June

Monday 1st

• Cleaning up May notes
• Updates dependencies in Ephemeris-Notes build/publish action
• Adding IsSpectatorPlayer tag to player entities
• Researching laser damage physics: Lasers

Tuesday 2nd

• More laser research
  • Airy disc size
  • Jitter
  • Game system:
    • Measure range to target
    • Calculate $M^2$, partly from initial ship design, partly from runtime modifiers (e.g. thermal stress)
    • Calculate airy disc spot size at that range
    • Apply jitter
    • Use that spot size to calculate damage
• Developing orbit editor gizmo
  • Extracting orbit sampling jobs from KeplerBodyLineRenderer
  • Fixed bug in kepler sampling job
    • Same bug is duplicated in KeplerBodyLineRenderer!
    • Porting KeplerBodyLineRenderer to use extracted code, instead of keeping a copy

Wednesday 3rd

• Unit tests for new kepler sampling jobs extracted from line renderer
• Using polyline rendering for drawing orbit flattened into XZ plane
  • Debugging why it's not rendering at all
    • Weird ztest settings
  • Need to add support for closed polylines
  • Thickness can't go narrower than the base mesh
  • Adding job to normalise polyline distances
  • Removed polyline, doesn't really help with understanding orbit and can be confusing
  • Adding ground track
    • Rotations are weird?
      • Kepler orbits calculate in swizzled coordinate frame (XYZ)
      • Using plane normal instead
  • Colouring orbit based on Y coordinate (above/below plane)
    • Adding more points into the line either side of the zero crossing, for a sharper colour transition

Thursday 4th

• More colours
• Fixing transparency on flattened orbit (polylines in general)
• Experimenting with lines attached to periapsis
  • Dropline (periapsis to plane)
  • Droparc (periapsis to plane)
  • Arc from periapsis to ascending/descending node (whichever is closer)
• Refactoring prototypes
  • Split settings into sensible modules
  • Share all orbit related calculations

Friday 5th

• Changing all lines that are not in the orbit plane to grey, subtle hint on plane alignment colour == in plane.
• Adding markers at the foci
• Drop arcs from latus rectum points to plane
• Simplifying colours scheme:
  • Flat plane guidelines
  • Orbit plane guidelines
  • Out of plane guidelines
• Adding support for open orbits (parabola/hyperbola)
  • New method for generating orbit points, sweeping over true anomaly range instead of time
  • All the guidelines are broken
  • Added a new pole line through planet
  • Not drawing both ascending/descending nodes if the orbit only crosses the plane once
  • Not drawing apoapsis lines on open orbit
  • Added back in ring drawing, ignoring rings that don't make sense
• Added in Luna prefab, rings are wrong scale?
• Adding helpers for drawing arrows
  • Straight
  • Curved

Monday 8th

• Fixing font error
• Experimenting with curved arrow thickness
• Added extra ring in XZ plane (where focus point droplines contact plane)
• Adding mouse raycast for grab handles
• Storing handle state (currently dragging etc)
• Working out what guide arrows to show
  • Periapsis
    • Altitude
    • Around orbit (in plane)
    • Inclination
  • Apoapsis
  • Descending node
    • Twist
      • Circle
      • Very twisty arrow

Tuesday 9th

• Investigating how to do input properly
  • Don't use Mouse directly, use a pointer
  • Add generic input actions for things (next point, next mode, move value)
  • Mouse and non-mouse will basically be two independent modes
  • Added input settings class, with actions for next/prev mode
• Refactoring orbit axes out into separate classes, to handle the extra state associated with input
  • Also refactoring out handles
  • Orbit editor -> handles -> axes
  • Twist axis
    • Ascending node
    • Descending node
  • Arc axis
    • Apoapsis in plane
    • Apoapsis perpendicular plane
    • Periapsis in plane
    • Periapsis perpendicular plane
    • Ascending node in XZ plane
    • Descending node in XZ plane
  • Linear axis
    • Periapsis alt
    • Apoapsis alt
    • Descending node alt
    • Ascending node alt
• Input issue: shift+tab doesn't work, it always triggers the tab action
  • Input consumption is enabled

Wednesday 10th

• Refactoring to add input stuff to new axis classes
• Implementing linear axis drag
• Added "orbit modifier" interface that accepts axis values and modifies orbit
  • Twist
    • Ascending
    • Descending
  • Arc
    • Implementing dragging
      • Dragging in the pi to 2pi range isn't working as expected?
        • Orbit normal calculation breaks as it crosses over zero
    • Apoapsis in plane
    • Apoapsis perpendicular
    • Periapsis in plane
    • Periapsis perpendicular
    • Ascending in XZ
    • Descending in XZ
  • Linear
    • Periapsis
    • Apoapsis
    • Ascending
    • Descending
• Need to lock out all other handles when any handle is active

