TOGBankClassic Changelog

[v1.8.1] (2026-10-09) - Oldest Order First

Bug Fixes

• BANKCOLLECT-SHARE-001: at the bank, Fulfill Oldest skipped a newer order for the same item. crimsonmane, Discord 2026-10-08: Fulfill Oldest "does not fulfill the oldest order first" when a second player orders the same item plus more. Mail:FindOldestBankFillableOrder (Modules/Mail.lua) measured every order against the whole bag count, so the stack I pulled for the older order also "covered" a newer order of the same item, and the newer one was never collected for -- it then could not be filled at the mailbox, and only that player's other items went out. The bags are now handed to open orders oldest-first, and each order's shortfall is what is left after the older orders' share. Tests/bankcollect_spec.lua pins both the search and the click-through. Not yet checked in a game client.
• FCFS-001: at the mailbox, Fulfill Oldest mailed a newer order ahead of an older one for the same item. The operator, 2026-10-09: "you are supposed to fill the oldest order first. first come first server". Mail:FindOldestServiceableOrder (Modules/Mail.lua) picked the oldest order the bags could fully fill and skipped any it could not, so a newer, smaller order for the same item went out first and spent the stock the older order was waiting on. I had described that skip as by design in this entry and in my Discord reply to crimsonmane; it contradicted the operator's rule. Now the first order for an item that cannot be filled yet holds its place: no newer order for that item is sent until it is. Orders for other items still go out, since they do not use the same stacks. Mail:FindMoreOrdersFor (the same person's other orders riding on one mail) now leaves the stock other people's older orders need, so riding along cannot jump the queue either. The three fulfillment searches share one "oldest" ordering and one item key. The envelope's tooltip and the Requests "?" help in Modules/UI/Requests.lua said unfillable orders "are skipped"; both now describe the first-come-first-served rule, as does the Fulfill Oldest feature line on the CurseForge page. Four examples in Tests/bankcollect_spec.lua, including crimsonmane's case. Not yet checked in a game client.

Internal

• SUITE-SPEED-001: the whole forward test run fits in one run again. Every spec builds a fresh copy of the addon, and libraries that load once per run (AceComm, AceEvent and the other Ace mixins, AceAddon's queue of addons waiting to load, VersionCheck's listener list) kept a strong reference to every earlier copy, so the run held tens of thousands of dead frames and each reset's garbage collection grew slower; late files took 40 seconds or more that take under one alone. env.reset in Tests/env_togbank.lua now clears those references, and a test window's close releases its tab contents instead of only hiding them. Measured: objects left after browse_spec 30,800 to 5,400. The forward run passes whole (1970 examples, 2026-10-09) and matches the three reverse parts (545 + 722 + 703). Clearing the handler lists means no spec can rely on an addon copy from an earlier file still hearing messages; the reverse parts passing is the check on that. No addon code changed.

[v1.8.0] (2026-10-08) - Donations Setting

The "Enable donations" setting is back, off by default, and the Donation popup can be turned off again. Checked in a game client by the operator on 2026-10-08 (the Donation popup, the tooltip toggle and the outstanding in-game checks from earlier releases all passed).

New Features

• TOOLTIP-TOGGLE-001: a setting to turn the bank stock off in item tooltips. The operator, 2026-10-07: "someone asked to be able to turn it off" / "we need to add one to settings, and have it on by default". There was no such setting: Modules/TooltipBankerInfo.lua drew its "Bankers:" block on every item tooltip unconditionally. General settings gain "Show bank stock in item tooltips" (db.global.bank.itemTooltipBankers, account-wide like the Mailbox switches, AceDB default true; Options:IsItemTooltipBankersEnabled also reads a missing key or no DB as on). The check is in TooltipBankerInfo:AppendTo, the one renderer, so it also covers TOGProfessionMaster's recipe tooltips, which call it. Tests/tooltipbankerinfo_spec.lua pins the default, the off state and the shipped toggle.

Bug Fixes

• MAILBOX-POPUP-001: the Donation popup could not be stopped on a bank character. GUNNEY, Discord 2026-10-07: "donation window pops up on all bankers even when unchecked and I cannot get it to stop poping up". Confirmed in the code: Mail:Scan (Modules/Mail.lua) stood aside only while the Mailbox window auto-opened, so unticking the General setting "Open the Mailbox window at a mailbox" (a global setting, so every banker at once) brought the per-mail Donation popup back in its place on every mailbox visit -- and v1.5.2 (UX-WATERFALL-001) had removed the per-character "Enable donations" box that used to turn the popup off. The only remaining off switch was "Enable for <player>", which also stops bank scanning. That "the popup returns when unticked" rule was my design choice in UX-WATERFALL-001, not a request, and it contradicted the setting's own description ("Untick to stop that").

• DONATIONS-BOX-001: the "Enable donations" setting is back, unticked by default. The operator, 2026-10-07: "we should have the setting to enable/disable donations back where it was, and have it unchecked by default." The per-character box returns to the Bank panel at its old place (order 2, db.char.bank.donations, Options:GetDonationEnabled), and it alone decides whether the Donation popup opens at a mailbox: Mail:Scan no longer consults the Mailbox window's setting or verdict, so the two settings are independent. The AceDB default changes from true to false; AceDB never wrote the old default to disk, so every character that never touched the box starts unticked. This replaces the first fix above, which had tied the popup to the Mailbox window's setting a second way. KNOWN COST: a banker who ticks both gets both windows, as before v1.5.2; gifts taken through the Mailbox window are credited either way. Tests/donations_spec.lua pins the popup following the box alone, the box's place, key and OFF default, and the real accessor.

• PKG-EMBED-001: the CurseForge app stopped installing LibDBIcon for Classic Era, TBC and MoP. The copy of LibDBIcon bundled for WoW Forever in v1.7.1 (Libs/LibDBIcon-1.0/LibDBIcon-1.0.lua, taken from TOGProfessionMaster) began with the packager's --@curseforge-project-slug: libdbicon-1-0@ marker. The BigWigs packager greps every packaged .lua for that marker and records the slug as embeddedLibrary (release.sh:1882-1889), overriding .pkgmeta's required dependency, so the upload told CurseForge LibDBIcon was embedded and the app did not install the standalone addon -- while the Era, TBC and MoP TOCs list it under ## Dependencies and never load the bundled copy. Marker removed; the harness .pkgmeta checker now reports libdbicon-1-0=requiredDependency for the upload, and Tests/wiring_spec.lua fails if the marker comes back. Whether v1.7.1's CurseForge file actually shows LibDBIcon as embedded was not checked on the site (not verified). The design is the operator's (2026-10-07): bundled ONLY for WoW Forever, an external required dependency everywhere else. .pkgmeta also lost all its comments, per the operator's no-comments rule of 2026-10-03; release.sh skips whole-line comments, so that change alters nothing uploaded.

Internal

• HARNESS-PIN c2fd446: Tests/wowapi moved from a49127f. Taken for WoWAPITesting's fix to harness thread 1a91c3ca. Under load the runner's per-file TIMEOUT used to escape as a bare lua.exe: (error object is not a string), with no file named; it is now a named failure. Adoption fix: two Tests/searchbox_spec.lua fixtures built a FontString with no font, which the harness now refuses as the client does (FontString:SetText(): Font not set); they use GameFontHighlightSmall, the real cells' font. Production was already right (Requests.lua gives every glow copy a font object).
• GATE-LOAD-001: the gate survives a busy machine. run.lua's per-file limit is wall time on Windows (WoWAPITesting, same thread), so it tripped whenever other desk runs shared the machine. Every forward part now passes --timeout=240, and Tests/run_reverse.lua passes --timeout=600 (browse_spec ran past 240 s late in the reverse order with three other runs in flight, and passed 697/0 at 600 s). Forward B2 is split into B3a and B3b, because under load its 52 files outran the desk's 15 minutes. CLAUDE.md names the new parts.
• BOARDS-REMOVED-001: the last boards are gone. docs/LIBRARY_CONTRACTS.md, docs/DEPENDENCY_CONTRACTS.md, Tests/HARNESS_CONTRACT.md, docs/contracts.md and docs/AUDIT.md were imported into the writ inbox and deleted. Writ, 2026-10-07: the operator's directive is "the boards should be removed after a migration"; git history keeps them. All 81 imported items were triaged the same day: 79 were already delivered, fixed or board prose. The other two were closed with a comment-level change: LOG-001's self-logging is kept on purpose and now says so (Modules/Output.lua), and Q1's false "nil when there is no link" claim was corrected (Modules/MailInventory.lua).

[v1.7.1] (2026-09-29) - WoW Forever

TOG Bank on WoW Forever, the fourth supported client. Checked in a Forever client by the operator on 2026-09-29: "everything seems to be working right now ... the display and everything is working, no more lua errors", and order highlighting "works now as intended". What that session could NOT check, for lack of other guild members running the addon: sync between players on Forever, and other players' names arriving over addon messages there.

New Features

• FOREVER-001: WoW Forever is a supported flavour (TOGBankClassic_Camelot.toc, new, Interface 16001; the operator, 2026-09-28: "i have wow forever now, we need to onboard to make that a supporeted version with it's own .toc and updating the .ps1"). The fourth TOC is the Era TOC with one deliberate difference: LibDBIcon-1.0 is not published for Forever (it is absent from the _classic_beta_ install; TOGProfessionMaster's Forever TOC embeds it for the same reason), so a hard ## Dependencies: entry would stop the whole addon loading there. That TOC lists it under ## OptionalDeps: instead and loads a bundled Libs/LibDBIcon-1.0/LibDBIcon-1.0.lua (MINOR 56, the copy TOGProfessionMaster ships) straight after LibDataBroker; LibStub keeps the higher MINOR if a standalone copy is also installed. TEMPORARY: once LibDBIcon-1.0 ships for Forever, make it a dependency again and delete the bundled copy. Every other dependency (Ace3, AceCommQueue-1.0, VersionCheck-1.0, DeltaSync, LibAceGUIWidgets, GuildRoster, ItemDB, ProfessionDB) already declares Interface 16001 or ships a _Camelot TOC. Not yet checked against the Forever client API (F:\Blizzard API Docs\wow-ui-source-forever) and not yet run in a Forever client.
• FOREVER-HIGHLIGHT-001: order highlighting reaches the WoW Forever bank. Modules/ItemHighlight.lua looked only for Classic's BankFrameItemN buttons and the -1 vault, neither of which Forever has. Its bank is the tabbed BankPanel (Forever Camelot/BankFrame.xml:85), which draws one tab from a pool of buttons. UpdateBankPanelHighlighting walks BankPanel:EnumerateValidItems() and reads each button's own GetBankTabID() / GetContainerSlotID(). A tab switch re-inits the pooled buttons with no bag event (GenerateItemSlotsForSelectedTab), so that is hooked to refresh, and the refresh clears every button it dimmed first. The dim-or-clear step for one slot is now one helper, HighlightBankButton, which the Classic vault and bank-bag loops also use. Spec'd in Tests/itemhighlight_spec.lua. Not run in a Forever client. The icon key on Forever's bank item button (icon or Icon) was not found in the Forever source; both are tried, and a button with neither is left undimmed rather than raising.
• FOREVER-HIGHLIGHT-002: the bags too. The operator, on Forever with no orders open, ticked "Highlight needed items" and nothing in the Combined Backpack greyed. That is not a Forever difference in what is needed: every item should grey. The bag path looked up Classic's ContainerFrameNItemM buttons, and Forever's modern container frames (ContainerFrameCombinedBags, and ContainerFrameContainer.ContainerFrames for split bags) draw pooled buttons with no such names. UpdateModernBagHighlighting walks each shown frame's EnumerateValidItems() and reads each button's GetBagID() / GetID() (Forever Mainline/ContainerFrame.lua:560, :566-567). It answers false on a client without those frames, so Classic Era, TBC and MoP keep the Classic path unchanged. Not run in a Forever client. Whether Forever's own button refresh resets the grey before our next refresh is not known.

