Keywords: milestone dag, dependency graph, critical path, feature clusters, roadmap overlay, implementation order, hard dependency, soft dependency
This page is the detailed dependency companion to ../18-PROJECT-TRACKER.md . It does not replace ../08-ROADMAP.md ; it translates roadmap phases and accepted decisions into an implementation-oriented milestone DAG and feature-cluster dependency map.
Use this page to answer:
What blocks what?
What can run in parallel?
Which milestone exits require which feature clusters?
Which policy/legal gates block validation even if code exists?
Where do Dxxx decisions land in implementation order?
Edge Kind Meaning Example
hard_depends_onCannot start meaningfully before predecessor exists M2 depends on M1 (sim needs parsed rules + assets/render slice confidence)
soft_depends_onStrongly preferred order; can parallelize with stubs M8 creator foundations benefit from M3, but can start after M2
validation_depends_onCan prototype earlier, but cannot validate/exit without predecessor M7 anti-cheat moderation UX can prototype before full signed replay chain, but validation depends on D052/D007 evidence
enables_parallel_workUnlocks a new independent lane M2 enables M8 creator foundation lane
policy_gateLegal/governance/security prerequisite DMCA agent registration before validating full Workshop upload ops
integration_gateFeature exists but must integrate with another system before milestone exit D069 setup wizard + D068 selective install + D049 package verification before “ready” maintenance flow is considered complete
M0 -> M1 -> M2 -> M3
├-> M4 (minimal online slice)
├-> M8 (creator foundation lane)
└-> M5 -> M6
M4 + M6 -> M7
M7 + M8 -> M9
M9 -> M10
M7 + M10 -> M11
Milestone Objective Maps to Roadmap hard_depends_onsoft_depends_onUnlocks / Enables
M0Tracker + execution overlay baseline Pre-phase docs/process — — M1 planning clarity
M1Resource/format fidelity + rendering slice Phase 0 + Phase 1 M0— M2
M2Deterministic sim + replayable combat slice Phase 2 M1— M3, M4, M5, M8
M3Local playable skirmish Phase 3 + Phase 4 prep M2— M4, M5, M6
M4Minimal online skirmish (no tracker/ranked) Phase 5 subset M3M5 (parallel)M7
M5Campaign runtime vertical slice Phase 4 subset M3M4 (parallel)M6
M6Full campaigns + SP maturity Phase 4 full M5M4M7
M7Multiplayer productization Phase 5 full M4, M6— M9, M11
M8Creator foundation (CLI + minimal Workshop) Phase 4–5 overlay + 6a foundation M2M3, M4M9
M9Scenario editor core + full Workshop + OpenRA export core Phase 6a M7, M8— M10
M10Campaign editor + modes + RA1 export + ext. Phase 6b M9— M11
M11Ecosystem polish + optional AI/LLM + platform breadth Phase 7 M7, M10— Ongoing product evolution
Order Milestone Why It Is On the Critical Path
1 M1Without format/resource fidelity and rendering confidence, sim correctness and game-feel validation are blind
2 M2Deterministic simulation is the core dependency for skirmish, campaign, and multiplayer
3 M3First playable local loop is the gateway to meaningful online and campaign runtime validation
4 M4Minimal online slice proves the netcode architecture in real conditions before productization
5 M5Campaign runtime slice de-risks the continuous flow/campaign graph stack
6 M6Full campaign completeness is a differentiator and prerequisite for final multiplayer-vs-campaign prioritization decisions
7 M7Ranked/trust/browser/spectator/community infra depend on both mature runtime and online vertical slice learnings
8 M8Creator foundation can parallelize, but M9 cannot exit without it
9 M9Scenario editor + full Workshop + export core unlock the authoring platform promise
10 M10Campaign editor and advanced templates mature the content platform
11 M11Optional AI/LLM and platform polish should build on stabilized gameplay/multiplayer/editor foundations
Lane Start After Primary Scope Why Parallelizable
Lane A: Runtime CoreM1M2 -> M3 -> M4Core engine and minimal netcode slice
Lane B: Campaign RuntimeM3M5 -> M6Reuses sim/game chrome while net productization proceeds
Lane C: Creator FoundationM2M8CLI/minimal Workshop/profile foundations can advance without full visual editor
Lane C₂: P2P EngineM2M8 → M9p2p-distribute standalone crate (D076 Tier 3); core engine through NAT/uTP needed for M8 Workshop delivery; full CAS/federation for M9
Lane D: Multiplayer ProductizationM4 + M6M7Needs runtime and net slice maturity plus content/gameplay maturity
Lane E: Authoring PlatformM7 + M8M9 -> M10Depends on productized runtime/networking and creator infra
Lane F: Optional AI/PolishM7 + M10M11Optional systems should not steal bandwidth from core delivery
Section Topic File
Execution Ladders Granular foundational execution ladder A-G (RA first mission loop through completion) execution-ladders.md
Clusters M0-M1 Feature cluster dependency matrix: M0 (Tracker/Overlay) + M1 (Resource/Render Fidelity) clusters-m0-m1.md
Clusters M2-M4 Feature cluster dependency matrix: M2 (Deterministic Sim) + M3 (Local Skirmish) + M4 (Minimal Online) clusters-m2-m4.md
Clusters M5-M6 Feature cluster dependency matrix: M5 (Campaign Runtime) + M6 (Full Campaigns) clusters-m5-m6.md
Clusters M7 + Addenda Feature cluster dependency matrix: M7 (MP Productization) + cross-milestone addenda clusters-m7-addenda.md
Clusters M8 Feature cluster dependency matrix: M8 (Creator Foundation) clusters-m8.md
Clusters M9 Feature cluster dependency matrix: M9 (Scenario Editor + Full Workshop) clusters-m9.md
Clusters M10-M11 Feature cluster dependency matrix: M10 (Campaign Editor + Modes) + M11 (Ecosystem Polish) clusters-m10-m11.md
Gates and Mappings UX surface gate clusters, policy/external gates, external source study mappings, mapping rules, feature intake gates-and-mappings.md