{
  "version": "6.2.8",
  "url": "https://tracking.clearabovecamera.com/updates/CompTools_Setup_6.2.8.exe",
  "sha256": "9e77eafc94cae271ff47bd33de345970d7b4fb5eb766e1a23dfe81958d68aa45",
  "summary": [
    "FIX: the audio variometer no longer clicks / cuts out \u2014 a beep's pitch and",
    "FIX: the HUD vertical-rate (VS) readout is far steadier and more accurate \u2014 it",
    "FIX: navigating to the Landing pin no longer drops a second \"baggie\" marker",
    "FIX (crew companion): the Landing pin no longer shows a duplicate blue target",
    "FIX (crew companion): the map / settings / tool buttons no longer appear",
    "CHANGE: the descent trace follows the wind layers a little more closely \u2014 each",
    "FIX: a set Landing pin no longer reloads as a stray waypoint named \"Landing\"",
    "FIX: a very long flight (or long emulator run) no longer gets progressively"
  ],
  "notes": "Competition Tools \u2014 Release Notes\nWhat's new from v6.2.7 to v6.2.8\n==================================================\n\nSUMMARY\n-------\n- FIX: the audio variometer no longer clicks / cuts out \u2014 a beep's pitch and\n  cadence are now held steady for the whole beep instead of being recomputed\n  mid-tone (which truncated the sound and caused the clipping).\n- FIX: the HUD vertical-rate (VS) readout is far steadier and more accurate \u2014 it\n  now uses the same robust 8-second smoothed climb estimate the solver and\n  descent tools already trust, instead of a raw two-point GPS-altitude rate.\n- FIX: navigating to the Landing pin no longer drops a second \"baggie\" marker\n  dot on it \u2014 the Landing is a destination, not a scoring drop target;\n  bearing/distance/ETA and the nav wind line still work.\n- FIX (crew companion): the Landing pin no longer shows a duplicate blue target\n  dot and its 100 m target ring stacked on the red landing flag.\n- FIX (crew companion): the map / settings / tool buttons no longer appear\n  \"blacked out\" when the site is added to an iPhone Home Screen.\n- CHANGE: the descent trace follows the wind layers a little more closely \u2014 each\n  layer holds its own wind across more of its band and turns over a narrower zone\n  (less inter-layer line smoothing), so a sharply-different layer bends the trace\n  more instead of being smoothed flat.\n- FIX: a set Landing pin no longer reloads as a stray waypoint named \"Landing\"\n  after an app restart \u2014 the landing now persists and comes back as a landing\n  pin (its hidden nav-target mirror is rebuilt, not saved as a waypoint).\n- FIX: a very long flight (or long emulator run) no longer gets progressively\n  laggy \u2014 the live GPS track now extends by appending the new segment instead of\n  redrawing the whole track every fix (which grew to ~1 s per fix hours in).\n\nDetails below.\n\nFIX: AUDIO VARIOMETER CLICKING / CUTTING OUT\n--------------------------------------------\n- Bug: the vario's beep clicked and cut out, especially while the climb/descent\n  rate was changing (i.e. most of a real flight).\n- Root cause: the synth recomputed each beep's cadence AND its raised-cosine\n  window length on every 40 ms audio chunk from the live vertical acceleration,\n  so a rate change mid-beep truncated the envelope part-way through \u2014 an audible\n  click / dropout.\n- Fix: the cadence and pitch are now latched for the whole beep and only\n  re-evaluated at a beep boundary (across the silent gap, where a pitch step is\n  inaudible and the waveform phase stays continuous). Each beep now starts and\n  ends cleanly at zero, and a per-chunk cap stops a post-stall burst skipping a\n  beep.\n\nFIX: HUD VERTICAL-RATE READOUT WAS ERRATIC\n------------------------------------------\n- Bug: the VS (vertical rate) card jittered and read inaccurately.\n- Root cause: it computed a raw two-point \u0394altitude/\u0394time over ~0.5-1 s of\n  quantized, noisy GPS altitude with only light smoothing.\n- Fix: it now reads the FlightTracker's robust 8-second linear-regression climb\n  estimate \u2014 the same filter the wind-sample weighting and the descent\n  start-rate already use \u2014 falling back to the old two-point value only until\n  that estimator has enough history.