UI Refresh Architecture Decision Matrix

Updated: 2026-06-26 · Audited: 2026-06-26 · Work Item: AB#1821 · Decision deadline: March 15, 2026

Executive Summary: SDUI is confirmed as the approach. Decision pending: which SDUI implementation?
Recommendation: Path 2 (Hybrid SDUI) for Fall 2026, with Path A (Full SDUI) upgrade in 2027.

Path Summary

Option	Timeline	Fall 2026?	Best For
Path 2: Hybrid SDUI ⭐	7.5–9.5 months	✅ Tight but feasible	Hit deadline, upgrade later
Path A: Full JSON SDUI	13–15 months	❌ Misses by 6+ months	Maximum flexibility, 2027 launch
Path B: Blazor Hybrid	5–6 months	✅ Yes	HTML from CMS, WebView acceptable
Path C: FlagstoneUI + SDUI	14–16 months	❌ Misses by 8+ months	Pixel-perfect — HIGH RISK
Path D: Mixed (Native + Blazor)	7–8.5 months	⚠️ Tight	Pragmatic balance, mixed UX OK

Critical: Must decide by March 15, 2026 to start foundation work in Week 3.

Confirmed Requirements

Requirement	Answer	SDUI Impact
Update speed	Same-day (within hours)	✅ SDUI mandatory
Change types	All (content, styling, layout, nav, flags, A/B)	✅ Full SDUI required
MEC layouts	Different per MEC	✅ Layout renderer required
Launch plan	DAL + UAL pilot, Sept 2026	✅ 6-month timeline
Offline	Existing strategy works	✅ No new architecture needed
HTML source	Trusted ALPA only	✅ No sandboxing needed

Decision Matrix: Questions → Architecture Path

Q6: Look & Feel

Answer	Eliminates	Favors	Reasoning
Q6.1: Pixel-perfect consistency critical	Blazor Hybrid, Native XAML	FlagstoneUI + JSON SDUI	FlagstoneUI solves cross-platform quirks
Q6.1: Platform-native OK	FlagstoneUI	JSON SDUI, Mixed	Leverage native iOS glass + Android Material
Q6.2: Full iOS glass aesthetic	Blazor, FlagstoneUI	JSON SDUI (native)	WebView can't do glass, FlagstoneUI is flat
Q6.5: Web-like acceptable	—	Blazor Hybrid, Mixed	Makes WebView viable
Q6.5: Must feel native	Blazor (full)	JSON SDUI, FlagstoneUI+SDUI	Rules out pure WebView
Q6.6: Mixed architecture OK	—	Native + Blazor	Performance screens stay native

Q7: Performance

Answer	Eliminates	Favors	Reasoning
Performance screens <200ms	Blazor (those screens)	Native/JSON SDUI	WebView overhead too high
Content screens <2s OK	—	Blazor viable	Relaxed constraint
All performance issues unacceptable	Blazor Hybrid	JSON SDUI, FlagstoneUI	No WebView risk

Q8: Content Richness

Answer	Eliminates	Favors
Simple text/buttons	—	Any SDUI
Rich HTML from CMS	JSON SDUI (complex)	Blazor Hybrid, Mixed
Embedded media	—	Blazor (easier)

Q9: Risk Tolerance

Answer	Eliminates	Favors
Stable tech = critical	FlagstoneUI	Blazor, JSON SDUI (native)
OK with WebView	—	Blazor Hybrid, Mixed
OK with custom renderer	—	JSON SDUI

Path A: Full JSON SDUI

Effort: 52–60 weeks · Launch: April–June 2027 · ❌ Misses Fall 2026

Phase Breakdown: Phase 1 (6w): Remote ResourceDictionary loader, feature flag system, version management. Phase 2 (10w): JSON schema design, component registry, JSON-to-XAML renderer, data binding, schema validation. Phase 3 (8w): Analytics, rollback mechanism, A/B testing, staged rollout, security validation. Phase 4 (28–36w): Dashboard · Document Library · Committee Pages · Media Hub · Navigation · Personalization · Hotels/Reviews · Offline FTDT.