Bug Fixes

• FOREVER-002: TOG Bank raised a Lua error at login on WoW Forever (the operator, from game, 2026-09-28: bad argument #2 to '?' (Usage: local success = self:HookScript(scriptTypeName, script [, bindingType])) at Modules/TooltipBankerInfo.lua:117). The Forever client's GameTooltip has no OnTooltipSetItem script -- nothing in wow-ui-source-forever references it -- and builds item tooltips through TooltipDataProcessor.AddTooltipPostCall(Enum.TooltipDataType.Item, ...) (Blizzard_SharedXMLGame/Tooltip/TooltipDataHandler.lua:199). TooltipBankerInfo:Initialize now tries the old hook under pcall first, so every client that has it keeps exactly its behaviour, and falls back to the item post-call, limited to GameTooltip (the only tooltip the old hook reached). pcall rather than HasScript: the offline harness does not claim the client's HasScript semantics. Tests/tooltipbankerinfo_spec.lua: a client refusing the script gets the post-call, which renders through AppendTo on GameTooltip and leaves any other tooltip alone. TooltipDataProcessor added to .luacheckrc and .luarc.json. Not yet re-run in a Forever client.
• FOREVER-NAME-001: a banker on WoW Forever was not a banker to itself, so the Bank settings page never appeared (the operator, 2026-09-28: "i set myself as gbank, and show up on the roster ... but i don't have the options in the settings menu to make myself visible as a banker"; /togbank listed them as Quebeldorf Flamehammer). Forever characters have a first name and a surname: the roster lists First Surname, and UnitName("player") returns the two as separate values (Blizzard's Blizzard_FrameXMLUtil/Camelot/NameUtil.lua:4-12 joins them with Constants.CharacterNameSeparatorConsts.CHARACTERNAME_SURNAME_SEPARATOR). Guild:GetPlayer used only the first, so the player was Quebeldorf-Realm, matched no roster key, IsBank(self) was false and Options:InitGuild never registered the page. The same wrong name reached every self-check (own scans, own requests, own-echo guards). New Guild:GetPlayerFullName() (Modules/Guild.lua) appends the second value as a surname only on a client that has that separator constant and only when it is not the realm (the documented second return is unitServer, and UnitPopupUtils.lua:126 shows the slot can carry either); it is used by GetPlayer and the request dialog's fallback (Modules/UI/Search.lua). Era, TBC and MoP have no such constant and keep the bare name. Tests/guild_spec.lua: four examples (surname joined and IsBank true; realm in the second slot ignored; a surname already in the first slot not doubled; no constant, second value ignored). Constants added to .luarc.json. Not yet re-run in a Forever client, and other players' names arriving over addon messages on Forever have not been checked for the same split.
• PENDING-STATE-001: a banker's own bank read "Old format" between its scan and its first publish (the operator, 2026-09-29, on a brand-new Forever banker's Bankers tab: "there is only one format of data on newer clients, right?"). Guild:GetAltStaleness returns "v1" for any held copy with no canon, and a canon is minted only AT publish -- so a banker whose scan was stored but held by the MULTIPC-001 publish gate (the guild had not answered yet; the status bar's countdown) was labelled old-format. New state "pending": our OWN bank, no canon, and Bank.deferred set. The Bankers tab says "Not published yet" in yellow (Modules/UI/Browse.lua STATE_TEXT / STATE_COLOR / STATE_RANK), the Inventory tab is not painted red and its tooltip explains the hold (Modules/UI/Inventory.lua). A later version named by another PC still wins ("behind"), and anyone else's canon-less copy is still "v1". Every flavour, not only Forever -- any new banker in a quiet guild saw it. Tests/tabstaleness_spec.lua: one example walking pending -> behind wins -> another banker stays v1 -> published is current.
• FOREVER-BANK-001: on WoW Forever every bank-character scan raised an error, so a banker's items never reached the Guild Bank window (the operator, 2026-09-29, with BANK debug on: OnUpdateStop called / Calling Scan() and nothing after, then Inventory/Scan.lua:77: bad argument #1 to 'GetContainerNumFreeSlots' from Bank:Scan on closing the bank). Forever has no BANK_CONTAINER, NUM_BANKGENERIC_SLOTS or NUM_BANKBAGSLOTS -- the Forever tree defines only NUM_BAG_SLOTS -- because its bank is the modern bank-TAB system: each purchased character tab is a container whose id comes from C_Bank.FetchPurchasedBankTabData(Enum.BankType.Character), which is how Forever's own Blizzard_UIPanels_Game/Camelot/BankFrame.lua:91-98 counts them. New TOGBankClassic_Constants.BankContainers() (Modules/Constants.lua) is the one list of the character's bank containers: on Classic Era, TBC and MoP the vault then the bank-bag range, the same ids in the same order as before; on Forever the character tabs. The ACCOUNT (warband) bank is left out on purpose -- it is shared by every character on the account, so each alt would publish the same items as its own. Inventory/Scan.lua walks the tabs when there is no BANK_CONTAINER (Scan:ScanBankTabs; the Classic branch is untouched) and treats tabs reporting no slots as "bank out of sight", keeping the stored bank rather than emptying it (INV2-VAULT-001). Bank:FindItemsInBank and Mail:FindEmptyBankSlot walk the same list; on Forever both would have raised the same way. Tests/scan_spec.lua: four examples (both character tabs read and the account bank not; no tab or no slots returns nil; a whole ScanAll completes; the Classic list is the vault then BankBagRange). Tests/itemhighlight_spec.lua's BANKSLOT-001 one-spelling guard accepts BankContainers as the shared spelling and pins that it is built on BankBagRange. C_Bank added to .luacheckrc and .luarc.json. FOREVER-BANK-002 (the same session's self-audit): the tabs are read only while the bank window is open (Bank.atBank, set on BANKFRAME_OPENED, cleared after the close's own scan in Modules/Events.lua), because a tab reporting slots is not proof its contents are readable away from the bank, and an empty read there would have replaced the stored bank on the next vendor or mailbox scan; scan_spec pins nil away from the bank. Item highlighting on Forever's bank followed as FOREVER-HIGHLIGHT-001, above.
• FOREVER-NAME-002: on WoW Forever a character's name carries no realm, and TOG Bank added one (the operator, 2026-09-28: "could it have something to do with the fact that Forever doesn't have <name><realm> it has <first name><last name>"). Forever runs regional unique names (RegionalUniqueNamesEnabled(), which exists only in the Forever tree): the full name is the identity, with no realm. The guild roster returns it bare, Forever's own AceDB keys the character as First Surname (AceDB-3.0.lua:263-278 in the _classic_beta_ install), and LibGuildRoster keys its roster by the bare full name (LibGuildRoster-1.0.lua:1264, :1345, :1659). TOGBank's memberRoster is keyed by the library's names (Guild:_RefreshFromRosterLib), but every lookup went through NormalizeName, which appended -<realm>, so on Forever every direct roster lookup (IsBank's fast path, IsInCurrentGuildRoster, IsPlayerOnline's fallback, whisper targets) missed. On a regional-names client NormalizeName now keeps a bare name bare and drops a realm suffix, and GetPlayer is the bare full name -- the library's rule, in the library's words. The earlier FOREVER-NAME-001 surname join is now gated on the same switch instead of on the separator constant. Era, TBC and MoP have no RegionalUniqueNamesEnabled and take the old path unchanged. Tests/guild_spec.lua: five examples (the player is the bare full name and IsBank finds them under the library's key; a realm suffix is dropped; a surname already in the first slot is not doubled; no switch, or a switch answering false, changes nothing). The zero-data banker record saved under the old -ClassicBetaPvE key is removed by the existing stale-stub cleanup at the next roster rebuild. Not yet re-run in a Forever client.
• FOREVER-EVENTS-001: on WoW Forever the addon stopped half-way through setting up its event handling, so a banker's bags were never stored and the Guild Bank window stayed empty (the operator, 2026-09-28: bags full, Guild Bank window "0 items across 0 banks", no debug output with BANK and uncategorized debug on). Events:RegisterEvents called the bare ChatFrame_AddMessageEventFilter, which on Era and Forever alike is only a deprecation alias of ChatFrameUtil.AddMessageEventFilter (Blizzard_DeprecatedChatInfo/Deprecated_ChatFrame.lua:51, loaded only while the loadDeprecationFallbacks CVar is on). On the operator's Forever client it was nil, so the call raised and everything after it was skipped: the mailbox hooks, the share timer and eventsRegistered. It now calls ChatFrameUtil.AddMessageEventFilter, the same function the alias points at on Era, TBC and MoP (ChatFrameFilters.lua:195 in each tree), with the alias only as the fallback. A sweep of every global Forever defines only in its Blizzard_Deprecated* files found two more, both in the debug chat tab (Modules/Output.lua): ChatFrame_RemoveAllMessageGroups and ChatFrame_RemoveAllChannels, now called as the chat frame's own ChatFrameMixin methods, present on all four clients. Tests/events_spec.lua: RegisterEvents finishes with the alias absent and installs the offline-whisper filter through ChatFrameUtil. Not yet re-run in a Forever client; whether this is the whole reason the scan stored nothing there is NOT established -- the empty debug output is consistent with it, but the scan's own path was not shown to raise.

