LootFeed: Roadmap

LootFeed — verification and development roadmap

Source-backed v0.2.2 baseline

• WoW original 1.12 / Lua 5.0-style APIs; TOC interface 11200.
• Five ordinary loot notifications with queued coalescing, item quality/icon/quantity, cash icon and exact Quest Item tooltip detection.
• Separate roll rows created as needed, with Need/Greed/Pass, dice-hover details, winner and honest uncertainty states.
• Hover pauses the visible stack and queued insertion; configurable direction, scale 0.6–1.8, duration 2–20s, anchor placement.
• Optional personal item-chat suppression through a wrapped ChatFrame_OnEvent and a showLootSpam CVar side effect for roll detail.
• Account-wide LootFeedDB with x, y, scale, duration, direction and mute.
• Optional roll preview /lfeed testroll; personal item preview /lfeed test.
• Octo Loot → Loot Feed migration helper and its PowerShell fixture test.
• Published v0.2.2 ZIP/checksum, notes reporting historic syntax/migration tests; current docs audit didn't execute tests.

P0 — actually validate the interaction

1. Install v0.2.2 on an intended Vanilla 1.12 client. Verify TOC, two Lua files, slash commands, saved settings and Lua error-free loading.
2. Loot one item, multiple items, money, a stacked item, a real quest-bound item and a normal item required by a quest. Verify only genuinely flagged Quest Items display the special label.
3. Send rapid duplicate loot events. Check that queued identical links coalesce, while already visible items retain their own row history and five ordinary rows do not expand indefinitely.
4. Test Scale 0.6/1/1.8, Duration 2/6/20, direction up/down, locked and unlocked drag, anchors on different UI scales, saved position after reload.
5. Hover an item and its tooltip: visible expiration and insertion should pause, but new loot may still queue. Check item cache retry when GetItemInfo initially has no data.
6. Enable /lfeed chat off: only self-item loot chat should disappear. Other players' loot, group-roll details and money should continue through intended channels. Test with other addons wrapping ChatFrame_OnEvent.
7. Record the value of showLootSpam before installation and after login. The source changes it to 1 when it was 0 and does not restore it; decide whether this side effect is acceptable and document actual observed behavior.

P1 — group roll state and ambiguity

8. Test real Need/Greed/Pass rolls in parties/raids, including numeric roll results, all-pass, winner, interrupted/cancelled local prompt and resolution after the local window closes.
9. Verify the source's local-prompt-close vs group-resolution distinction: CANCEL_LOOT_ROLL marks promptClosed but does not finalize the record.
10. Check a roll with an initially missing item link; the addon polls pending roll IDs for up to three seconds.
11. Test two simultaneous rolls of the same item link. Tooltip should disclose ambiguity rather than assign numbers/choices to an arbitrary copy.
12. Test a missing result crossing its original roll deadline, then an authentic late result. The status should indicate result unavailable until a real result arrives, without inventing a winner.
13. Validate the displayed 15-second post-resolution lifetime, hover pause, up to 300-second internal completed-record retention, and unlimited extra roll rows relative to five ordinary rows.
14. Test /lfeed testroll preview and confirm demo names/roll numbers are visibly marked as a preview, not a real event.
15. Test localized LOOT_ROLL_* constants including positional specifiers; test LOOT_ITEM_* messages separately, since that pattern compiler is simpler.

P2 — migration, packaging, accessibility

16. Run tests/test_upgrade.ps1 with PowerShell on temporary data; capture real test output. Next migrate a backup copy of actual OctoLoot SavedVariables and inspect results before attempting live user-data migration.
17. Download and checksum-match LootFeed.zip and SHA256SUMS.txt; inspect internal TOC and files. This audit checked release metadata only.
18. Test row contrast, hover targets, tooltip accessibility, tiny dice pips and animation on smaller/low-FPS screens. Decide whether reduced-motion options would be useful.
19. Add a scoped synthetic Lua regression test for Rolls.lua state transitions: duplicate item-link ambiguity, delayed winner, provisional timeout, localization and closed prompt.
20. Decide and document repository licensing; no LICENSE exists in this main tree.

Open creative choices (unapproved)

Does hover pausing the entire feed feel right in intense loot moments? Should the addon ask before changing showLootSpam? Should the group roll display offer greater uncertainty cues for missing participant eligibility? Do motion, scale and lifetime need per-category controls? Should the saved-data migration be easier on Linux?

Record actual creator decisions, implementation commits and in-game test results when these questions are answered.

Document information

Creator / decision authority: Babuni.dev.
Status: test plans and open options. They do not authorize new code, design or dependency changes.