CAPTURE FLOW

The same five-step progression from the newcomer onramp, drawn as a flow. Each lane represents one onramp step.

If you’re reading this in Obsidian, the .canvas version in the repo root has the same layout as an interactive board. The Mermaid version below renders directly in GitHub Pages.

flowchart LR
    %% Step 1 — Phone only
    s1_phone[Android phone<br/>Wi-Fi + BT + GPS] --> s1_app[WiGLE Wifi Wardriving<br/>Android app]
    s1_app -->|auto-upload| wigle((WiGLE.net))

    %% Step 2 — Cross-post
    s2_phone[Android phone<br/>same as Step 1] --> s2_csv[.wiglecsv.gz export<br/>WigleWifi-1.6]
    s2_csv --> s2_w2w[wigle-to-wdgwars<br/>v1.6.2 --setup wizard]
    s2_w2w -->|POST /api/upload-csv| wdgwars_csv((wdgwars.pl<br/>/api/upload-csv))
    s2_csv -.->|or direct to WiGLE| wigle

    %% Step 3a — On-device uploaders
    s3a_cardputer[M5 Cardputer + GPS] --> s3a_bruce[LOCOSP Bruce fork<br/>v1.0-wdgwars]
    s3a_bruce -->|direct| wdgwars_csv
    s3a_xiao[XIAO ESP32-C5/S3/C6/C3<br/>or T-Dongle C5<br/>C3 is 2.4 GHz only] --> s3a_piglet[Piglet firmware<br/>hamspiced/piglet, 213★]
    s3a_piglet -->|web UI direct| wdgwars_csv
    s3a_pager[Hak5 Pineapple Pager<br/>+ u-blox 7 GPS] --> s3a_payload[LOCOSP Pineapple payload<br/>SYNC NOW button]
    s3a_payload -->|direct| wdgwars_csv
    s3a_rj[Raspberry Pi + LCD 1.44 + GPIO] --> s3a_rjpayload[wdgwars_upload.py<br/>Raspyjack payload]
    s3a_rjpayload -->|direct| wdgwars_csv

    %% Step 3b — Capture-only, PC feeder
    s3b_hw[Classic ESP32 / CYD<br/>+ GPS module] --> s3b_marauder[Marauder v1.13.0<br/>writes WigleWifi-1.4 to SD]
    s3b_marauder -->|SD pull| s3b_sd[wigle-to-wdgwars<br/>pads 1.4 to 1.6]
    s3b_sd --> wdgwars_csv

    %% Step 4 — WDGWars-only slots
    s4_sdr[RTL-SDR<br/>+ 1090 MHz antenna] --> s4_dump[dump1090 / readsb]
    s4_dump --> s4_muninn[Muninn<br/>adsb-to-wdgwars v2.2.1]
    s4_lora[LoRa node<br/>Heltec / TTGO] --> s4_mm[MeshMapper CSV]
    s4_mm --> s4_heimdall[Heimdall<br/>meshcore-to-wdgwars v0.8.1]
    s4_muninn --> gungnir[gungnir v0.1.3<br/>HMAC + retry + cooldown]
    s4_heimdall --> gungnir
    gungnir -->|signed JSON<br/>POST /endpoint/upload/| wdgwars_json((wdgwars.pl<br/>/api/upload/))

    %% Styling — CRT palette matching the on-ramp brand
    classDef destinations fill:#3a2a05,stroke:#fbbf24,stroke-width:2px,color:#fbbf24
    classDef ondevice fill:#063312,stroke:#00e436,color:#00e436
    classDef pcside fill:#053946,stroke:#00e5ff,color:#00e5ff
    classDef transport fill:#3a0010,stroke:#ff004d,stroke-width:2px,color:#ff004d
    class wigle,wdgwars_csv,wdgwars_json destinations
    class s3a_bruce,s3a_piglet,s3a_payload,s3a_rjpayload ondevice
    class s2_w2w,s3b_sd,s4_muninn,s4_heimdall pcside
    class gungnir transport

Reading the diagram

Notes

• One source can feed two destinations. A WiGLE Android export is accepted by both WiGLE.net (direct) and wdgwars.pl (via wigle-to-wdgwars).
• Two upload paths on wdgwars.pl side. Bulk Wi-Fi/BLE CSVs go to /api/upload-csv (multipart). ADS-B + LoRa mesh go to /api/upload/ as signed JSON via gungnir.
• The mesh slot takes MeshCore and Meshtastic, not just MeshCore. The lane above shows the MeshCore path (LoRa node → MeshMapper CSV → Heimdall) because that’s the feeder that exists today. Since 2026-08-12 the slot itself also accepts Meshtastic records, told apart by a network field. See the Hardware survey §1a. No Meshtastic-native feeder has been built yet.
• The /endpoint/* mirror exists as a Cloudflare-L7 bypass for clients that burst-POST. gungnir v0.1.2+ uses it by default. Hand-rolled clients should target /endpoint/* rather than /api/* if they POST in bursts.

For the full progression with hardware suggestions and skill prerequisites at each level, see the Newcomer onramp. For the firmware × chip support matrix behind each tier, see the Hardware survey.