Thursday 11th

• Continuing with handles from yesterday
  • Ascending XZ
  • Descending XZ
  • Apoapsis perpendicular
  • Periapsis perpendicular
• Locking handles so that multiple cannot be selected at once
• Adding slight input damping, smooths out rapid mouse moves
• Unifying twist axis to just be a special case of arc
  • Need to change the line that connects mouse to twist ring
    • Split into a base "curved arrow" class, with arc and twist subclasses

Friday 12th

• Input mode for twist axis
  • Click, drag to ring to set start point, drag around ring to set angle
  • Capture initial inclination when dragging out of the ring, set inclination based on that + delta
  • Draw angle ticks

Monday 15th

• Adding "orbit validator" which checks that drag handles are not creating an invalid orbit
• Switching some orbital methods to take Radian instead of double, verifying no unit confusion
• Removed arc from apside to node (broken in some quadrants, doesn't really help visually)
• Scanning orbit to find closest point to mouse
  • Need to calculate the true anomaly for the selected point
• Refactoring open/closed orbit sweep to unify them into the same method
  • Removing camera adaptation, just render lots of points instead
  • Keeping track of true anomaly for each orbit point, returning this so mouse selection can use it

Tuesday 16th

• Adding flattened orbits back in
• Experimenting with an indicator showing selected orbit point
  • Crosshair, aligned with orbit radial out
  • Line from mouse to orbit
    • Change style when outside radius
  • Separate setting for selection stuff
    • Move radius setting over
• Refactoring:
  • Move orbit path to Values struct
    • Some work can be done in a job while other drawing is happening
    • Simpler lifetime management of native resources
  • Add ability to create/destroy additional TrueAnomaly markers, usable by other systems
    • Color
    • Size
    • Pulsing?

Wednesday 17th

• Trying out Project Auditor
  • Many many false positives for code
  • Useful results for content
    • Tweaking importer settings for SFX and music
    • Setting up default importers for audio with Preset Manager
      • Applying to lots of audio clips
    • Investigating creating a tool to force apply presets to folders
      • Not easy, don't bother for now
    • Enabled mipmap streaming
    • Expanded async upload buffer
• More tests
  • Span sort
  • Root finder edge cases
  • RigidTransformDouble
  • Plane3D
  • Exclude pure data components to remove noise in coverage report
  • Fixed some memory leaks in tests
  • More stuff I forgot to take note of
  • PolylinePathGenerator
• Added UnityEvent for clicking on orbit and selecting a position

Thursday 18th

• Converting OrbitEditor into prefab
  • Fix render layer
  • Input handling
  • Icon is tiny when zoomed out
    • Constant screen size
• Ui panel to show orbit info
• Todo:
  • Create orbit container in spawn mode
  • Create editor in spawn mode
  • Bind them
  • Create spaceship at selected point

Friday 19th

• Experimenting with pork chop generator
  • Potentially incorrect unit for velocity (km/s to m/s) in kepler calculation
    • Switched to central difference for now
  • Creating job to find minimums in a native array
  • Creating a colourmap that can be used in jobs
  • Job to convert double array to texture (with colourmap)
  • Applying updates to texture is very slow?
    • Solving lambert problem 1E6 times is overloading the job scheduler - other jobs are going slow because there's no space to schedule them. Something to be aware of when scheduling very heavy work in the future!
  • Splitting work up into a quadtree, subdividing nodes that need it

Saturday 20th

• Bilinear interpolation of quadtree data to texture

Sunday 21st

• Triangulating quadtree, rendering as mesh with vertex colours

Monday 22nd - Friday 26th

• Too hot to think

Friday 26th

• It's still too hot
• Improving mesh approximation of porkchop plot
  • Adding centre vertex with triangle fan (better interpolation)
  • Rejecting some nodes (very high dv value that we don't care about) for subdivision, using the extra details elsewhere
  • Sharing vertices between adjacent triangles in mesh
  • Still looks worse than texture (bilinear) version at all levels of detail

Monday 29th

• Improving menu planet randomisation
• Researching space based solar power projects, notes in timeline
• Disabled fleet selection in lobby
  • Remove all the supporting code for syncing fleet choice
• Setting up selection of orbit point for spawning ships

Tuesday 30th

• Investigating potentially incorrect velocity calculation in kepler orbit
  • Unit test says analytical and central difference produce the same results
  • Pork chop generator finds huge differences!?
  • Mass mixup! For an orbit of Earth around Sol mass used should be Sol mass not Earth mass!
  • Tightening up some unit tests around this area
• Removed FleetStorage, refactored into ShipStorage
• Removed fleet selection related bits from lobby
• Add ship selection dropdown to in game spawn menu