Shopware Migration

Appearance

Executive Summary ​

  • SXSW velocity sustains but review pipeline crisis deepens: +5 stories to 68% (634/926) marks third consecutive week of +15/week acceleration after June crisis, but 14 stale tickets >30 days (up from 10 on 8/12) — led by SXSW-1125 at 63 days — signal infrastructure/schema work calcifying in review states with no visible deployment path. 292 stories outstanding still risks October delivery even at this improved pace.
  • CE Shopware holds at 91% but Deploying dominates pipeline: +4 stories with Emil's 116 commits and Scott's 45 commits sustain momentum, but 30 tickets sit in Deploying (median 27 days) awaiting production deployment — Blake's DMI/Lists work (28 commits) lands CE DMI subgraph + StockingLists UI wiring, but 14 tickets (CESW-817 and descendants) remain stuck >30 days in deployment limbo.
  • ECMD velocity collapses after last week's breakthrough: +2 stories to 23% (104/446) with just 16 commits — Sebastian/Gabriel reallocation effect from last week (45 stories, 45 commits) does not repeat, and new story creation drops from +80 to +2, suggesting backlog grooming paused. Root cause unclear from commit data alone.
  • Baker sustains progress, Homans/Peirce show hidden velocity: Baker adds +1 story to 48% (173/362) with 25 commits, marking third week of sustained single-digit gains. Homans and Peirce both remain at 0/0 in Jira (expected for newly-created boards) but feature parity matrices show +1 Done feature each this week (36/35 total) — Lino/Pablo/Joshua's 255 combined commits drive real checkout/PDP/cart work not yet reflected in boards.
  • Gemaire infrastructure blockers prevent test closure: Zero board movement (0/79, fifth week) despite 9+4 commits from Damian/Nikolay — GSHOP-165/156/118 (MiniCart, Branch Finder, Google Maps) block 9 downstream test/QA tickets, preventing any "Done" closures until root-cause fixes land.

Progress Overview ​

SXSW  [████████████████░░░░░░░░] 68% (634/926)
CESW  [█████████████████████░░░] 91% (614/672)
ECMDSW[█████░░░░░░░░░░░░░░░░░░░] 23% (104/446)
GSHOP [░░░░░░░░░░░░░░░░░░░░░░░░]  0% (0/79)
BSWM  [███████████░░░░░░░░░░░░░] 48% (173/362)
HMSW  [░░░░░░░░░░░░░░░░░░░░░░░░]  0% (0/0)
PPSW  [░░░░░░░░░░░░░░░░░░░░░░░░]  0% (0/0)

1. Shopware Platform - SXSW Initiatives ​

68% complete (634/926) — up +5 stories from 629 yesterday, +29 week-over-week. Third consecutive week of sustained +15/week velocity after June crisis, but 292 stories outstanding still risks October delivery at current pace.

Week-over-week: +29 stories closed vs +36 new stories added — net scope growth continues (total +7), but closure rate improved from last week's +19 closed vs +36 added. This is the third week in a row of sustained acceleration (was +29/+36 on 8/12), suggesting the post-June velocity collapse has finally stabilized. However, the program is still 6+ weeks behind mid-July projections, which anticipated 68% by early August.

Review pipeline aging accelerates sharply: 14 tickets now >30 days in review/dev states (up from 10 on 8/12), led by SXSW-1125 (Kubernetes/EKS Hosting for Domain Layer) at 63 days in Active Development. This ticket alone has no visible forward progress in 9 weeks — Chris Nitchie assigned, but no commits visible in tracked repos, no sub-tasks closed, and no timeline communicated. Infrastructure/schema work is calcifying: SXSW-1156 (Compare Shopware Commercial Plugin) at 21 days, multiple schema design tickets at 12-21 days (SXSW-1173, SXSW-1113, SXSW-1159). These are blocking downstream work across all BU boards — e.g., Homans/Peirce cannot decompose their boards until core ERP account schema (SXSW-1173) stabilizes.

