At fractional UI scales (e.g. 0.71 game scale + Windows display scaling)
the raw fractional pin offset landed icons between physical pixels,
rendering blurry and offset from the engine-drawn native quest blips.
Snap the offset to a whole physical pixel via the minimap's effective
scale so Questie's icons sit on the same grid as native markers.
The cutoff hid icons beyond its yard value before checking the minimap's
visible radius, so the default 100 hid icons that were clearly on the
minimap (view radius is 133-466 yd by zoom). Now it only clips icons
outside the visible circle: HBD's pin renderer gates the cutoff on
dist > 1, and QuestieMap FadeLogic raises the effective cutoff to at
least the current minimap view radius via new HBDPins:GetMinimapRadius().
Reverts the previous approach of stripping char.complete (which mutated
completed-quest data). Correct rule: a quest in the player's log is active
and should show its turn-in location regardless of the completed flag.
- AddFinisher now trusts the live quest log: if the quest is in
currentQuestlog and not failed, the '?' finisher draws regardless of the
(possibly stale, e.g. prior-prestige) char.complete flag. No completion
data is modified. Reverts the AcceptQuest and QUEST_QUERY_COMPLETE strips.
- Learner: OnTargetChanged now learns quest-giver/turn-in NPC spawns
(it previously only cached the GUID), so targeting a turn-in NPC records
its location. OnQuestComplete/OnQuestTurnedIn fall back to the 'target'
unit when the 'npc' gossip unit is already cleared, so the finisher NPC
is reliably learned on turn-in.
selene 0 errors; busted 145 successes / same 7 pre-existing failures.
A quest completed in a prior Ascension prestige stays in char.complete.
On re-accept it is active in the log again, but the stale completion flag
was never cleared — and AddFinisher only draws the turn-in '?' when
not char.complete[questId]. So the finisher was suppressed and the quest
showed '(Complete)' with stale objective text while standing at the
turn-in NPC (e.g. Aggression 8334 at Lanthan Perilon 15281).
An active in-log quest is now kept out of char.complete:
- QuestieQuest:AcceptQuest clears char.complete[questId] on accept.
- The async QUEST_QUERY_COMPLETE handler also strips any quest currently
in QuestiePlayer.currentQuestlog, so the server's completed list can't
re-flag a re-accepted quest.
Restores the turn-in '?' and normal objective tracking. selene 0 errors;
busted 145 successes / same 7 pre-existing failures.
#7: Completed quests kept showing as available '!'. The server completed-
quest list arrives async via QUEST_QUERY_COMPLETE, often after available
quests were first drawn. Recalculate available quests once char.complete is
populated by that event so completed quests are removed (also clears the
already-completed subset of #8).
#9: An accepted quest's available '!' could linger on the minimap until
/reload. UnloadQuestFramesByDataType unloaded the frame but left its name in
questIdFrames and _G; it now removes those references so the icon is torn
down immediately on both map and minimap.
#10: Added 'Hide repeatable quests below level 60' (Icons tab, off by
default) to hide repeatable available quests (e.g. the Ascension Callboard)
from map/minimap until level 60. Enforced in _DrawQuestIfAvailable and
ShouldBeHidden so existing minimap pins are removed too; reappear at 60.
Lint: selene 0 errors. Tests: 145 successes / same 7 pre-existing failures.
- Use 3.3.5a-native Minimap:GetViewRadius() (instance method) instead of
C_Minimap.GetViewRadius (nil on 3.3.5a) with C_Minimap fallback. Returns
116.67 at zoom 5 instead of the hardcoded 125 from the broken lookup table.
- Replace minimapWidth = (GetWidth()/2) * (mapRadius/155.52) with
minimapWidth = GetWidth() * GetScale() / 2. The new formula yields the
actual half-width of the visible minimap in screen pixels, the correct
multiplier for diffX*minimapWidth when diffX is normalized by mapRadius.
- Comment out (do not delete) all QDMATH / UPDATE entering / ICON debug
blocks as --[[ DEBUG: ... --]] for future regression investigation.
- Remove loaded debug scripts from Questie-X.toc and add to .gitignore
(drift_test, radius_debug, spawn_calibration, etc.). Scripts preserved
in tests/ for reference.
- Add .gitignore entries for tests/ lowercase variant (case-sensitive
on Linux/macOS) and additional debug script patterns.
- Update CHANGELOG.md and docs/changelog.html with fix entry.
- Refresh handoff.md to reflect RESOLVED status and iteration 4 history.
Verified: pins now stay anchored to world positions across all zoom levels
on both Stock UI and ElvUI. Per-frame diffX ~0.35 yards during walking
translates to ~0.21 pixels of pin movement (correct for 1 yard of world
movement).
The uiMapID==1241 branch in drawMinimapPin was an experimental overlay
that ran BEFORE the standard minimap projection and rotation transforms.
It duplicated the projection math while bypassing the rotation matrix,
used a different edge-clamp path, and is the primary suspect for why
pins visually follow the player on Sunstrider Isle.
The standard minimap path (handle rotation -> adapt delta to mapRadius
-> apply shape mask -> float-on-edge or clamp) already handles all
pins uniformly regardless of zone. Removing this experimental branch
lets 1241 pins use the same proven path as every other zone.