Internal

• FOREVER-001: the lockstep specs cover the Forever TOC. Tests/wiring_spec.lua pins it as the Era TOC with only its Interface line and the LibDBIcon difference changed, so a module added to one TOC only still fails; the per-TOC "is this module loaded" examples in browse, item, logapi, mailbox, pricelist, sendresult, togtoolsbridge, usable, windowalpha and guildroster_integration now include it. LIBDBICON-DEP-001's "not vendored" example keeps the three other TOCs and names Forever as the exception.
• wow-version-replication.ps1 also mirrors into _classic_beta_, the WoW Forever install.
• Harness pin ee56a45 -> a49127f; frames.setScreenSize adopted. Tests/browsepersist_spec.lua's standUp now makes one frames.setScreenSize(2560, 1440) call. It replaces three hand writes, one of them into the harness's private UIParent._fixedRect (harness thread 36d5a55f). The undragged-window scale example is back at scale 2. Its 1.25 workaround was hiding a gap in my own hand-rolled resize, not a harness defect.
• REVERSE-TIMEOUT-001: the reverse runner passes --timeout=240. This pin brings in a 60 s per-file CPU limit. In the a-end reverse slice, browse_spec took 97 s where it takes 9 s alone, because every frames.reset() fully collects what earlier files left alive (~450 MB, ~150k frames by then). All 696 examples passed once the limit was raised. Tests/run_reverse.lua now also forwards flags after <part> <parts> to run.lua (--times), so a slice can be measured. The timeout was escaping as (error object is not a string) with no file named; reported to WoWAPITesting on thread 5f29b50a and fixed there in 2b63397 (now a named TIMEOUT line and one failure). The harness's new .pkgmeta checker (pkgmeta.lua) reports no FAIL here; its six WARNs are the defensive ignore entries that match no tracked file today (tmpclaude-*, *.bak and the like), kept on purpose. KNOWN COST: a file that slows down in reverse order is now caught at 240 s, not 60. The retained heap is still SUITE-HEAP-001's.
• SPEC-NIL-001: 19 weak is_nil assertions strengthened in the five spec files above that writ's new spec-shape check flagged (Writ inbox thread 004cbf2f: is_nil on a literal key passes for ever if the literal is misspelt). Removed members are now pinned by an enumerated key set (browse, item); a state that is cleared is first asserted present (mailbox, pricelist, guildroster_integration). The PriceList debounce example now holds the timer, so it sees the latch set, one publish per burst, and the latch released -- it previously ran the timer synchronously and never observed the latch at all. The remaining 59 sites in 22 more spec files followed the same day, so writ's spec-shape check is clean repo-wide (Tests/wowapi excepted, per the thread). The method, in order of preference: a whole-table assert.same where the table should hold exactly something (L.opening, info.alts, the answer kinds a responder sent); an enumeration of keys by pattern where a removed member must not come back under any name (snapshot helpers, the Escape and store settings, the old-wire observer, vendored libraries); an assert.is_not_nil on the same expression BEFORE the state that clears it (batch state, pending sends, the legacy rows); a local constant shared by fixture and check for bracketed string keys, which the checker does not pair; and the public API where one answers the question (D:Entries(), Guild:IsBank, Switches:IsEnabled).
• LAGW-PERSIST-SCALE-001 + LAGW-FORGET-001: TOGBank's window-scale stand-ins retired onto LibAceGUIWidgets MINOR 35/36 (released as v0.2.7). The library's PersistWindow follows a live scale change itself for an AceGUI Frame and re-evaluates a FUNCTION floor on each change, so UI:PersistWindow now hands it the max(1, scale) floor rule as floorAt(px) and registers no listener of its own; Persist_OnScaleChanged, the window._togPersist record and UI:SetPersistedAnchor (SCALE-DOCK-001's re-dock hook, and its two callers in Requests.lua and Search.lua) are deleted. The re-dock hook existed only because the old listener re-ran PersistWindow, whose ApplyStatus cleared a hand-made dock; the library's listener raises size and bounds only, so a docked window keeps its anchor with nothing registered. UI:ForgetPersistedWindow is now W:ForgetWindow (which also restores a ClearFrame handle's own floor, SCALE-FLOOR-002). An older library copy without ForgetWindow keeps the pre-35 hand-removal of the library record, and gets NO floor (LAGW-OLD-FLOOR-001, below). browsepersist_spec: the live-scale examples pass unchanged on the library's listener alone; the released-window example checks the library's record; SCALE-DOCK-001's three examples now pin a dock surviving a scale change with nothing registered, a control proving ApplyStatus would clear it, and a released docked window left alone by the next scale change. Not run in a client.
• DS-422-001 / CLAMP-SCREEN-001: 11 offline failures fixed, all from two sibling-library releases the specs had not caught up with (the addon code was right in each case). Whole suite now 1939 forward and 1939 reverse, 0 failed.
  • DeltaSync v4.4.0 (MINOR 22) parks a broadcast whose numbers a listener cannot yet resolve and replays it once the table lands. A viewer now requests straight from the banker's replayed login broadcast, with no offer and no version query, seconds after adopting the table. fullsync_spec pinned the old offer + query round (4 examples): the first-sync example now asserts the shorter path and that it goes to the author; "four cold viewers" counts the slots while they drain; "already current" counts offers from the second broadcast on; SYNCED-001 reads "pending" as the broadcast leaves, not two seconds later when the receipt has genuinely arrived. congestion_spec's park example stamped its offer v = 99, and a table version is a timestamp, so 99 now reads as a table BEHIND the viewer's and is dropped rather than parked; it uses the viewer's own version.
  • LibAceGUIWidgets v0.2.7 (MINOR 36): PersistWindow clamps a saved window to the screen. browsepersist_spec saved windows 1200-2000 wide on the harness's 1024x768 screen (6 examples). Its fixture now stands up a 2560x1440 screen (tunable, size AND the anchor rectangle, guarded by screenMismatch), saved tops sit on-screen, and the undragged-window example uses a 1.25 scale: that window's CENTER still resolved against a 1024x768 parent after the resize (cause not established), so at 2x it hung off the left edge and the clamp correctly moved it. TOGBank's own scale listeners were searched for the cause: only the standalone Requests window re-runs PersistWindow (Requests_OnUIScaleChanged, Modules/UI/Requests.lua:1345); Browse's (Browse.lua:1181) only re-lays its lists. A harness helper that resizes the screen coherently is requested on WoWAPITesting's inbox (harness thread 36d5a55f). Delivered and adopted later in this release (the harness-pin entry above): the _fixedRect write and the 1.25 scale are gone.
  • Coverage moved, stated so nobody looks for it here: fullsync_spec's "four cold viewers" now drives four requesters through DeltaSync's replay path, so the simultaneous-login-broadcast storm is covered only by congestion_spec's storm example.
• LAGW-OLD-FLOOR-001: an outdated LibAceGUIWidgets (no ForgetWindow, before MINOR 35) is handed no window floor (Peer Review on self-audit 587af509, finding 2). Such a copy never gives a released ClearFrame's resize handle its own floor back, and AceGUI's frame pool is shared by every addon, so TOGBank's floor would have ridden a pooled frame into another addon's window. With no floor set there is nothing to leak. KNOWN COST: on an outdated library TOGBank windows can be dragged below their minimum; UI:PersistWindow prints one warning per session asking the player to update the library. browsepersist_spec: an outdated stand-in is handed width and height only and warns once across two windows; the control shows a current library still gets the floor as a function.

[v1.7.0] (2026-09-26) - The Shopping List

GuildShoppingList merged into TOG Bank (GSL-MERGE-001; the operator, 2026-09-25: "ok, we can do the gsl integration in v1.7.0"). Design and build order: docs/GSL_MERGE.md. The operator's standing rule for the whole merge: "i don't want to build new stuff as much as possible i want to use the libraries" -- LibProfessionDB, LibItemDB through Inventory/Resolve, LibAceGUIWidgets.