Key blockers:

  • SXSW-1125 (Kubernetes/EKS) at 63 days: No commits, no sub-tasks, no timeline — infrastructure work invisible to velocity metrics but critical to all-BU rollout.
  • SXSW-1156 (Commercial Plugin Analysis) at 21 days: Rune Laenen assigned, blocks unknown downstream work (no explicit links, but likely Account/Order management features).
  • Schema design tickets stalling: SXSW-1173 (erpAccountSetDefault), SXSW-1113 (Parts List schema), SXSW-1159 (UOM verification) all >12 days — Paul Briscoe assigned to multiple, but no resolution timeline visible.

Repository activity: 58 commits this week from 12 contributors (vs 68 commits from 10 contributors last week) — slightly lower absolute volume, but broader contributor base suggests less reliance on single-person heroics. Paul Briscoe (39 commits, 10 closed tickets) and Ilir Spaho (37 commits, 4 closed tickets) dominate, both working on domain resolver implementations (Fees, Pay-on-Account, FBT, Quote migration). Rob Simmons (29 commits, 4 closed tickets) ships reference-only products and product restrictions. This is good depth — three high-output contributors shipping cross-BU platform work rather than one person carrying the board.

Risk: 292 stories outstanding at current 29/week velocity = 10 weeks to completion, pushing delivery to late October. This is acceptable if velocity holds, but any reversion to June levels (14/week) blows past year-end. Review pipeline aging is the real concern — infrastructure tickets stalling at 60+ days suggest systemic decomposition/handoff issues, not just capacity constraints.

6. CE Shopware (CESW) ​

91% complete (614/672) — up +4 stories from 610 on 8/18, +27 week-over-week. Holding steady after last week's -170 scope swing (backlog cleanup, not real work loss). Emil Stewart's 116 commits and Scott Kovach's 45 commits sustain momentum, but review pipeline dominated by Deploying status (30 tickets, median 27 days) — work landing in code but stuck awaiting production deployment.

Week-over-week: +27 closed vs +30 added — net scope growth minimal (+3), and closure rate healthy at 27/week. This is CE's normal operating velocity (was 27 closed on 8/12 as well), suggesting the chaos from 8/11's -170 swing is fully behind them. The board is stable and predictable again.

Deploying status dominates: 30 tickets sit in Deploying (median 27 days), led by CESW-915/914 at 41 days (Blake Saunders: conduit-ng cache-bypass and customer_company_id filter). These are not blocked — they're awaiting production deployment after successful staging validation. The 27-day median suggests a ~1-month deployment cadence, which is acceptable for infrastructure changes but creates visual clutter in the review pipeline (30 tickets = 5% of total board scope).

DMI/Lists work ships but deployment lags: Blake Saunders' 28 commits this week land CE DMI subgraph (CESW-817) and StockingLists UI wiring (CESW-818), closing the gap between CE Magento and CE Shopware DMI functionality. 14 tickets in the CESW-817 family remain >30 days in Deploying, but this is expected deployment lag, not a stall. The work is done; it's just not in production yet.

Emil Stewart sustains high output: 116 commits (highest single contributor this week across the entire program), closing 29 tickets — mix of QA work (CESW-989 Warranty, CESW-951 FrequentlyBought), infrastructure fixes (CESW-1717 Order Queue retries, CESW-1707 Cart item removal), and Salesforce/branch/document bugs. This is the same Emil who drove +50 closed tickets in prior weeks — CE's velocity is directly proportional to Emil's capacity, which is a risk if he becomes unavailable.

Scott Kovach ships punchout + user management: 45 commits close CESW-1595 (user add/edit/remove failures), CESW-1691 (Contact sync batching), and advance CESW-1325/1344 (punchout provisioning). This is real BU-specific feature work (punchout is CE-only), not just boilerplate sync.

Stale tickets remain a concern: 14 tickets >30 days, same list as last week (CESW-915/914/1191/1181/1179/1048/1047/1046/874/872/871/870/817) — all Blake Saunders' DMI/conduit-ng work in Deploying. No new stale tickets added this week, which is good (suggesting review pipeline is not backing up further), but the existing batch needs production deployment to clear.