Also cleaned up:
- stale lastInstanceId reference in throttled PIN debug trace
- unnecessary data.uiMapID guard in QuestieDebugMinimapPin gate
- Sunstrider-specific minimap debug block in UpdateMinimapPins
- /qdrift: one-shot diagnostic that traces the full player-position chain
(GetCurrentPlayerPosition -> GetWorldCoordinatesFromZone), checks
mapIdToUiMapId entries, shows pin world coords vs player world coords,
and verifies zone center bounds
- Throttled [QD] PIN debug in drawMinimapPin (1/sec) showing pinWorld,
playerWorld, distance, mapRadius, and diff values
- Enhanced [QD] CONV/CONV2 debug in GetCurrentPlayerPosition showing
raw coords, world coords after conversion, and zone coords after
back-conversion
Purpose: diagnose why minimap pins follow the player after the areaId fix
Root cause of minimap pin drift on Sunstrider Isle: on Ascension,
GetCurrentMapAreaID() returns 3430 for Eversong, but mapIdToUiMapId only
had the WotLK mapping (463→1941). This meant actualUiMapId was nil in
GetCurrentPlayerPosition(), causing the parent→child zone conversion to
be skipped entirely. Eversong-zone coords passed through tagged as
uiMapID 1241, producing wrong world positions and pin drift.
Fix: add Ascension areaId→uiMapId mappings (3430→1941, 3431→1241) to
mapIdToUiMapId after UiMapData init. Now the conversion path triggers
correctly: Eversong coords → world coords (via 1941 bounds) →
Sunstrider zone coords (via 1241 bounds).
One-line fix in Compat.lua. No changes to HBD.lua or projection logic.
Adds _ComputeZoneSpaceDiff to HBD.lua that computes minimap pin
projection in the parent zone's coordinate space (Eversong 1941) instead
of using the world-coordinate delta. On Ascension, GetPlayerMapPosition
returns Eversong-relative coords for players on Sunstrider, but
GetWorldCoordinatesFromZone(1241) treats them as Sunstrider-relative,
producing a wrong player world position and causing pins to drift or
move opposite to the player.
Fix: (1) Store original zone coords on minimap pins at registration
(AddMinimapIconMap). (2) In drawMinimapPin, recover player's actual
Eversong zone coords by reversing the 1241-bounds conversion. (3)
Convert pin's 1241 zone coord to Eversong zone coord via 1241-world.
(4) Delta in Eversong zone-space, scale to pixels using parent-width/mapRadius.
Falls back to existing world-space math for pins without zone coords.
- Fix: isSunstrider block in _MergeSpawnEvidence now checks IsAscensionProtected
before writing learner data to npcDataOverrides. Without this guard, each Mana
Wyrm kill overwrote AscensionDB's z1241=5 data with learner z3431 coords.
- Fix: Clustering disabled for zone 1241 (Sunstrider Isle) in _DrawObjectiveIcons
so all 5 AscensionDB spawn pins display individually instead of collapsing to 2.
- Fix: Ghost pin loop in AvailableQuests.lua -- 'while frames[i]' was iterating a
string-keyed table with a numeric index (never iterated). Changed to pairs().
- Fix: NormalizeSpawnZoneKey now uses ZoneDB.areaIdToUiMapId for all zones so
learner evidence is stored under map IDs (e.g. 1241) not area IDs (e.g. 3431),
matching AscensionDB's key space. Applied in LearnNPC and _StoreGuidSpawnEvidence.
- Fix: isSunstrider detection in _MergeSpawnEvidence updated from hardcoded
zoneId==3431 check to IsSunstriderNativeZone() since zone IDs are now normalized
to map IDs at storage time.
- zoneDB: remap areaId 3431 → uiMapId 1241 (Sunstrider Isle) so quest 8335
pins render on correct map instead of Eversong Woods (1941)
- compat/HBD: remove 1241→1941 redirect (Sunstrider pins now handled by zoneDB)
- compiler: QuerySingle override check skips empty tables (next() guard) so
spawn data overrides don't fall through to nil base DB values
- MapIconTooltip: nil guards for minLevel/maxLevel to prevent crashes
- Map/QuestieMap: remove stale 1241→1941 redirect comment
- HBD.lua: Cache GetPlayerWorldPosition/GetPlayerZonePosition at 50ms
intervals instead of hammering GetPlayerMapPosition() every frame.
Invalidate cache on PLAYER_ENTERING_WORLD and ZONE_CHANGED_* events.
Expected ~97% reduction in C API position calls (45,570 -> ~1,200 / 10min).
- zoneDB.lua: Replace O(n) linear scan in GetAreaIdByUiMapId with an
O(1) reverse lookup cache (uiMapIdToAreaIdCache) built at Initialize().
Cache is kept in sync by ApplyCustomZones and name-match fallback now
caches its result so subsequent calls are also O(1).
- Tooltip.lua: Throttle GameTooltip OnUpdate hook to 100ms intervals
(was firing every frame at 60-144 Hz). Added _tooltipLastText cache
to avoid redundant GetText() + CountTooltip() calls when nothing changed.
- QuestieArrow.lua: Hoist _HasMissingCompletedFlag, _GetCompleteIconType,
_CollectFinisherSpawns, and _CollectObjective out of _CollectQuestTargets
to module-level functions. These were re-created as closures on every
UpdateNearestTargets call (1 Hz). Shared per-cycle context is published
via _arrow_* module upvalues to avoid closure capture overhead.
- QuestieLearnerComms.lua: Four micro-optimizations:
(1) Reduce ProcessQueues ticker 0.2s -> 0.5s (still 7x faster than
minChatInterval of 3.5s).
(2) Cache hidden channel ID at init; lazy refresh on disconnect.
(3) Drop LibDeflate compress level 9 -> 1 (fraction of CPU cost).
(4) O(1) messageCacheCount counter replaces O(n) pairs() size scan.
Xurkon