\n\nFIX: LANDING PIN DREW A SECOND MARKER WHEN NAVIGATED TO\n-------------------------------------------------------\n- Since the Landing pin became a real navigation target, navigating to it drew\n  the marker-drop \"baggie\" projection dot on top of it \u2014 but you don't drop a\n  marker on your own landing spot.\n- Fix: the marker-drop projection is skipped when the Landing pin is the nav\n  target (treated like a virtual, no-drop target), so no second dot lands on the\n  red triangle. Bearing / distance / ETA and the nav wind line still work.\n\nFIX (CREW COMPANION): DUPLICATE LANDING DOT + 100 M RING\n--------------------------------------------------------\n- The crew Chase webpage drew a duplicate blue target dot \u2014 and its 100 m target\n  ring \u2014 stacked on the red Landing flag, because the Landing was being\n  published to the crew as a ground target as well as a landing pin.\n- Fix: the Landing is published only as the landing pin now, so the crew shows\n  just the red flag (with its parcel outline) \u2014 no second dot, no ring.\n\nFIX (CREW COMPANION): CONTROLS \"BLACKED OUT\" ON IPHONE HOME SCREEN\n------------------------------------------------------------------\n- Added to an iPhone Home Screen (standalone PWA), the Map/Sat chips and the\n  round map buttons (parcel / centre / locate / settings) rendered as dark,\n  unstyled blobs that blended into the map.\n- Root cause: those frosted-glass controls used backdrop-filter without the\n  -webkit- prefix that iOS still requires, so the blur dropped out and the\n  dark-translucent buttons sank into the dark basemap.\n- Fix: added -webkit-backdrop-filter so the frosted controls render on iOS.\n\nCHANGE: DESCENT TRACE FOLLOWS THE WIND LAYERS A LITTLE MORE CLOSELY\n-------------------------------------------------------------------\n- The descent trace + marker-drop projection look the wind up per altitude by\n  interpolating between the profile's layers. A full linear interpolation spread\n  a distinct layer's wind across the whole gap to its neighbours, so a sharply-\n  different layer barely bent the line \u2014 it was only fully felt at one exact\n  altitude.\n- The lookup now holds each layer's wind across the middle of its band and\n  transitions over a narrower central zone (a plateau-and-ramp instead of a full\n  linear lerp), so a distinct layer curves the trace more \u2014 without the hard\n  kinks a nearest-layer snap used to cause. The predicted landing point uses the\n  same lookup, so it tracks the layers a touch more too. (The bin size is\n  unchanged at 50 ft.)\n\nFIX: A SET LANDING RELOADED AS A STRAY \"LANDING\" WAYPOINT ON RESTART\n--------------------------------------------------------------------\n- Bug: after setting a Landing pin and restarting the app, the landing came back\n  as an ordinary waypoint named \"Landing\" (in the list / on the map) instead of\n  the red landing pin.\n- Root cause: making the Landing pin a nav target (6.2.7) keeps a hidden mirror\n  waypoint behind the scenes. State save was persisting THAT mirror as a normal\n  waypoint (and didn't store its hidden flag), while the landing pin itself\n  wasn't saved at all.\n- Fix: the landing PIN is now persisted and restored as a landing (its nav\n  target rebuilt automatically); the hidden mirror is no longer saved. Saves made\n  before this (no landing stored) simply restore no landing.\n\nFIX: PROGRESSIVE LAG ON A VERY LONG FLIGHT\n------------------------------------------\n- Bug: the app got steadily laggier the longer a flight ran (real flights and\n  long emulator runs alike), eventually stuttering badly hours in.\n- Root cause: the live GPS track was torn down and redrawn IN FULL on every\n  recorded point, so the redraw cost grew with the track length \u2014 measured at\n  ~1.2 s per fix at ~29,000 points (O(n^2) over the flight). Every other per-fix\n  stage stayed flat and tiny.\n- Fix: the live track now APPENDS just the new segment each point instead of\n  redrawing the whole polyline (a full redraw happens only on a new track, an\n  Edit-style restyle, or a reset). Per-fix cost is now constant regardless of\n  flight length; the balloon marker was always drawn separately, so nothing feels\n  different \u2014 it just stops slowing down.\n\nHelp center updated for the above.\n"
}