Risk: CE is healthy — 91% complete, 58 stories outstanding, sustainable velocity. The Deploying backlog is visual noise, not a real bottleneck. The real risk is Emil's single-person dominance: 116 commits this week, 29 closed tickets last week. If Emil goes on vacation or gets reassigned, CE velocity will crater.

8. ECMD Shopware (ECMDSW) ​

23% complete (104/446) — up +2 stories from 102 on 8/18, +50 week-over-week. Minimal movement this week (16 commits) after last week's breakthrough (+45 stories, 45 commits from Damir/Lucian) — Sebastian/Gabriel reallocation effect does not repeat, and velocity collapses back to near-zero.

Week-over-week: +2 closed vs +2 added — net zero scope growth, which is a sharp reversal from last week's +50 closed vs +72 added. Last week showed the first real decomposition progress after 4 weeks of stall; this week reverts to stasis. Commit volume also drops sharply: 16 this week vs 45 last week. Root cause unclear from commit data alone — no major holidays, no obvious external blocker.

Repository activity: Just 16 commits this week (Damir: 11, Lucian: 4, Michelle: 1), down from 45 last week. Damir's 11 commits close 6 tickets (ECMDSW-387/397/394/383/382/370) — mix of boilerplate sync, product catalog import fixes, and login flow changes. This is real work, but single-digit ticket closures at 23% complete (342 stories outstanding) is unsustainable. At 2 stories/week, ECMD is 171 weeks from completion (3+ years). Even last week's 50 stories/week was only 7 weeks to completion, which was the first plausible timeline this board has ever shown.

Decomposition stalls: +2 new stories this week (vs +80 last week) suggest backlog grooming paused. Last week's +80 was deliberate epic decomposition (Damir/Lucian breaking down Infrastructure & Architecture, Catalog & Product epics into stories for the first time). This week's +2 is just minor adjustments to existing work. No new epics created, no new stories added to existing epics. This is the opposite of progress — the board is regressing back to the "epics with zero stories" state that plagued it for 4 weeks prior to 8/18.

Program_notes context: Sebastian and Gabriel were moved onto Baker/ECMD on 7/28 specifically to fix the board-vs-reality gap (5-10x scope explosion with commit volume far too low). Week 1 post-reallocation (8/4-8/11) showed no movement, as predicted. Week 2 (8/12-8/18) showed the breakthrough (+45 stories, 45 commits). Week 3 (8/19) reverts to near-zero movement. This does not match the expected pattern ("Week 2-3: expect board movement to start reflecting actual commit activity again"). Either the reallocation is not working, or there's a 1-week sprint cadence causing alternating weeks of decomposition vs implementation.

Risk: ECMD is structurally blocked — not by a single ticket, but by lack of sustained decomposition. The board can't move if epics don't get broken into stories. Last week proved this board can move when decomposition happens (50 stories closed in one week), but this week proves decomposition is not self-sustaining. If next week also shows <5 closed stories, escalate to Sebastian/Gabriel directly: what is blocking decomposition? Is it a sprint cadence issue, or is the underlying epic structure still too coarse to decompose?

9. Gemaire Shopware (GSHOP) ​

0% complete (0/79) — unchanged for fifth consecutive week. Commit activity persists (Damian: 9, Nikolay: 4 this week), but no board movement. Feature parity matrices unavailable for Gemaire, so cannot assess real progress outside Jira.

Week-over-week: +0 closed vs +4 added — net scope growth continues (+4 new stories), total now 79 (was 75 on 8/12). Scope is growing faster than work is landing, which is expected for a board in active decomposition, but five weeks of zero closures is extreme even for that phase.

Infrastructure blockers prevent test closure: GSHOP-165 (MiniCart Subtotal, Damian assigned), GSHOP-156 (Branch Finder Not Setting Preferred Branch, Damian assigned), and GSHOP-118 (Google MapID / Google Cloud Console Setup, Mike D'Ottavio assigned) each block 1-2 downstream test/QA tickets. Total: 9 test tickets blocked by 3 root-cause issues. None of the 3 blockers have moved in 12-14 days. Until these ship, no test tickets can close, and the board stays at 0%.

