diff --git a/architecture.md b/architecture.md
index f3a542d..cbef961 100644
--- a/architecture.md
+++ b/architecture.md
@@ -802,14 +802,14 @@ These were previously open questions and are now resolved:
 - **Stream ID assignment**: stream IDs are assigned by the controller, not generated locally by nodes. This ensures both ends of a stream report the same ID and the graph can be reconstructed by correlating node state reports.
 - **Single port per node**: one TCP listening port handles all traffic — video frames, control messages, state queries — in both directions. Dedicated per-stream ports on separate sockets are a future option but not the default.
 - **First delivery mode**: low-latency (no-buffer) mode is implemented first. No frame queue anywhere in the pipeline — V4L2 dequeue goes directly to transport send; received frames render immediately and are dropped if the display is behind.
+- **Drop policy**: per-output configurable. Both drop-oldest (recency) and drop-newest (continuity) are supported; the policy is set at stream open time.
+- **Stream ID remapping at relay**: no remapping — stream IDs pass through unchanged. The relay forwards frames with the same stream ID they arrived with. Site-to-site gateways may need to translate IDs at the boundary but that is a future concern handled at the gateway, not in the relay itself.
+- **Transport for relay edges**: TCP only for now. UDP and shared memory (for local hops) are future considerations; the transport abstraction should accommodate them without the relay needing to care which is in use.
+- **Byte budgets**: soft limits with hysteresis — two thresholds (start dropping, stop dropping) to avoid thrashing at the boundary.
+- **Relay scheduler**: strict priority first. Additional policies (round-robin, weighted round-robin, deficit round-robin, source suppression) are documented in the Relay Design section and will be added later. The scheduler interface is pluggable so policies are interchangeable without touching routing logic.
 
 ---
 
 ## Open Questions
 
 - What is the graph's representation format — in-memory object graph, serialized config, or both?
-- Drop policy for completeness queues: drop oldest (recency) or drop newest (continuity)? Should be per-output configurable.
-- When a relay has multiple inputs on an encapsulated transport, how are streams tagged on the outbound side — same stream_id passthrough, or remapped?
-- What transport is used for relay edges — TCP, UDP, shared memory for local hops?
-- Should per-output byte budgets be hard limits or soft limits with hysteresis?
-- Which relay scheduler policies should be implemented first, and what is the right interface for plugging in a custom policy? Minimum viable set is probably strict priority + weighted round-robin.