The breakthrough — N3GZ packet capture

Through the amateur radio community, a debug log was obtained from operator N3GZ, running DXLab Commander 16.2.8 with a real FlexRadio FLEX-6500 (firmware 4.1.5.39794). The 1.8 MB log captured the complete handshake sequence between Commander and the FlexRadio, plus several minutes of normal operation including spot delivery.

This single artifact was the foundation of all subsequent protocol work. From it, the complete wire format was reconstructed — initial burst structure, command/response pairs, the spot-support gating logic, and the spot-relay protocol itself.

Without N3GZ’s capture, the reverse-engineering effort would have required access to a real FlexRadio transceiver. The amateur radio community’s willingness to share debug logs made this possible from a single capture.

Validation and demonstration

Once the Flex emulator handshake completed correctly and Commander began pushing spots, the integration was validated end-to-end:

• Cluster source — Reverse Beacon Network (RBN), with sustained spot rates of 500+ per minute
• Pipeline — cluster → SpotCollector → Commander (Flex Signature mode) → Bridge (Flex emulator) → TCI → Thetis panadapter
• Latency — approximately 30 ms end-to-end per spot (TCP receive → parse → TCI send), comfortable headroom over the 8 spots/sec arrival rate
• Colour preservation — foreground (need-category) and background (QSL status) colours delivered to Thetis exactly as configured in SpotCollector
• No drops, no spot ID collisions, no protocol violations during sustained operation

Evidence captured: timestamped full-screen screenshots showing all four applications — SpotCollector, Commander, Bridge debug, and Thetis panadapter — simultaneously, plus debug logs from Bridge, Commander, and Thetis’s TCI server. The same callsigns were traceable across all artifacts.

The complete demo package was delivered to Dave Bernstein AA6YQ as the milestone proving Stage 1 of the agreed two-stage development plan.

Two-stage agreement with the DXLab author

Dave AA6YQ proposed and agreed the work in two stages:

This two-stage approach validates the integration approach before committing to changes in Commander itself. Once Stage 2 ships, the Flex emulator code in Bridge can be retired in favour of a much simpler native-mode handler.

Future architecture

Once Dave ships native Thetis support in Commander (Stage 2), Bridge’s architecture converges to a focused minimum:

In this final form, Bridge reads each line from Commander’s TCP stream. If the line is a spot directive, it parses and translates to TCI. Otherwise, it forwards the raw bytes unchanged to Thetis’s CAT server.

This delivers several architectural benefits:

• Single TCP channel from Commander’s perspective — easier to firewall, easier to debug
• CAT path can’t break — Bridge isn’t interpreting CAT, only routing bytes
• No CAT translation work — Thetis’s CAT server already speaks Commander’s CAT dialect
• Bridge surface area shrinks dramatically — perhaps 200–300 lines of C# in steady state
• Flex emulator code retired — no longer needed once Commander emits the same wire format natively for Thetis
• Local DXView cache retired — the native path will carry both foreground and background colours, no local lookup needed

The value of this work beyond Bridge

The reverse-engineered Flex spot protocol documentation has utility beyond Bridge itself:

• Other SDR consoles could consume Commander’s spot stream by impersonating a Flex
• Other loggers could implement compatibility with the protocol Dave will ship for native Thetis support
• Future developers can understand the wire format without re-deriving it from packet captures
• The DXLab author has a reference for how the protocol is observed and consumed externally

Bridge is built with an explicit understanding that its job is to disappear into the platform. As DXLab and Thetis converge on common protocols, Bridge becomes thinner. The reverse-engineering work captured here ensures that knowledge survives that convergence.

Acknowledgements