Damian Bogdanowicz ships 9 commits this week: Closes GSHOP-153 (Search Fails on mobile), GSHOP-149 (Shop for Products: Forbidden on mobile), GSHOP-148 (Search falls back to Magento on desktop), GSHOP-147 (Shop for Products dropdown links broken on desktop), and GSHOP-82 (Add to Cart). That's 5 closed tickets, but none moved to Done status in Jira — all closed as duplicates, obsolete, or merged into parent epics. This explains the 0/79 board state despite 9 commits: the work is landing, but Jira tracking is broken. Tickets close as "won't fix" or "duplicate" rather than "Done", so the done count never increments.

Nikolay Prudnikov ships 4 commits: No ticket closures visible this week, but ongoing work on GSHOP-159 (Arrow Graph - Quotes), GSHOP-160 (Recently Viewed Products), GSHOP-141 (Expose Wingman as Fusion-direct subgraph). These are all graph/subgraph work, not frontend, so no user-facing features ship yet.

Risk: Gemaire's board-vs-reality gap is the worst in the program. 5 weeks at 0/79, 9+4 commits/week, 5 tickets closed this week — but board shows zero progress because closed tickets aren't tracked as "Done". This is either a Jira workflow misconfiguration (tickets closing into wrong status) or a deliberate choice to not mark anything Done until full integration testing passes. Either way, it's invisible to velocity metrics. Recommend manual audit: what % of Gemaire's 79 stories are actually Done but not marked as such in Jira?

10. Baker Shopware (BSWM) ​

48% complete (173/362) — up +1 story from 172 on 8/18, +64 week-over-week. Third consecutive week of sustained single-digit gains (+11 last week, +1 this week), marking real progress after 4 weeks of stall prior to 8/12. 189 stories outstanding.

Week-over-week: +1 closed vs +1 added — net zero scope growth, and closure rate minimal. However, this is the third week in a row of forward movement (+64 total over last 7 days, +11 over prior 7 days), suggesting the Sebastian/Gabriel reallocation effect (flagged in program_notes as starting 7/28) is now fully visible and sustainable. Prior to 8/12, Baker was stuck at 35% (109/312) with zero movement for 3 weeks — the current 48% (173/362) represents 64 closed stories in 7 days, which is the fastest velocity this board has ever shown.

Repository activity: 25 commits this week, all from Jerrold Smith (25 frontend commits) — addresses, dashboard notifications, footer. This is a sharp drop from last week's 80 commits (Tyler: 45, Jerrold: 25, Sebastian: 10), but Jerrold's work is real BU-specific feature implementation (BSWM-153 Addresses, BSWM-154 Default Addresses, BSWM-214 Dashboard Notifications, BSWM-336 Footer), not just boilerplate sync. These are user-facing Baker-only features shipping for the first time.

Feature work visible in commit log: Sebastian Poblete's 6 commits on BSWM-439 (Order History Domain API changes) close a real Baker-specific feature (order history schema extensions). Cleo Hister's 15 commits on BSWM-407 (Warranty Search - Warranty Details) and BSWM-375 (Offers Plugin - Add correct UOM to price offer) ship real integrations. This is the first week where Baker's commit log shows recognizable user-facing features rather than just infrastructure/boilerplate work.

Shopping Lists progress: BSWM-414 (Shopping Lists - Details View) in active development per Cleo's commit log, though not yet closed. This is a core BU-specific feature (lists are Baker-only, not shared with other BUs) and represents the kind of work that justifies a dedicated BU board rather than folding everything into SXSW.

Risk: Baker's velocity is sustainable but capacity-constrained. +1 story this week (vs +11 last week) suggests a sprint cadence or resourcing bottleneck — Jerrold is the only active contributor this week, down from 3+ last week. 189 stories outstanding at 1-11 stories/week = 17-189 weeks to completion, which is wildly variable. If next week also shows single-digit movement, escalate to Tyler/Sebastian: is this a sprint boundary, or is capacity drying up?