New Features

• GSL-MERGE-001 step 2: the shopping list's model (Modules/CraftList.lua, new; TOGBankClassic_CraftList). What the guild wants crafted (a LibProfessionDB recipe) or gathered (a plain item, GuildShoppingList's Miscellaneous bucket), and the gather window. Writers (D2, the operator: "so the GSL player is a tag like gbank, use what is in the GSL app for that. the officers/gm should be able to edit the shopping list"): the GM, officers, and a home member whose PUBLIC note carries [GSL] (GuildShoppingList's own marker, GuildShoppingList.lua:799). Sync: four fields on the existing guild-settings payload -- craftList and craftListStamps merged entry by entry (a removal keeps its stamp; the bankerOwners rule), gatherStart / gatherEnd field by field. BroadcastSettings now sends from a [GSL] member, and ApplyRemoteSettings takes ONLY the list from one. The four fields are left out of SettingsCanon, so a v1.6.1 client with the same settings still matches (KNOWN COST in the code). A sister guild's list never crosses. Demand: a recipe expands into its reagents (LibProfessionDB, by item id) for what the HOME bank does not already hold of the crafted item; each reagent row nets the home bank (Guild:GetAltItemTotal) and your bags (C_Item.GetItemCount); items are named through Inventory/Resolve. ProfessionDB is a required dependency in every TOC and .pkgmeta (slug libprofessiondb). No window yet (step 3). Offline: Tests/craftlist_spec.lua (new, 20, on the real ProfessionDB Vanilla Alchemy data; the canon example goes red with the exclusion removed). Locations: Modules/CraftList.lua, Modules/Guild.lua, the TOCs, .pkgmeta, .luacheckrc, .luarc.json.
• GSL-MERGE-001 step 3: the two tabs (Modules/UI/CraftList.lua, new; TOGBankClassic_UI_CraftList). In the Guild Bank window, after the existing tabs so none moves: Shopping List for everyone (a View dropdown between what the guild wants, with its profession, how many the home bank holds and how many are still needed, and the reagents that takes, with Needed, Bank, You and Missing), and List Setup for the list's writers only (a search across LibProfessionDB recipes and, with no profession picked, LibItemDB items; a How many box; left-click puts an entry on the list or changes its count, right-click takes it off; the gather window on LibAceGUIWidgets' date picker). Built on the existing machinery, per the operator's rule: Browse's tab group and BuildSimpleStrip, UI/RowList, the shared chrome and status line, items named through Inventory/Resolve. Profession names and icons come from the client's C_TradeSkillUI.GetTradeSkillDisplayName / GetTradeSkillTexture (present in the Era, TBC and MoP trees). Browse:SyncTabs re-renders the tab strip on open when the set of tabs changed, since the writer role moves with the roster, and Browse:RememberedTab now validates its fallback too, so a demoted writer is not reopened on a tab the strip no longer carries. A refused edit says whether the player may not edit the list or the list already holds that count. The quality-coloured item name is one helper, UI:QualityText, which the Browse rows now use as well. Offline: Tests/craftlistwindow_spec.lua (new, 13, on the real Guild, ProfessionDB Alchemy data, LibItemDB for the item search, AceGUI, Browse and the date pickers; red with List Setup shown to everyone, and red with every open rebuilding the tab strip). C_TradeSkillUI comes from the harness (WoWAPITesting ee56a45, thread b856f5a5). Modules/UI/CraftList.lua, Modules/UI/Browse.lua, the three TOCs, .luacheckrc, .luarc.json.
• GSL-MERGE-001 step 4: the Reagent Tracker (Modules/UI/CraftTracker.lua, new; TOGBankClassic_UI_CraftTracker; D5). GuildShoppingList's detached overlay as a small window of its own, showing the Shopping List tab's reagent roll-up (UI_CraftList:ReagentRows, the same columns), sorted by Missing, persisted as tracker. It repaints on a list change, on BAG_UPDATE_DELAYED while shown (registered on open, dropped on close, as ItemHighlight does), when bank data lands (UI_Inventory:RefreshSoon now fans out to it on its own debounce, since it is often up with neither bank window open), and on a roster change (Browse:Refresh). It opens from a Tracker button on the Shopping List strip and from the new /togbank tracker. It is deliberately left out of Escape: a gathering overlay that the Escape stand-in counted would close on every Escape pressed in the world instead of the game menu opening. Offline: six examples in Tests/craftlistwindow_spec.lua; the bank-data one was red before the RefreshSoon fan-out existed, and the bag-update one fires the real event through the harness and asserts the registration on open and its removal on close (Peer Review F5, red with the registration taken out). Modules/UI/CraftTracker.lua, Modules/UI/Inventory.lua, Modules/UI/Browse.lua, Modules/UI/CraftList.lua, Modules/Chat.lua, the TOCs, .luacheckrc, .luarc.json, README.txt.
• GSL-MERGE-001 step 5: GuildShoppingList's list comes across, once (D9; CraftList:ImportGSL, CraftList.ResolveGSLName). On a roster refresh (Guild:RefreshOnlineCache, both the library and the legacy path, beside the officer-floor publish), a client that may edit the list and whose guild has never had one (no entries and no stamps) reads GuildShoppingList_SavedItems ("<name> x<N>") and its gather dates (GuildShoppingList_Config.GatherStartDate/GatherEndDate), writes the entries, and publishes once. Each name is matched through LibProfessionDB by recipe name, then effect, then both with a trailing parenthetical stripped, and last as a plain item by its exact name through LibItemDB (GuildShoppingList's Miscellaneous turn-in items). A spelling two recipes or two items share maps to neither (REQ-001); an ambiguous name does not fall through to the effect. Repeats are summed. The entries are stamped CraftList.IMPORT_STAMP (1), older than any TOGBank edit, because the first roster refresh can come before this client has heard the guild's list: stamped "now", an import would have overwritten the guild's counts and brought back entries it had removed (Peer Review F1 on self-audit 13180dbf). What matches nothing is reported with Output:Response, which no setting mutes, and the names are kept in db.global.gslImport.unmappedNames (F2). GuildShoppingList's variables are account-wide, so the import runs once per account. KNOWN COST in the code: on an account with writers in two guilds the list goes to whichever guild such a character logs into first; the gather dates ride the ordinary settings stamp, so the race can still let imported dates replace a guild's newer ones; and on a client that has never held the guild's settings, an import racing an existing list is merged into it rather than refused (Peer Review on daa08281). The migration lives in Modules/CraftList.lua, beside the list's other writers, rather than Modules/Database.lua as the design first said, because it writes through the list's own rules. Offline: six Tests/craftlist_spec.lua examples on the real ProfessionDB Alchemy and Enchanting data and the real LibItemDB, including the race (red with the import stamped now), a recipe effect two recipes share (Health +5, red with the ambiguity marking removed) and a name nine items share (Punctured Voodoo Doll, red with the two-ids refusal removed); the importing examples went red with the roster hook removed. And one Tests/xguildfleet_spec.lua example: a home [GSL] bank character's list reaches its guild, and neither a whisper of it nor a federation pull puts it in the sister guild (red with the XGUILD-SETTINGS-001 gate taken out). Modules/CraftList.lua, Modules/Guild.lua.
• GSL-MERGE-001 step 7: order from a Shopping List row (D6; UI_CraftList:OnListRowClick, UI_CraftList:OrderSource). A left click on a row in either view (a wanted item, the item a wanted recipe makes, or a reagent) opens the existing request dialog on the bank character holding the most of it. The row comes from Browse:BuildRows, the Browse tab's own rows, so a sister guild's bank character counts as it does there, and the click goes through Browse:OnBrowseRowClick, so ordering closed, not-for-sale and the request limit refuse it exactly as on the Browse tab. A view-only bank and a hidden row are skipped when choosing. An enchant (no item) and an item no bank character holds say so on the status line; Shift/Ctrl-click links or previews; a right-click does nothing. A tie goes to the first bank character in name order. Offline: eight Tests/craftlistwindow_spec.lua examples (six red before the change). The spec's bank fixture now runs the real Guild:RebuildBankerRoster from the banker's gbank note, the path a live client takes on GUILD_ROSTER_UPDATE, rather than a hand-set roster (Peer Review on self-audit 93c01c04). Modules/UI/CraftList.lua.
• WANTED-BREATH-001: Still needed breathes (the operator: "can you have the wanted stuff 'breath' using the function from the widgets library?"). On the Shopping List tab's Wanted view the Still needed number pulses through LibAceGUIWidgets' W:Breathe (the call the help icon, the cancelled date and the stay-online line make) while it is above zero, and stops at full alpha when the bank covers it. The breath animates a frame, so the column builds its own cell (a frame holding the number) and the list's onRowRender paints it; a pooled row reused for a met entry is stopped, not left dim. Offline: one Tests/craftlistwindow_spec.lua example (red before). Modules/UI/CraftList.lua.
• SETUP-SPLIT-001: List Setup split, TOGPM's shopping-list box on top (the operator: "split the list setup into whats on the list on top and the selection bit on the bottom. lets make it look like TOGPM's 'shopping list' bit at the top of the professions page"). What is on the list is now a bordered box at the top of the tab (UI_CraftList.SetupBox, a RowList): icon, quality-coloured name, type, then TOGPM's - count + and red x. It is capped at 40% of the tab's height and scrolls beyond that, as TOGPM's is. The search results sit under it, and an empty search no longer lists the list there. - from 1 and x take an entry off; every control goes through the list's own writers and the same may-edit gate as a result click (UI_CraftList:StepEntry / RemoveEntry / WriteEntry). TOGPM's "!" alert and reagent expander are not carried: they are per-character craft features with no meaning for a guild list. OnSetupRowClick's branch for on-list rows could no longer be reached and is removed. Browse:AnchorLists calls UI_CraftList:AnchorSetup to put the box over the results. Offline: the old "empty search lists the list" example is replaced by four (box over results, the controls, a member refused, the height cap). Modules/UI/CraftList.lua, Modules/UI/Browse.lua.
• GATHER-BREATH-001: the gather dates breathe (the operator, of the Shopping List tab's "3 wanted -- gathering 2026-09-26 to 2026-10-03": "the date isn't breathing"). That line was what "the wanted stuff" in WANTED-BREATH-001 meant, and I built the Still needed pulse instead; the Still needed pulse stays unless the operator wants it gone. The status bar gains Instance:SetLeftBreathing(on): the left text is AceGUI's FontString on the pooled status background, which cannot host the breath (BREATH-001), so the first request moves it onto a frame of the bar's own (cached on the pooled background, as the centre host is, POOL-CHROME-001) and the bar's release hands it back, still. UI_CraftList:Draw asks for the breath while the Shopping List tab shows a gather window; HideLists, which every tab switch passes through, stops it, so no other tab's line breathes. Offline: two Tests/statusbaryield_spec.lua examples on the real bar (the breath and its release) and one Tests/craftlistwindow_spec.lua example (on, both views, off on Browse). Modules/UI/StatusBar.lua, Modules/UI/CraftList.lua.
• SETUP-SPLIT-002: the box's scrollbar came in at 3 rows (the operator: "i don't mind the scrollbar, but we need to expand more than 3 items before we introduce it. how does TOGPM do it?"). TOGPM shows every row until its 40% cap and pads the section past the rows. My box was sized to exactly its rows, and RowList counts rows with floor, so one pixel lost to the client's pixel snapping dropped the last row and showed the scrollbar. The box now carries UICL.SETUP_BOX_SLACK (2). Offline: a Tests/craftlistwindow_spec.lua example shaves a pixel off the box and requires three rows and no scrollbar (red before). Not yet seen in a client. Modules/UI/CraftList.lua.
• GSL-SENDTO-001: the Shopping List tab says who to send items to (the operator: "on the shopping list tab, we need to show who is the GSL character, so people know who to send items to"). A "Send items to:" line on the tab's strip names every home-guild member with [GSL] in their public note (CraftList:GSLPlayers, the same test as IsGSLPlayer), online ones first in green, offline ones grey and marked, and tells a guild with none how to name one. Painted by UI_CraftList:Draw, which a roster change reaches, so online state stays current. Offline: one Tests/craftlistwindow_spec.lua example. Modules/CraftList.lua, Modules/UI/CraftList.lua.
• ITEM-SEARCH-001: every item is searchable on List Setup, whatever the profession (the operator: "we need to ensure the eko's that make juju's are on the list, why don't we just have the itemdb searchable on the GSL add list?"). The search skipped LibItemDB whenever the Profession dropdown named a profession, so an item no profession makes (the E'kos are Quest items) could not be found from a profession view. A profession now narrows the recipes only. Offline: a Tests/craftlistwindow_spec.lua example loads LibItemDB's shipped Quest data and finds all seven E'kos (12430-12436) with Alchemy picked. KNOWN LIMIT: LibItemDB's search is a plain substring, so "eko" without the apostrophe still finds nothing; a punctuation-insensitive match is requested from ItemDB (contract e860cddb848a). Modules/UI/CraftList.lua.
• SETUP-REAGENTS-001: made items show their reagents, TOGPM's expander (the operator: "i would like to do a TOGPM like thing, where each 'made' item shows the mats to make it. so we can say we need 5 flask OR x reagent y reagent etc. then it can generate a complete reagent list based on the end items required"). In the List Setup box a made item carries TOGPM's +; a click opens its reagents under it (indented, at per-craft count times the count wanted, no controls) and - closes them. The box is no longer sortable, as TOGPM's is not, so a reagent stays under its item. The complete roll-up across every end item is the Shopping List tab's Reagents view (CraftList:Demand), unchanged; a raw reagent can still be put on the list directly as an item. Offline: one Tests/craftlistwindow_spec.lua example. Modules/UI/CraftList.lua.
• SHOPLIST-SPLIT-001: the Shopping List tab laid out like List Setup (the operator: "the + and the shopping list totals need to be on the shopping list tab as well. the guild needs to see how many dreamfoil etc they need to send ... the shopping list tab needs to be like the list setup tab, the items on the top, but instead of a search bit on the bottom, the bottom rows should be all the reagents we need to make the items in the list"). The wanted items are a bordered box on top (unsorted, so an item's reagents stay under it) and the full reagent roll-up (CraftList:Demand, Needed / Bank / You / Missing) is always shown under it; the Show dropdown that switched between the two is removed. A crafted item's + opens its reagents at the count still to make (wanted less what the bank holds of it), so an opened item agrees with the roll-up. Because a row click orders the item (step 7), the + is a button of its own in the first column (buildArrowCell), not a row click as on List Setup. Both boxes are now one builder, one height rule (UICL:BoxHeight) and one expander (UICL.WithReagents), so the two cannot drift. The status line reads "N wanted, N reagents, N still missing". Offline: two new Tests/craftlistwindow_spec.lua examples; the four that drove the old dropdown now open the tab as it is. Modules/UI/CraftList.lua.
• GSL-BANK-001: the [GSL] character is the shopping list's banker (the operator: "we also need what is on the GSL player (bank/bags/mail) just like a banker. they are the 'special' banker for the shopping list" / "the missing tab should be what the GSL banker is missing"). One rule now decides what a guild note makes a character a bank, noteBankRole in Modules/Guild.lua, and the six sites that decide who is a bank ask it instead of testing for gbank on their own: GetBanks' fallback, _SisterBankers, RebuildBankerRoster, IsViewOnlyBank's fallback, and both roster builds (library and legacy) plus the sister roster. SenderHasGbankNote deliberately stays gbank-only, because it grants settings and donation authority, not bank status. An earlier draft of this entry said "every site", which missed it; and because the sister roster stores a [GSL]-only character as isBank, its sister-guild branch gave such a character that authority until it was changed to re-read the note for gbank (new Tests/xguild_spec.lua example). [GSL] in the PUBLIC note makes a banker flagged gsl (the field on memberRoster, read by the new Guild:IsGSLBank), so the GSL character's own client scans and publishes its bags, bank and mail through the banker pipeline. It is view-only unless the note also says gbank: the guild sends to it, and does not order the shopping list's stock back out. It therefore shows on the Browse and Bankers tabs as a view-only bank, and BankerStores strips [GSL] from its note. On the Shopping List tab the reagent roll-up gains a GSL column (CraftList:GSLHolds), Bank no longer counts the GSL character (CraftList:BankHolds, both through HomeBankersHold), and Missing is the need less what the GSL character holds; Bank and You are shown beside it, not netted out. The Reagent Tracker shows the same columns. Found while wiring it: my new rule made IsViewOnlyBank's fallback depend on being a banker, which broke the existing "GBANK VIEWONLY" case (the banker test is case-sensitive); the fallback keeps the marker's own answer. Offline: two Tests/craftlistwindow_spec.lua examples (the note rule, and the GSL column); the roll-up and tracker examples in craftlistwindow_spec and craftlist_spec now expect the new Missing. Modules/Guild.lua, Modules/CraftList.lua, Modules/UI/CraftList.lua, Modules/UI/CraftTracker.lua.
• GSL-YOU-001: You and GSL disagreed on the GSL character (the operator 2026-09-26, with a screenshot of 120 against 160 Dreamfoil: "when i'm on the GSL toon, the GSL and You columns should match"). CraftList:YouHold was C_Item.GetItemCount(itemID), bags only. A bank character (the [GSL] one included) now reads its own scanned total (Guild:GetAltItemTotal: bags, bank and mail), the number the GSL and Bank columns read for it; anyone else gets bags plus bank (includeBank, Classic Era ItemDocumentation.lua:242). The mail half is only as fresh as the character's last mailbox visit, since the inbox can be read only while it is open (Modules/Bank.lua:303). Offline: a Tests/craftlistwindow_spec.lua example with bag and mail stock on the GSL character. Modules/CraftList.lua.
• TRACE-REALM-001: /togbank dev trace <name> traced the wrong realm (the operator's own trace, 2026-09-26). A bare name normalised to the PLAYER's realm, so "trace Togcloth" typed on Old Blanchy traced Togcloth-OldBlanchy, which does not exist, and printed the real Togcloth-Azuresong as "banker on the roster: NO". A bare name now resolves to the one bank character (then the one roster member) with that name on any connected realm; an ambiguous name keeps the old reading. Offline: a Tests/trace_spec.lua example. Modules/Chat.lua.
• ESC-SPLIT-001: one window list was doing two jobs (Peer Review, inbox a0556886). UI.ALPHA_WINDOWS was both the transparency-slider list and the Escape registry, so the Reagent Tracker could not get a slider without also being closed by Escape. Escape now reads its own UI.ESCAPE_WINDOWS (the seven existing windows); the tracker is in ALPHA_WINDOWS only, so it has its own Appearance slider and is also moved by Recenter Windows. Offline: a Tests/windowalpha_spec.lua example pins the two lists against each other, and a Tests/craftlistwindow_spec.lua example drives the tracker's slider and checks the Escape stand-in stays hidden with only the tracker open; both went red with the tracker added back to the Escape list. Modules/UI.lua, Modules/UI/CraftTracker.lua.