Pros: Full layout control · Platform-specific handlers (iOS glass, Material Design) · Native performance · No experimental dependencies · Maximum flexibility · Instant rollback during incidents · Full A/B testing (layout + content) · Each MEC can have completely unique layouts.

Cons: Cannot hit Fall 2026 target (misses by 6+ months) · Highest complexity and development cost · Complex QA (must test all possible JSON combinations) · Performance overhead (runtime rendering) · Requires robust backend CMS.

Path B: Blazor Hybrid

Effort: 20–24 weeks · Launch: July–August 2026 · ✅ Can hit Fall 2026

Phase Breakdown: Phase 1 (3–4w): Blazor WebView integration, bridge to native services, offline caching, security hardening. Phase 2 (17–20w): Dashboard · Document Library · Committee Pages · Media Hub · Navigation · Personalization · Hotels/Reviews · Offline FTDT stays native.

Pros: Proven technology · Web development skills · HTML rendering built-in · Dynamic CSS for MEC styling · Fastest to market.

Cons: WebView overhead · Cannot do iOS glass effects · Prior bug (#1891) demonstrates WebView risks · Less "native" feel · Performance concerns for complex interactions.

Path C: FlagstoneUI + JSON SDUI

Effort: 56–64 weeks · Launch: May–July 2027 · ❌ HIGH RISK — Do not choose for Fall 2026

FlagstoneUI is v0.0.1 (only 4 controls, experimental). Complex hybrid approach with longest timeline. Best for: Pixel-perfect consistency + dynamic updates, if you can tolerate both FlagstoneUI risk AND 2027 timeline.

Path D: Mixed Architecture (Native + Blazor)

Effort: 28–34 weeks · Launch: September–October 2026 · ⚠️ Tight

Phase Breakdown: Phase 1 (3–4w): Blazor WebView integration for content screens. Phase 2 (8–10w): Flight Finder + FAR 117 Calculator (keep existing, refresh UI). Phase 3 (17–20w): Dashboard (Blazor) · Document Library · Committee Pages · Media Hub · Navigation · Personalization · Hotels/Reviews.

Pros: Best of both worlds · Flight Finder/FAR 117 stay native · MEC content gets dynamic updates · Can hit Fall 2026 · Lower risk than Full SDUI.

Cons: Inconsistent feel (users may notice native vs web) · Two codebases to maintain · Team needs both skill sets · Navigation complexity between paradigms.

Path 2: Hybrid SDUI — Recommended

    RECOMMENDED for Fall 2026 with 2027 upgrade path to Path A.

    Provides remote styling/branding without app updates. Feature flags for gradual rollout and A/B testing. Fixed component layouts (XAML), dynamic styling. Same layout structure for all MECs, customized appearance.

Effort: 31–38 weeks · Launch: October–November 2026

Phase Breakdown: Phase 1 (6w): Remote ResourceDictionary loader, feature flag system, version management & caching, fallback/offline handling. Phase 2 (25–32w): Dashboard (3–4w) · Document Library Search (4–5w) · Committee Pages (2–3w) · Media Hub (4–5w) · Navigation Overhaul (2–3w) · Personalization UI (2w) · Hotels/Reviews (6–8w) · Offline FTDT UI Refresh (2w).

Pros: MEC branding/styling changes without App Store approval (hours to deploy) · Feature flags enable safe gradual rollout per MEC · Moderate complexity, manageable risk · Can upgrade to Path A later without major rewrite · Can hit Fall 2026 target.

Cons: Cannot change layout structure remotely (add/remove/reorder components) · All MECs must use same component arrangement · Limited A/B testing (style variations only) · Major feature additions still require app store updates.

Phased Approach Strategy

2026 — Phase 1: Launch Hybrid SDUI (Path 2) — Delivers 80% of SDUI value with 60% of effort. Hits Fall 2026 deadline. Proves SDUI concept with lower risk. MEC branding customization without app updates. Feature flags for safe rollout.