12. Homans Shopware (HMSW) ​

0% complete (0/0) — board remains completely empty, exactly as expected per program_notes for newly-created board (created 8/18). All 31 epics remain in Discovery status with zero stories decomposed. However, feature parity matrix shows 36 Done features (up from 35 last week), and 255 commits from Lino/Pablo/Joshua this week drive real checkout/PDP/cart work not yet reflected in Jira.

Feature parity progress (real signal): 36 Done, 33 In Progress, 4 Testing, 59 Not Started — net +1 Done this week (was 35 last week). Active work: Checkout (WIP), PDP (WIP), Cart (WIP), Account (WIP), User Management (WIP), Branch (Done), Search (Done), Credit Card Processing (WIP), Company (WIP), Lists (Done), Documents (Done). This is the real progress indicator for Homans — Jira is 3-6 weeks behind reality, per program_notes.

Repository activity: 255 commits this week from Lino Rallo (59), Pablo Romero (157), Joshua LaRose (21), German Berselli (45), Rob Simmons (69), Sebastian Poblete (126), Tyler Hawkins (80), Robert Dumitru (46), Gabriel Guzmán (40), Damir Korpar (16) — all working on homans-next-frontend repo or shared boilerplate packages. Lino's 59 commits close SXSW-1309/1299/1281/1276 (CatalogFlow branch filters, ConduitLocations resolver, branch selector updates), which are Homans-specific implementations even though tickets live in SXSW. Pablo's 157 commits ship Quick Order, Quick View, PDP enhancements (SXSW-1267/1266/1238). Joshua's 21 commits close SXSW-1356/1272/1263/1249/909 (tax plugin, catalog refactor, Homans checkout).

Expected behavior per program_notes: "HMSW has epics but all are in Discovery status with no stories yet — this is expected for a board just starting decomposition, not a stall." This is exactly what's happening. The board is 1 day old. Expecting Jira decomposition to catch up to 6 weeks of frontend work (255 commits this week alone) in 1 day is unrealistic. The program_notes explicitly called this out: "If feature_parity data is present in the context, use it as the real progress signal for Homans... given the Jira gap above."

Risk: Homans has no Jira risk — the board is working as designed. The feature parity matrix (36 Done, +1 this week) is the real signal, and it's healthy. The risk is cross-board attribution: Lino/Pablo/Joshua are shipping Homans work but closing SXSW tickets because there are no HMSW tickets to close yet. This makes SXSW velocity look higher than it is (some of those +5 SXSW stories this week are actually Homans work) and Homans velocity invisible. Recommend: accelerate HMSW epic decomposition so contributors can start closing HMSW tickets instead of SXSW tickets.

13. Peirce Phelps Shopware (PPSW) ​

0% complete (0/0) — board remains completely empty, same as last week. Zero epics, zero issues. Jira project exists but work hasn't started. However, feature parity matrix shows 35 Done features (up from 34 last week), same as Homans, suggesting Peirce work is also hidden in SXSW ticket closures and boilerplate repos.

Feature parity progress (real signal): 35 Done, 34 In Progress, 3 Testing, 62 Not Started — net +1 Done this week (was 34 last week). Active work: Checkout (WIP), PDP (WIP), Cart (WIP), Account (WIP), User Management (WIP), Branch (Done), Search (Done), Credit Card Processing (WIP), Company (WIP), Lists (Done), Documents (WIP). This is identical to Homans' feature set (35 vs 36 Done, same WIP list) because both are built on the same boilerplate and share most implementations.

Repository activity: Same 255 commits as Homans (Lino/Pablo/Joshua/German/Rob/Sebastian/Tyler/Robert/Gabriel/Damir all work on peirce-frontend repo in addition to homans-next-frontend). Pablo's 157 commits this week include 31 on peirce-frontend (vs 32 on homans-next-frontend), suggesting equal effort across both BUs. Lino's 59 commits include 2 on peirce-frontend (branch selector, pricing/inventory). Joshua's 21 commits include 5 on peirce-frontend (tax plugin, catalog refactor).