Bug Fixes

• XGUILD-SETTINGS-002: a sister guild's officer still had settings authority behind one branch (Peer Review F4 on self-audit 13180dbf, traced by them). SenderHasGbankNote, SenderIsGM and SenderIsOfficer each read the SENDER'S guild roster, so they answer true for a sister guild's bank characters, officers and GM -- right where they authorize that guild's bank data, wrong for settings. The XGUILD-SETTINGS-001 branch at the top of ApplyRemoteSettings was the only thing between another guild and every officer setting. SettingsSenderAuthorized now refuses a sender placed in another guild (the same test that branch uses), and ApplyRemoteSettings asks it rather than repeating the three helpers, so the refusal holds even without that branch. The helpers are unchanged. Offline: the Tests/xguildfleet_spec.lua list example asserts the home bank character has a gbank note but no settings standing from the sister client (red without the new line), and with the sister branch disabled the home list still stays out. Modules/Guild.lua.
• SETTINGS-TIE-001: two clients holding different settings at the same stamp never agreed (Peer Review F1 on the step-2 self-audit). An equal stamp was taken from whichever payload arrived last, so two officers' writes in the same second flipped each client back and forth; the [GSL] path also spelled the same rule differently. Both paths now go through one adoptSettingsField: a newer stamp wins, an unset field is seeded, and an equal stamp goes to the greater canonical value, so every client picks the same winner. Found while wiring it: ApplyRemoteSettings moved the held version BEFORE judging the fields, so a field with no stamp of its own read as exactly the payload's stamp; it now moves after. KNOWN COST in the code: a plain member that sees a tie while the writer holding the greater value is offline asks the other writer once a minute until the two meet. The shopping list's entry merge (CraftList:Merge) and the bank-character owner merge (Guild:MergeBankerOwners, Peer Review on the follow-up audit) had the same defect per entry: on an equal stamp the list now takes the greater count and the owners the greater name. KNOWN COST in both: a removal or a clear counts as nothing and so always loses a tie, and comes back. Modules/Guild.lua, Modules/CraftList.lua; Tests/craftlist_spec.lua (five examples; the settings and owner ones go red on the old rule).
• POOL-CHROME-001: a released TOG Bank window left its "?", gear, share button and status-bar text on AceGUI's pooled frame (Peer Review F5, which asked for a trace; the trace found the defect, wider than asked). UI:DressWindow cached the icons on the frame and never hid them, so the next window built on that frame, another addon's included, showed our help icon and a working gear. The share button's OnShow hook still showed it on a bank character. StatusBar:Attach built a NEW centre frame and right-hand text on every attach, left the old ones shown with their last text, and stacked one more OnShow hook per life that redrew a dead bar. Now: one release dispatcher (UI:OnWindowRelease, because AceGUI holds a single OnRelease per widget) runs the chrome's and the status bar's clean-up. The bar's sections are cached on the pooled bar and reused, and its show/hide hooks are set once per frame and read the frame's current bar. Seen offline only, not in a client. Modules/UI.lua, Modules/UI/StatusBar.lua; Tests/windowchrome_spec.lua (red without the release step).

