usabel
3 Months Ago
storing "point types" in PathInterpolator to let per-usecase definitions of types along path points - initially only having circular road ring define some path types for BridgeExitEntry and BridgeSegment to have it sit above water instead of bringing terrain up to it
3 Months Ago
added a bunch more boat layouts for testing
3 Months Ago
basis of test to print table of front-on drag of a set of different copypaste boats, going to be used for retuning drag-by-angle values
3 Months Ago
3 Months Ago
drag consistency fix - track generated rays on each drag projection plane, in some situations it could result in generating an extra ray and alter drag values - all tests now passing
3 Months Ago
added test case to retest specific known-failing drag consistency tests - adjust TestFixture to allow it to run non-public test methods so my dodgy BoatRotFailSetup test parameter struct doesn't need to be public
3 Months Ago
merge from parent
3 Months Ago
merge from main
3 Months Ago
client compile fix
3 Months Ago
handling direct transition between different parent entities properly, the coordinate system changes so it needs to reset tracking, also removed some bad state resets - added convar to broadcast ddraw information on dropped item clip checks and collider state to _all_ players (help info makes it clear)
3 Months Ago
merge from main
3 Months Ago
removed CLIENT preprocessor requirement from TestRunner window/pipeline
3 Months Ago
rewrote paste test to monitor linear/angular velocity instead - locked playerboat tests behind just Server define, too many listen-server issues to reliably test it in combined
3 Months Ago
added test to see if pasted boats are positionally stable to check assumption of other tests (they aren't)
3 Months Ago
3 Months Ago
added simple drag-by-angle consistency tests - replaced galleon copypaste data with square/rectangle/triangle simple boat shapes
3 Months Ago
testlist export
3 Months Ago
basic test that boat can be pasted and is floating in water
3 Months Ago
merge from main
3 Months Ago
refactored to ScopedOcean to setup fixture-scoped flat ocean
3 Months Ago
minimum setup required to run a test with ocean
3 Months Ago
merge from droppeditem_improvements (compile fix)
3 Months Ago
merge from droppeditem_improvements (compile fix)
3 Months Ago
compile fixes
3 Months Ago
merge from droppeditem_improvements
3 Months Ago
wrapped DroppedItem RB-removal behaviour behind `droppeditem.remove_rb_on_sleep` convar (defaults off, very experimental) - only changes existing dropped items as they become inactive/active, doesn't go through and remove bodies from anything
3 Months Ago
dropped item improvements with vehicles - item clipping check data stored in LS, behaves properly with boats now instead of falling through all the time - experimenting with destroying/adding RB components instead of setting kinematic
3 Months Ago
kind of hacky BC fix to allow scaling BC without breaking the positions of chains - adjusts LengthModifier by inverse of minimum lossyScale to simulate larger in WS to cancel out the incorrect scale when adjusting transforms in LS with bonestates - need a more legitimate fix for VMs as I hadn't realised how much scale was effecting it
3 Months Ago
merge from wakeaiz_tickrate_optim
3 Months Ago
further decrease wakeaiz tickrate and added jitter, we don't need to tick every fixed update for players entering such giant volumes
3 Months Ago
BC - handle null root chain entries and chains that have no nodes
3 Months Ago
merge from bc_null_chain_fix
3 Months Ago
merge from wakeaiz_tickrate_optim
3 Months Ago
drastically reduce tickrate of WakeAIZ when it becomes active, at worst this will add +10% to the sleepDelaySeconds (usually set to 30s) but will significantly reduce the perf impact of it
3 Months Ago
fix BC serialization issue
3 Months Ago
committing test multichain setup for reference
3 Months Ago
added support to BC for branching chains - rotation is a little iffy, but behaves itself enough to not warrant adding extra constraint control against each child
3 Months Ago
merge from non_trigger_wake_AI - chunky merge, ran s2p on a bunch of monuments
3 Months Ago
s2p for both oilrigs to re-apply fixes from `s2p_particle_flatten_fix`
3 Months Ago
merge from main - kept destination on oilrig prefabs, will rerun s2p
3 Months Ago
refactor to allow subscriptions to nodes of the Grid<>, this lets us sleep the polling when the coarse query grid is completely empty and resume ticking it when it is populated again