Expected behavior per program_notes: "PPSW currently has zero epics and zero issues of any kind — the Jira project exists but work hasn't started; note this plainly rather than treating it as a data gap or error." This is even more extreme than Homans — PPSW has no Jira structure at all, not even Discovery epics. However, the feature parity matrix (35 Done) proves work is happening; it's just not tracked in Jira yet.

Risk: Same as Homans — no Jira risk, but cross-board attribution hides real velocity. Peirce contributors are closing SXSW tickets instead of PPSW tickets because there are no PPSW tickets to close. Recommend: create PPSW epics and decompose into stories so contributors can start attributing work correctly. Until then, Peirce velocity will remain invisible.

Review Pipeline Aging ​

145 tickets in review (vs 145 last week) — pipeline size stable, but internal distribution worsening. Deploying status now 30 tickets (up from 30 last week, median 27 days up from 26), and "Active Development" >30 days grows to 14 tickets (from 10 on 8/12).

Stale tickets >30 days: 14 total (vs 10 on 8/12), led by:

  • SXSW-1125 (Kubernetes/EKS Hosting for Domain Layer): 63 days in Active Development — no commits, no sub-tasks, no timeline. Chris Nitchie assigned. This is the longest-stalled ticket in the program.
  • CESW-915/914 (conduit-ng cache-bypass, customer_company_id filter): 41 days in Deploying — Blake Saunders, awaiting production deployment after successful staging validation. Not blocked, just slow deployment cadence.
  • CESW-1191/1181/1179/1048/1047/1046/874/872/871/870/817 (DMI/Lists work): 27-35 days in Deploying — all Blake Saunders, all awaiting production deployment. Same as above.

Review states by median age:

  • Deploying: 30 tickets, median 27 days, p90 35 days, max 41 days — CE dominates (14 tickets). This is not a blocker — these are validated and awaiting production cutover. The 27-day median suggests a ~1-month deployment cadence for infrastructure changes.
  • Active Development: 28 tickets, median 0 days, p90 15 days, max 63 days — SXSW-1125 is the only outlier at 63 days. Rest are <15 days, which is healthy.
  • QA: 18 tickets, median 14 days, p90 21 days, max 21 days — GSHOP dominates (10 tickets). These are blocked on infrastructure fixes (GSHOP-165/156/118), not aging due to QA capacity.
  • Code Review / Merge: 10 tickets, median 13 days, p90 14 days, max 14 days — all GSHOP, all blocked by same infrastructure fixes.
  • In Review: 15 tickets, median 2 days, p90 12 days, max 21 days — healthy, no systemic aging.

No new stale tickets added this week — the 14 >30 days were all present on 8/12, which is good (pipeline not backing up further). However, the existing batch is not clearing — SXSW-1125 has been 60+ days for 2+ weeks with no visible forward progress.

Systemic issue: Infrastructure/schema work (SXSW-1125, SXSW-1156, SXSW-1173) ages without resolution because it has no clear decomposition path. These tickets are "Active Development" but have no commits, no sub-tasks, and no progress updates. They're not blocked by downstream dependencies — they're stuck in design/architecture phase with no visibility into when they'll land. This is a program-wide risk multiplier: every BU board depends on stable ERP account schema, K8s hosting, and commercial plugin analysis, but none of those are on a timeline.

Repository Activity & Team Velocity ​

832 commits this week from 26 unique contributors (vs 740 commits from 25 contributors last week) — highest single-week volume since program tracking began. However, distribution is uneven: top 5 contributors (Pablo: 157, Sebastian: 126, Emil: 116, Tyler: 80, Rob: 69) account for 66% of total volume (548/832). Remaining 21 contributors average 13.5 commits each.