Internal

• Changelog archive: the whole v1.6.0 section moved to CHANGELOG_ARCHIVE.md. Forward part A8a (A5a plus craftlist_spec and craftlistwindow_spec) replaces A5a in the gate.
• The real LibItemDB in craftlist_spec (Peer Review F2): one describe loads the installed ItemDB with its shipped Vanilla Miscellaneous, Consumable and enUS name data and asserts both entries resolve through it; the existing naming example pins the no-library client fallback.
• CurseForge page checked with the harness's cfhtml.lua (declared as the writ suite cfhtml): well-formed, 64,542 characters. It flagged two CurseForge links (TOG Profession Master, TOG Tools) against the Discord-only rule for player pages; both names are now plain bold text. docs/Curseforge_Description.html.
• The whole-client test env was missing Modules/CraftList.lua. Tests/env_togbank.lua's MODULE_ORDER is a second copy of the TOC's load order, and every fleet client built from it silently dropped a heard shopping list, so a fleet example about the list would have passed for the wrong reason. Added where the TOC loads it, with Modules/Usable.lua, which it also lacked. MODULE-ORDER-001 (Peer Review F6): a Tests/wiring_spec.lua example now reads the Era TOC and requires MODULE_ORDER to equal its non-UI module lines in order (red with Usable.lua removed).
• Release docs for v1.7.0: README.txt brought level with the CurseForge page and the code -- version 1.7.0, the shopping list in Key Features and the tab lists, a [GSL] setup section, the You column counting bags and bank, List Setup (not the Shopping List tab) setting the gather dates, the Reagent Tracker's transparency slider, a v1.7.0 compatibility note, Changelog Highlights and ProfessionDB under Libraries. The CurseForge page gains ProfessionDB under Required (it was a TOC dependency and missing there), List Setup in both tab lists, the [GSL] setup step, the You column wording, and two New notes: older versions keep syncing but do not see the list or the [GSL] stock, and ProfessionDB is required.
• GSL-MERGE-001 step 6, TOGBank's half: a "Guild Shopping List" section in the CurseForge page's Core Features (telling a GuildShoppingList guild to keep that addon enabled until a list writer has logged in once, since the import can only read its saved variables while it loads), the Shopping List tab in the page's tab lists, and a Shopping List section in README.txt. GuildShoppingList's final release is requested from its own workspace (contract 27dae11a3e42). docs/Curseforge_Description.html, README.txt, docs/GSL_MERGE.md.
• Harness pin 23d11b1 -> ee56a45 (WoWAPITesting delivered C_TradeSkillUI, thread b856f5a5). The spec-local C_TradeSkillUI stand-in in Tests/craftlistwindow_spec.lua is deleted; the spec now runs on the harness's.
• Gate: forward 470 + 504 + 961 = 1,935, reverse 537 + 711 + 687 = 1,935, none failed.

[v1.6.1] (2026-09-25) - Sister-Guild Failover

New Features

• ESC-001: Escape closes the Guild Bank window, and a setting turns Escape off. A player: "Bank doesnt close with esc press, but togpm does"; the operator: "it should have esc close the window by default." The escape frame was only CREATED by the first open of Inventory, Search, Requests or Donations, so the Guild Bank window opened on its own never had Escape; closing Inventory also hid it and took the Guild Bank window down. Now one stand-in (TOGBankClassicEscProxy) on UISpecialFrames, shown while any window is open, closes them all. Appearance setting "Close windows with Escape" (db.global.closeOnEscape, default on). _G.TOGBankClassic is gone. Offline: Tests/escape_spec.lua (new, 10). Locations: Modules/UI.lua, Modules/Options.lua, every Modules/UI/*.lua window.
• XGUILD-SEARCH-001: the Bankers search matches a sister guild's name. The operator, 2026-09-25: "the guild tag name needs to be searchable, so i can search for bankers in the sister guild." Browse:FilterBankerRows passes the row's guildName to SearchMatch. Offline: Tests/browse_spec.lua (+1). Locations: Modules/UI/Browse.lua.
• SSYNC-001: /togbank ssync [<name>] starts a sister-guild sync now. The operator, 2026-09-25: "do we have a /togbank command to start a cross guild sync? i need that, especially for testing, and if i could do /togbank ssync <name> that would be helpful. it could do a roster lookup so i can be lazy with the realm." With no name, each listed sister guild's usual peer is asked at once, past the cycle's pending and silent-for-an-hour waits. With a name, that sister member is found in the sister rosters (realm optional, any case; Guild:FindSisterMembers) and asked directly -- the hash-list ask and the requests index behind it, the cycle's own AskFederationPeer -- with any "left alone" mark on it cleared. It refuses, sending nothing and saying why, when the Sister-guild bank is off, no sister guild is listed, the name is not a sister member, it matches characters on more than one realm, or the member is offline. Offline: Tests/xguild_spec.lua (one example). Locations: Modules/Guild.lua (SisterSyncCommand), Modules/Chat.lua, README.txt.
• XGUILD-NUDGE-001: a GreenWall guild hears you open the bank, and its members pull at once (docs/XGUILD_SYNC.md 4.6, build-order step 7 -- the optional last step, gated on steps 2-6 being seen in game, which the operator closed on 2026-09-16). GreenWall's bridge cannot carry the sync (a CHANNEL send is hardware-gated on Classic, a segment is 255 bytes, the handler is not told the guild -- 1.1 of the design note), but ONE line from a click can ride it: Browse:Toggle -- the minimap button, /togbank, the legacy window's Browse button -- sends hlq:<addon version> through GreenWallAPI.SendMessage when the sister-guild bank is on, GreenWall is loaded and its guild channel is JOINED (a segment GreenWall cannot send is parked for its next flush, which may run from its timer, so an un-joined channel refuses rather than risks a blocked send), at most once per sync cycle. A federated client hearing it (GreenWallAPI.AddMessageHandler, registered from Core's init) notes the version and treats the line as a TOGBank message from the sender (Guild:UpdateOnlineMember(..., "greenwall-nudge")), so OnFederationPeerProven asks that guild for its hash list NOW instead of on the next ten-minute cycle -- after Guild Roster places the sender in a LISTED sister roster; echo, a home guildmate and a name only GreenWall's confederation knows are ignored. Nothing runs without GreenWall (## OptionalDeps: GreenWall in every TOC, so it loads first when present). Offline: Tests/xguild_spec.lua step 7 (three examples; the API's signatures pinned against the installed GreenWall/API.lua). Not seen in a client. Locations: Modules/GreenWall.lua (new), Modules/UI/Browse.lua, Modules/Guild.lua, Core.lua, the TOCs, Tests/env_togbank.lua, .luarc.json, .luacheckrc.