2027 — Phase 2: Evaluate Full SDUI Upgrade (Path A) — Based on 2026 feedback: do MECs actually need different layouts, or is styling enough? What's the real-world update frequency? If needed, upgrade adds JSON renderer on top of existing UI (24 additional weeks). Phase 1 foundation is reusable.

Decision Tree

Start: SDUI Required (confirmed)
│
├─ Q6.1: Visual consistency critical?
│  ├─ YES → Path C (FlagstoneUI+SDUI) or Path A with careful handlers
│  └─ NO → Continue
│
├─ Q6.2/Q6.3: iOS glass or Material Design 3?
│  ├─ YES → Path A (JSON SDUI with native handlers)
│  └─ NO → Continue
│
├─ Q6.5: Web-like acceptable for content?
│  ├─ YES → Path B (Blazor) or Path D (Mixed)
│  └─ NO → Path A (JSON SDUI)
│
├─ Q7: Performance critical everywhere?
│  ├─ YES → Path A (JSON SDUI) or Path C (if consistency also critical)
│  └─ NO (relaxed for content) → Path B or Path D viable
│
├─ Q8.1: HTML from CMS?
│  ├─ YES (rich content) → Path B (Blazor) strongly favored
│  └─ NO (simple content) → Path A or Path C
│
└─ Q9: Stable tech priority?
   ├─ YES → Path B (Blazor, proven) or Path A (Native SDUI)
   └─ NO (can tolerate experimental) → Path C (FlagstoneUI+SDUI) viable

Predicted Outcome

Most likely: Path 2 (Hybrid SDUI) with 2027 upgrade path to Path A.

Path 2 wins because: SDUI foundation (as required) · Can hit Fall 2026 (tight but doable) · Native rendering (no WebView risks like #1891) · Consistent UX (unlike Path D mixed approach) · Upgrade path to Full SDUI in 2027 · De-risks investment (validate concept before full JSON renderer).

If…	Choose…
Fall 2026 is firm	Path 2 (Hybrid SDUI), plan Path A upgrade for 2027
Timeline can slip to 2027	Path A (Full JSON SDUI) from start
WebView acceptable + HTML from CMS	Path B (Blazor) fastest option
Strict performance everywhere	Rules out Path B; leaves Path 2 as only viable option

Critical Path (Path 2)

Today: March 2, 2026 · Target: Fall 2026 (September–November) · Available time: ~6–9 months

March 2026 (Week 1–2): Architecture decision finalized
March 2026 (Week 3–4): SDUI foundation begins (ResourceDictionary loader, feature flags)
April 2026 (Month 1): Foundation complete, feature UI begins
May–August 2026 (Months 2–5): Feature UI development
September 2026 (Month 6): Internal testing, UAL/DAL pilot prep
October 2026 (Month 7): UAL/DAL pilots launch
November 2026 (Month 8): Refinements based on pilot feedback

Risk buffers: 2–3 week buffer for unknowns · Prioritize core features first (Dashboard, Committee Pages, Media Hub) · Hotels/Reviews can slip to Phase 2 if needed (6–8 week savings).

Must start by mid-March 2026 to hit Fall target. Architecture decision deadline: March 15, 2026.

Next Actions

Finalize architecture decision by mid-March 2026 ⏰ URGENT — Recommend Path 2 (Hybrid SDUI) with 2027 upgrade option
If choosing Path 2: Begin SDUI foundation work immediately (Week 3–4 March)
Schedule Communications questionnaire workshop (1–2 hours) — still needed for Q6–Q9
Update sprint plan with chosen approach and detailed timeline
Staff planning: Ensure team capacity for 7.5–9.5 month effort

Reference Documentation

UI Refresh Communications Questions — Requirements questionnaire (Q6–Q9 outstanding)
Stakeholder Email Draft — Progressive disclosure email for Communications team
Epic AB#1821 — UI Refresh — March 2026 sprint plan preserved in the Epic discussion (sprint plans live on the board)
Component Library Build Plan — Three-track component implementation status

Prepared by: ALPA Bot · Date: 2026-03-02 18:20 UTC · Related: AB#1821