Team distribution:

  • Watsco Platform: 326 commits (Pablo: 157, Lino: 59, German: 45, Robert: 46, Mitchell: 2, Tiberio: 6, Htun: 3) — 39% of total program volume. This team drives all cross-BU work (boilerplate, SXSW platform, Homans/Peirce frontends).
  • Human Element: 176 commits (Rob: 69, Paul: 39, Joshua: 21, Ilir: 37, Cristian: 13, Adrian: 7) — 21% of total. This team drives domain API work (ERP accounts, pricing/inventory, reference-only products, fees).
  • Baker/ECMD: 215 commits (Sebastian: 126, Tyler: 80, Gabriel: 40, Jerrold: 25, Damir: 16, Cleo: 15, Lucian: 4, Nick: 3) — 26% of total. This team split across Baker and ECMD boards.
  • CE: 189 commits (Emil: 116, Scott: 45, Blake: 28) — 23% of total. This team drives CE Shopware board exclusively.
  • Gemaire: 14 commits (Damian: 9, Yovany: 1, Girish: 1, Nikolay: 4) — 2% of total. This team has the lowest output in the program despite a 79-story board.

Velocity by layer:

  • Frontend: 522 commits (63% of total) — Pablo, Sebastian, Tyler, Jerrold, German, Lino dominate. This is the highest frontend volume ever recorded in a single week.
  • Graph (Domain API): 246 commits (30% of total) — Rob, Paul, Joshua, Ilir, Cristian, Adrian, Gabriel, Blake. This is healthy depth — 7+ contributors shipping cross-BU domain work.
  • Shopware (plugins/backend): 64 commits (8% of total) — Emil, Scott, Blake, Nathan, Lucian, Nick, Cleo. This is the lowest Shopware volume in 4 weeks, suggesting a shift toward frontend/graph work this cycle.

Cross-board attribution issue persists: Lino/Pablo/Joshua close SXSW tickets (SXSW-1309/1299/1281/1276/1267/1266/1238) but their work is Homans/Peirce-specific (evidenced by repo targets: homans-next-frontend, peirce-frontend). This inflates SXSW velocity (+5 this week) and hides Homans/Peirce velocity (0/0 on both boards). Recommend: create HMSW/PPSW tickets for BU-specific work so contributors can close the correct board's tickets.

Single-person bottlenecks:

  • SXSW-1125 (Chris Nitchie, 63 days): No commits visible. Either working in a non-tracked repo (unlikely) or stalled in design phase with no code yet. Escalate for timeline.
  • CE Deploying backlog (Blake Saunders, 14 tickets >30 days): Not a bottleneck — work is done, awaiting deployment. However, Blake is the only CE contributor who can resolve this (he owns the DMI/conduit-ng stack), so if Blake becomes unavailable, the 14 tickets stay stuck.
  • Gemaire infrastructure blockers (Damian Bogdanowicz, 3 tickets blocking 9 downstream): GSHOP-165/156/118 have no commits this week despite 9 total from Damian. Either Damian is working on different tickets (5 closed this week, none were the blockers), or the blockers are harder than expected. Escalate for ETA.
mermaid
gantt
    title Projected Completion (based on current velocity)
    dateFormat YYYY-MM-DD
    todayMarker stroke-width:2px,stroke:#ef4444
    section Core
    SXSW (likely range)   :active, 2026-08-19, 2026-10-28
    section Business Units
    CE Shopware (stable)  :2026-08-19, 2026-09-30
    Baker (if sustained)  :2026-08-19, 2027-01-27
    ECMD (if last wk rpts):2026-08-19, 2026-09-30
    ECMD (if this wk rpts):crit, 2026-08-19, 2029-08-19
    Gemaire (blocked)     :crit, 2026-08-19, 2030-08-19

Notes on projections:

  • SXSW: 292 stories at 29/week = 10 weeks → Oct 28. This assumes velocity holds; any reversion to June levels (14/week) pushes past year-end.
  • CE: 58 stories at 27/week = 2 weeks → Sep 30. This is the most stable projection in the program.
  • Baker: 189 stories at 1-11/week range = 17-189 weeks. Used mid-range (6/week) → Jan 27. Highly variable.
  • ECMD: 342 stories at 2/week (this week) = 171 weeks → Aug 2029. At last week's 50/week = 7 weeks → Sep 30. The true timeline depends on whether last week's decomposition breakthrough repeats or this week's stall is the new normal.
  • Gemaire: 79 stories at 0/week for 5 weeks = infinite. Blocked by infrastructure fixes (GSHOP-165/156/118) with no ETA. Projected Aug 2030 as a placeholder for "undeliverable at current state."