Bug Fixes

• POOL-CLUSTER-001: a released Requests window handed on its broom and envelope. ReleaseWindow (role change, or the body moving into the Guild Bank tab) returned the frame to AceGUI's shared pool with the cluster and officer overlay still shown; they now hide first. Not seen in game. Tests/requestsactions_spec.lua +1. Locations: Modules/UI/Requests.lua.
• REQ-GATE-001: a stranger could pull the guild's request history (Peer Review on audit 31294783, F5). requests-index / requests-by-id now answer only IsInCurrentGuildRoster senders (a guildmate, or a sister member while the switch is on). Also F4: SendBankerOwnersTo skips the receiver's own guild's entries. Tests/xguild_spec.lua +1. Locations: Modules/Chat.lua, Modules/Guild.lua.
• XGUILD-SETTINGS-001: officer settings crossed between sister guilds, and who runs each bank character mostly did not. The operator, 2026-09-25: "the banker metadata isn't syncing though", then "we should NOT be syncing the officer settings between sister guilds, this would allow any guild to target another guild and force it into being a sister." A sister guild's bank character passed SenderHasGbankNote (the sister roster's note), its officers SenderIsOfficer (member.isOfficer) and its GM SenderIsGM (rank 0), so ApplyRemoteSettings adopted that guild's request limit, shop, discount, Sister-guild bank switch and rank floor as OURS whenever its stamp was newer. And AnswerFederatedAsker sent settings only when the answering client was a bank character or officer, while a pull usually asks a plain member -- so the owners rarely crossed. Now a sister member's payload yields only its own guild's bank characters' owners (merged entry by entry, the field stamp moved with it), from ANY member; the answer to a sister guild's ask is the owners alone (Guild:SendBankerOwnersTo), from any client. Supersedes docs/XGUILD_SYNC.md D1/D4/D6 on settings. Offline: Tests/xguild_spec.lua (+2, one existing example rewritten), Tests/xguildlive_spec.lua step 6 (the request limit no longer crosses). Locations: Modules/Guild.lua, docs/XGUILD_SYNC.md.
• SCALE-DOCK-001: a Window and Text Size change pulled Search and Requests off the Inventory window. Peer Review's residual on 593238c3: the scale listener re-applies AceGUI's status table, whose ClearAllPoints dropped the hand-made dock. UI:SetPersistedAnchor lets a window re-dock right after. Offline: Tests/browsepersist_spec.lua. Locations: Modules/UI.lua, Modules/UI/Search.lua, Modules/UI/Requests.lua.
• SCALE-FLOOR-002: a released window could still be resized by a Window and Text Size change. LibAceGUIWidgets now registers its own scale listeners -- the resize handle's and PersistWindow's record (widget._lagwPersist) -- and both went into AceGUI's addon-wide pool with the frame. UI:ForgetPersistedWindow now removes the record's listener (through the library's public OnScaleChanged(owner, nil); the library has no forget call, asked for on its inbox 3d96003a) and puts a ClearFrame's handle back on the bounds it had before UI:PersistWindow raised them (recorded there as handle._togOrigBounds). The handle is NOT destroyed: it is made once in the ClearFrame constructor, and my first version of this fix destroyed it, which would have left the frame's next owner with no resize grips -- caught by this session's own audit before release. Also browsepersist_spec's SCALE-DOCK example used a width under the doubled floor, which the library now correctly raises. Offline: Tests/browsepersist_spec.lua (+1 ClearFrame example). Locations: Modules/UI.lua, Tests/browsepersist_spec.lua.
• SETTINGS-STALE-001: an officer whose client still held an old or default setting could push it to the whole guild by changing a DIFFERENT setting. Found live 2026-09-25: the Sister-guild bank was unticked guild-wide, so the sister guild's inventory handshakes were answered "busy" by every client (a non-member fails isValidPeer, Modules/P2P.lua:136) while its request queries -- which have no membership gate -- kept crossing. The operator: "we need to make sure someone logging on for the first time doesn't uncheck it, as that's the default. how do we guard this?" SETTINGS-004's per-field stamps diff a write's payload against a snapshot of the last settings sent or applied, and that snapshot was EMPTY at two moments -- a session's first write before any settings were heard or sent, and every write after settings were RECEIVED (ApplyRemoteSettings cleared it) -- and an empty snapshot stamped EVERY field as freshly written. Now the snapshot is the held settings from the moment the record is bound (Guild:Init) and whenever settings are applied (Guild:SnapshotSettings), tied to that settings table; a write that still finds none diffs against the DEFAULTS (Guild:DefaultSettingsFields), so a default value is never stamped by a write that did not change it. KNOWN COST: v1.6.0 clients still have the old behaviour until they update, so an officer on v1.6.0 can still do this. Offline: Tests/settings_spec.lua, two examples, both red with the fix reverted. Locations: Modules/Guild.lua.
• XGUILD-BOUNCE-001: a sister-guild bank character who logged off stayed "online" to the other guild, which kept asking them for the bank. The operator, straight after v1.6.0: "build complete end to end tests for the sister guild sync, still having issues with that". Guild Roster never hears "X has gone offline." for another guild's member: its presence for one is a sighting stamp that only ages out (PRESENCE_TTL, 900 s) -- and its guild relay carries every name still inside that window and the receiver re-stamps them at receive time (LibGuildRoster-1.0.lua:3976, :4194, :3297, on a 270 s relay interval), so a member who logged off reads online for as long as two guildmates keep relaying (filed to Guild Roster's inbox 2026-09-17). The one authoritative "not online" a client gets for them is the server bouncing a whisper (ERR_CHAT_PLAYER_NOT_FOUND_S), which Events:CHAT_MSG_SYSTEM already turned into UpdateOnlineMember(X, false) -- a flag on memberRoster that IsPlayerOnline, FederationPeer and _AddSisterMembers never read for a sister member. So the sister viewer whispered its hash-list ask to the logged-off banker, waited the whole cycle, marked them silent for an hour and only then asked the other member, while the Bankers tab said "yes". Now the bounce is remembered (Guild.sisterBounced, frame clock) and outranks the library's sighting for the library's own window (Guild:SisterSightingHolds, read by all three sites; the library's stamps are NOT compared against it, because a relay re-stamps within the minute -- measured); a TOGBank message from the member clears it at once; and when the bounced whisper was our ask, the ask is withdrawn, the member is left alone for FEDERATION_SILENT_FOR, and the guild is asked through FederationPeer's next choice NOW (Guild:OnSisterMemberBounced). Design: docs/XGUILD_SYNC.md D9. Offline: Tests/xguild_spec.lua (bounce, re-ask, clear, window) and Tests/xguildlive_spec.lua step 8 on the real transport stack. Locations: Modules/Guild.lua.
• LINKCLICK-001: a modified click on an item link does what it does everywhere else in the game. The operator: "what shift+click and ctrl+click does to the links in the addon? i'd like it to mirror game functionality". Read in both client trees this addon ships (classic_era, classic_anniversary): a modified click on any item link runs HandleModifiedItemClick (Blizzard_ItemButton/Classic/ItemButtonTemplate.lua:137) -- IsModifiedClick("CHATLINK") (Shift by default, but the PLAYER'S binding) inserts the link through ChatFrameUtil.InsertLink (the open chat box, else the auction house's search box, else a macro being edited), IsModifiedClick("DRESSUP") (Ctrl by default) previews it with DressUpItemLink. The bare ChatEdit_InsertLink exists on both clients ONLY as an alias of the real function in Blizzard_DeprecatedChatInfo/Deprecated_ChatFrame.lua:43, behind loadDeprecationFallbacks, and nothing in either tree calls it. Four defects against that: (1) the item slots (UI:EventHandler, Inventory and Search windows) read IsShiftKeyDown / IsControlKeyDown, so a player who rebound "chat link" or "dress up" got nothing; (2) the Browse and Shop rows handled Shift only, and a Ctrl-click REQUESTED the item; (3) every insert called the deprecated alias, nil with that CVar off; (4) Events:RegisterEvents hooked the alias for the "shift-click fills the Search box" feature (UI:OnInsertLink), so the hook never saw a shift-click on a bag item or a chat line -- only this addon's own slots -- and raised outright with the CVar off. Now UI:HandleLinkClick(link) runs the client's dispatcher (the same logic spelled out for a client without it), consumes the click whenever a link modifier is held so no plain-click action runs under one, and is what every site calls; UI:InsertLink prefers ChatFrameUtil.InsertLink; the hook is on the function the client calls, the alias only as a fallback, and neither present raises nothing. The mail window's shift-click (take everything, MAILCLICK-002) is unchanged: the game's own inbox has no modified-click handling (MailFrame.lua:828), so there is nothing to mirror. Offline: Tests/linkclick_spec.lua (new) and Tests/browse_spec.lua. Locations: Modules/UI.lua, Modules/UI/Inventory.lua, Modules/UI/Search.lua, Modules/UI/Browse.lua, Modules/Events.lua, .luarc.json, .luacheckrc.
• SCALE-FLOOR-001: with Window and Text Size under 100%, a window at its minimum had its filter strip past the border. A player's report against v1.6.0 (2026-09-17, the slider at 80%): "Window size at minimum has widget overflow after resizing/reloading text in config". Two defects, both mine from VISIBILITY-001. (1) LibAceGUIWidgets' PersistWindow multiplies a window's floor by the accessibility scale -- right for a window made of its own widgets, whose every region scales -- but TOGBank's floors are sums of FIXED widths too: the Browse strip's stock AceGUI dropdowns, edit boxes, checkbox and button (their template art is fixed; ScaleStockWidget grows only the text), the frame's insets, RowList's gutter. At 80% the Guild Bank window's floor fell from 784 to 627 px while the strip's first row still needed 738, so a window dragged to its minimum, or reloaded at a saved size the floor no longer raised, spilled. Now the pixel floor is never below the normal-size one (minW x max(1, scale), handed to the library as minW x max(1, scale) / scale so its own multiply lands there). KNOWN COST: under 100% a window cannot be dragged smaller than it can at 100%; the text shrinks, the minimum does not. (2) The library re-applies a floor on a scale change only through a resize handle, which an AceGUI Frame (every TOGBank window) has none of, so the bounds set at draw time stayed at the old scale until the next reload -- a window at the 100% minimum kept it at 200% with twice the text inside (finding filed to LibAceGUIWidgets' inbox 2026-09-17; TOGBank's own listener is the stand-in). UI:PersistWindow now re-runs itself on the scale signal for every window -- Inventory, Search, Requests, Browse, Mailbox -- raising a window sitting under the new floor and re-applying the bounds; nothing shrinks on a scale-down. Offline: Tests/browsepersist_spec.lua (three examples against the real library and the real Requests floor; two fail without the fix at 627 vs 784); forward A4 882/0, forward B2 936/0, reverse-order 1818/0 after it. Locations: Modules/UI.lua, README.txt.

Internal

• SUITE-HEAP-002, harness pin 23d11b1. browse_spec, requestsactions_spec and mailwindow_spec release their windows; env.releaseWindow drops Requests' frame caches. The GetClassColor stand-in and browse_spec's fake LibDBIcon are gone. Heap not measured.
• UI-OWN-TABLE-001: TOGBankClassic_UI is its own table, reading through to AceGUI-3.0; it was the shared library table itself (Peer Review 35130c29). Escape also closes the owner dialog.
• OWNERS-STAMP-001, harness pin 35f697b. One Guild:AdvanceOwnersStamp for both owners-merge routes (Peer Review 31294783). Harness 17d5211 -> 35f697b: GetClassColor stand-in, two specs on the rich layer, canonhash_spec loads Inventory/Sync.lua, windows released in two specs.
• USABLE-LIBITEMDB-001: "Usable by me" asks LibItemDB's ClassProficient (ItemDB v0.10.0, MINOR 27; ItemDB inbox 10a85af75c12). The five proficiency tables copied from Dibs and Usable.ArmorCap are deleted. One answer changed on purpose: a class off the tables may hold a shield. Offline: Tests/usable_spec.lua, Tests/browse_spec.lua (the installed library). Locations: Modules/Usable.lua.
• TOC-MISTS-001: a MoP Classic TOC, and the TBC TOC renamed _BCC -> _TBC (the operator, 2026-09-25: "make a mop toc", "update the _BCC toc to be _TBC"). New TOGBankClassic_Mists.toc at Interface 50504: the Era TOC byte for byte apart from the Interface line. Every required dependency already ships MoP (Ace3, VersionCheck-1.0, AceCommQueue-1.0, DeltaSync, LibAceGUIWidgets and LibDBIcon-1.0 list 50504; GuildRoster and ItemDB ship their own _Mists.toc, ItemDB with Data\Mists\ item data). Loads in a MoP client with no Lua errors (the operator, 2026-09-25, on a character in no guild); NOT VERIFIED: every guild feature -- nothing past loading has run in MoP, and the code has not been read against the MoP API. TOGBankClassic_BCC.toc is now TOGBankClassic_TBC.toc (same content), matching the fleet's GuildRoster_TBC.toc / ItemDB_TBC.toc. New lockstep guard Tests/wiring_spec.lua (TOC-LOCKSTEP-001): every TOC must equal the Era TOC apart from its Interface line; the ten specs that looped over two TOCs now loop over three. wow-version-replication.ps1 now also mirrors into _classic_ (MoP). Locations: TOGBankClassic_Mists.toc (new), TOGBankClassic_TBC.toc (renamed), .pkgmeta, wow-version-replication.ps1, CLAUDE.md, Tests/*_spec.lua.
• LIBDBICON-DEP-001: LibDBIcon-1.0 is a required dependency, no longer vendored (TOGTools inbox contract 1eab38f9, the operator's decision of 2026-09-25). An embedded copy only updates when TOGBank releases; as a CurseForge required dependency it updates for every player on its own. Libs/LibDBIcon-1.0/ is deleted, every TOC declares LibDBIcon-1.0 in ## Dependencies and no longer load the copy, and .pkgmeta lists the slug libdbicon-1-0. LibDataBroker-1.1 stays bundled on the operator's word ("change it to not get rid of libdatabroker"), although the standalone LibDBIcon also loads one (LibDBIcon-1.0/embeds.xml:6); LibStub keeps whichever copy has the higher MINOR. No Lua changed -- Modules/UI/Minimap.lua still resolves both through LibStub. KNOWN COST: a player installing by hand must also install LibDBIcon-1.0. Seen in an Era client 2026-09-25: the standalone (TOC ## LoadOnDemand: 1) loads and the minimap button shows; TBC and MoP not checked. Pinned by Tests/sendresult_spec.lua ("de-vendored LibDBIcon-1.0"). Locations: the TOCs, .pkgmeta, README.txt, CLAUDE.md.
• Peer review of this session's own audit, acted on (inbox 593238c3, reply 8e933d44). Three of its four points changed code. (1) c.offline is deleted from Tests/env_fleet.lua: one concept with two spellings, where the write-only one was the plausible-looking one, so a future guard written against it would have been true for a client taken offline through F.offline and false for one created with opts.offline. The reviewer also spotted that Tests/fullsync_spec.lua:864 was a hand-inlined F.offline -- the one call site that could not benefit from the change that gave F.offline its presence call -- so it now calls the helper and one of the three deletions came free. (2) GreenWall:Bridge no longer claims GetChannelNumbers()[1] is the guild channel, and the six-line caveat defending that claim is gone; it now accepts any joined channel. The reviewer's suggested #numbers == 0 was NOT taken, with a reason: GwChannel:is_connected calls a channel connected only when its number is non-zero, so a channel that exists un-joined reports 0 and a send on it is PARKED for the next flush -- the later non-hardware send this gate exists to prevent. The non-zero test is load-bearing; only the claim about which entry it is was wrong. (3) Nudge now stamps the cooldown on a refusal FROM GreenWall, so the one configuration that always fails there costs one attempt per cycle instead of one per window open; a refusal before the send still leaves the cooldown alone, so a client whose GreenWall loads a moment later does not wait a cycle. New example in Tests/xguild_spec.lua. The fourth point was routed rather than fixed: the desk's file move-span reports success after moving an incomplete span, which is writ's defect and is filed to its inbox. Locations: Modules/GreenWall.lua, Tests/env_fleet.lua, Tests/fullsync_spec.lua, Tests/xguild_spec.lua.
• GSL-MERGE-001, first step: GuildShoppingList read end to end, and the merge design written (docs/GSL_MERGE.md, new). The operator, 2026-09-16: "add to the todo list to merge the GSL addon into TOGBank, put it at the bottom", with the item's own first step being the read and the design before any code. Nothing is built and no code was touched. What the read settled: GuildShopping- List is a second, weaker implementation of machinery this addon already has -- its own guild-note role marker ([GSL]), its own whole-state newest-timestamp-wins broadcast, its own name-keyed cache of ONE player's bags and bank, its own 290-line calendar picker, its own two frames -- wrapped around one genuinely new idea, expanding a recipe into reagent demand. So the merge is a new Craft List tab on the rails that already exist (the stamped guild-settings payload for the list, the request log for orders, Guild:GetAltItemTotal for what the bank holds), not a port of its code. LibProfessionDB-1.0 already carries the recipe data keyed by spell id with reagents keyed by item id, which deletes GSL's eight vendored Recipes\*.lua tables and removes its name-matching at the root -- the same defect class as REQ-001. Build step 1 was then run rather than assumed (section 1.5): GSL's eight tables diffed against the library's Vanilla enUS data, 924 of 1,106 names matching exactly, and the 182 that do not are GSL's own shorthand (Crusader (enchant) against the library's Enchant Weapon - Crusader, whose effect field is literally Crusader; Blight (polearm) against Blight), names the Vanilla data does not carry and which look like later-expansion recipes (recall, not a lookup -- flagged in the document as its weakest claim), and the Miscellaneous bucket, which is seven Winterspring E'ko turn-in items and not recipes at all. So the shorthand cause is verified and large, and the migration gained two name fallbacks. Both open decisions were answered 2026-09-25: GM, officers and the [GSL] note tag edit the list, on a Shopping List tab plus an officer setup tab; the build is the release after v1.6.1. Locations: docs/GSL_MERGE.md.
• Harness pin bb80c1b -> 17d5211; both XGUILD-LIVE-001 stand-ins are deleted, and the fleet now takes a REAL bounce. (Two deliveries in the same move: 58ff098 below, and 4ca9566, which bumped LibItemDB-1.0's tocOmits 98 -> 99 after the drift this repo's newly-declared verify-libs suite caught at the previous pin. verify-libs is now 9/9, 0 failed.) WoWAPITesting delivered the two halves contracted yesterday (58ff098). (1) A logged-out client sends nothing: all three of the harness's send seams are gated on the sender being online, so Tests/env_fleet.lua no longer shadows each client's own C_ChatInfo while it is offline. (2) The server answers the sender of a whisper to a logged-out character with ERR_CHAT_PLAYER_NOT_FOUND_S, so F.bounceIfOffline -- which MANUFACTURED that line -- is gone. What replaced it is one frame per fleet client registered for CHAT_MSG_SYSTEM that hands the event to Events:CHAT_MSG_SYSTEM, because a fleet client never runs Events:RegisterEvents; the bounce is now real and only its delivery is wired here, with every line kept on client.sysLines so a spec can assert what the SERVER said separately from how the addon reacted. Two things cost time and are worth the next session knowing: that frame must be held on the client rather than in a local, because the frames registry keeps its candidate lists WEAK and a frame nothing references is collected and silently stops hearing its event (it received nothing at all, which reads exactly like the harness not firing); and Tests/xguildlive_spec.lua step 8 had to stop asserting WHICH whisper bounced. The stand-in only ever bounced whispers handed to it by TOGBank's own Core wrapper, so the first bounce was always that cycle's hash-list ask; the server bounces every whisper to a logged-out character, and the earliest here is the requests-index ask an earlier cycle staggered behind itself, so the viewer learns sooner and never spends the cycle's ask on the banker who is gone. The step now pins the outcome -- exactly one real bounce naming the banker, nothing further asked of it, every failover ask answered, the bank intact -- and the "our own ask bounced, re-ask at once" leg keeps its focused example in Tests/xguild_spec.lua step 4. Green in all three orders: forward A4 882/0, forward B2 939/0, reverse-order 1821/0. Locations: Tests/wowapi (pointer), Tests/env_fleet.lua, Tests/xguildlive_spec.lua.
• Harness pin 6181ba5 -> bb80c1b; the LAGW-RESIZE-FILE-001 stand-in is deleted. The contract filed earlier today (thread a6cfd6a8) was answered ALREADY SHIPPED: WoWAPITesting had landed 2aac908 -- LibAceGUIWidgets-Resize.lua in the LibAceGUIWidgets-1.0 manifest, in TOC order, tocOmits unchanged -- hours before the message, so the fix was a pin move and nothing else. The wrapper Tests/env_togbank.lua put round libs.load is gone. Three more behaviour changes ride the same move and needed nothing here, each checked rather than assumed: 8b36d23 routes the addon-message chatType case-insensitively (TOGBank's own report, inbox 22460a4b -- the recorded spelling is unchanged, so FILL-ALERT-001's "GUILD" assertions are untouched); bcd81e3 keys Settings.GetCategory on the numeric category id, which is what makes an AddToBlizOptions SUB-panel resolvable offline at all; 82387fa adds C_Spell.GetSpellName and C_Seasons. The Adoption log's warning for a library that splits a file -- a per-file load line is needed in any spec using wow.loadAddonFile as well as the manifest line -- does not reach this addon: the one by-path satellite load here (browse_spec.lua:326, the DatePicker) runs after libs.load. Green on the new pin in all three orders: forward A4 882/0, forward B2 939/0, reverse-order 1821/0. Locations: Tests/wowapi (pointer), Tests/env_togbank.lua.
• XGUILD-LIVE-001: the sister-guild sync, whole lifecycle, on the real transport stack (Tests/xguildlive_spec.lua, new). Four whole clients in two guilds over each client's real AceCommQueue, AceComm and ChatThrottleLib, in ONE example so each leg starts from the state the previous one left: both guilds log in on the clamp; one "Pull from" by name crosses the rosters, the gbank notes and presence on the library's own whispered pull and click-sent /who; both Bankers tabs list the other guild's bank, tagged; the sister viewer's cycle asks by whisper and holds the home bank with the author's canon and nothing between the guilds on GUILD; a sister member's order and the banker's fill (from Mail:ApplyPendingSend) cross by whisper at ALERT and are relayed once each way until all four clients agree; an officer's settings write and a who-runs entry ride the next ask without wiping the sister guild's own entry; the sister banker, which never pulled, reaches the home bank through its guildmate's relay and the home banker holds the sister bank; the home banker logs off (step 8, above); a quiet cycle opens no data leg. What its first runs found, 2026-09-17: (a) the invalid key to 'next' the previous session stopped on was MY spec bug -- next(F.held(hb, SB)) passed F.held's second return, the money total, as the key -- and steps 5-7 were passing behind it; (b) the harness lets a logged-out client's timers keep SENDING (six Guild Roster messages measured in the 660 s after logout, each re-stamping the banker online on the sister side) and drops a whisper to an offline character silently where the server bounces it -- both stood in by Tests/env_fleet.lua (F.offline shadows the client's own C_ChatInfo send seams, F.bounceIfOffline hands the bounce to Events:CHAT_MSG_SYSTEM) and filed as harness contract XGUILD-LIVE-001; F.presence no longer clears the library's stamps by hand, which had let the failover leg pass without the bounce the addon has to act on; (c) the Guild Roster relay finding and (d) XGUILD-BOUNCE-001, above. Two more spec gaps the same runs surfaced, both mine from earlier sessions: bankcollect_spec never loaded Inventory/Record and Inventory/Scan, which Bank's matchers have read since LINK-AUDIT-001 step 4 -- it passed in the declared halves on a Scan an earlier file left behind and failed alone (REQ-003, two examples); and LibAceGUIWidgets v0.2.0 moved its resize framework into LibAceGUIWidgets-Resize.lua, which the harness's hand-copied manifest does not list, so PersistWindow called a nil GetResizeHandle in 103 window examples -- Tests/env_togbank.lua loads the satellite when the library comes up without it (LAGW-RESIZE-FILE-001, harness contract filed; delete on adoption). The WHOLE forward run no longer finishes inside the desk's 15-minute budget, so the full-suite claim is the declared halves: forward A4 (altcompat..logwho incl. linkclick_spec, declared today) 879/0, forward B2 (mailbox..xguildlive, declared today) 936/0, reverse-order 1815/0 -- after LINKCLICK-001 as well.
• MDLINT-BACKLOG-001: every Markdown file in the repo lints clean. A whole-repo markdownlint (60 files) reported 100 violations, all pre-existing and all in five design docs nobody had linted since they were written: docs/MAIL_INVENTORY_DESIGN.md (31), docs/ORDER_FULFILLMENT_LOGIC.md (18), docs/MAIL_PERSISTENCE_INVESTIGATION.md (16), docs/TESTING.md (a UTF-8 byte-order mark, 8 headings, a table's pipe spacing, a double blank line) and docs/fulfill-button-plan.md (10 headings, a table). Every one was a blank line missing after a heading (MD022) or a table row's pipe spacing (MD060) bar the BOM and the blank; the text is unchanged. None of these files ships (docs is in .pkgmeta's ignore); the point is that the whole-tree lint is now a usable gate.

Releases v1.6.0 and older have been moved to CHANGELOG_ARCHIVE.md to keep this file under GitHub's 125,000-character release-body limit. Nothing was deleted; each section moved whole, at a version boundary.