MeshCore-Solo/docs/development/l2-roadmap.md

Wio Tracker L2 (ui-lvgl) roadmap

The L1 → L2 feature port and its audit are done (2026-09-26). This is the plan for what comes next, in order. Tick items off as they land.

Decisions already made:

• Buttons: WAKE (side, expander P00) turns the screen off / on. USER/BOOT goes back; holding it mutes / unmutes the device.
• Placeholders: the "+" popup next to a text field is the placeholder UI; it must offer every placeholder and appear at every field that sends text.
• Message history: kept on the SD card, about 100 per conversation to start with.
• OTA: firmware comes from this fork's GitHub Releases.
• PIN: screen lock only, for now.

1. Quick fixes

• Keyboard: a key that closes it (so far a message could only be sent).
• "+" placeholders at every field that sends text (compose, quick messages, bot replies).
• Short, terse descriptions on buttons and rows.
• WiFi off switch.
• Buttons as decided above.

2. Arduino-ESP32 3.x, PSRAM

• A separate L2 env on pioarduino 55.03.312-1 (Arduino-ESP32 3.3.12, IDF 5.5): Wio_Tracker_L2_companion_solo_lvgl_v3; the 2.0.17 env stays until the new one passes the hardware checklist.
• Port: speaker (new I2S driver), BLE (NimBLE), ESP-NOW (IDF 5 callbacks); HTTPS map download and LovyanGFX build unchanged (check on the device).
• Large buffers in PSRAM (psramBuf): trail drawing, tile cache, map marks, list rows, keyboard maps, message metadata. Internal heap free at runtime: 2.0.17 147.8 KB; 3.x 128.2 KB before, 147.9 KB after (static RAM 109 → 71 KB). Measured with -D UI_HEAP_REPORT. History and a larger trail come to PSRAM with stage 3.
• The 3.x env is the default (2026-09-26, after stages 5 and 6 were tested on it): Wio_Tracker_L2_companion_solo_lvgl is Arduino 3.3.12 and is the one published; 2.0.17 stays as Wio_Tracker_L2_companion_solo_lvgl_arduino2 (not published, can't self-update).
• Limits sized for the PSRAM, overridable per board so the nRF52 defaults stay (the L1 build is byte-for-byte unchanged): trail 512 → 4096 points (TRAIL_CAPACITY; the map drops points under 2 px apart when drawing, so a long trail doesn't slow redraws; 8 → 32 trail segments), waypoints 16 → 64, message history 48 → 256 channel and 32 → 128 DM entries (HIST_CH_MAX / HIST_DM_MAX; stage 3 moves it to SD), Nearby 32 → 64; in ui-lvgl: contact list 64 → 256 rows, rooms 16 → 32, conversation 30 → 50 bubbles, pickers 64 → 128, WiFi scan 12 → 20. The sim builds with the same limits. Measured after: internal heap 154.1 KB free (was 147.9), PSRAM 5.67 MB free.

3. Data on the SD card

• Message history on SD (HistoryStore.h, -D HIST_ARCHIVE): one file per conversation in /sdcard/meshcore/history (channels keyed by a hash of name + secret, contacts / rooms by key prefix), a ring of the history entries themselves, every new message and later change (relay echo, delivery) written at once. After a reboot the newest entries go back into the RAM ring (Messages list, previews); a conversation shows 50 at a time with "Older messages (n)" / "Newer messages" and reads the card, so it reaches back as far as the card keeps. Kept per chat: 100 / 250 / 500 (default) / 1000 / 2000, ~20 KB per 100 messages.
• Settings > Storage: SD card used / free with what takes it (maps, messages, GPX trails, other; files counted in the background), the internal flash, "Kept per chat", "Delete message history" (tap twice).
• Saved trails on the SD card: Map tools > Save writes a new /sdcard/trails/trail-YYYYMMDD-HHMM.trl each time (the internal slot without a card); Load lists them newest first (plus the internal slot, where the low-battery auto-save goes) -- each with distance, time and points, Load (the map frames it), GPX, Delete.
• Diagnostics > Live: GPS (off / no data / fix or no fix with satellites), why the device last started (crash, watchdog, low voltage...), and the last crash from the core dump partition (task and address).

4. UI layout

L1 splits its tools into many small screens because of the joystick and the 128x64 display; the L2 groups them where they are used instead.

• Tools: trail, live share and arrival alert stay in the map (its tools popup, with their options behind "Options"); the advert (send now, auto-advert) is in Nearby; Repeater, Admin (a list of repeaters and room servers) and Diagnostics are Home tiles next to Favourites, Compass, Clock and Bot. The melody editor stays in Sound.
• Settings in L1's order: display, sound, radio (with Bluetooth, WiFi), system, keyboard, contacts & messages.
• Every status icon L1 has (Bluetooth, GPS, alarm, mute, auto-advert, trail, live share, repeater) plus arrival alert; background modes in the accent colour instead of L1's blinking.

5. Security and internet

• Screen-lock PIN: 4-8 digits in NVS (lvport::loadPin), asked on every wake and after a reboot, 5 misses pause entry for 30 s; Settings > Display & power > SECURITY.
• Map tiles fetched live while online: the map queues tiles it is missing, TileDownloader fetches them one at a time over WiFi (connected on the first miss, dropped 30 s after leaving the map) and saves them to the card; Map > ☰ > Live tiles (NVS, on by default).
• One-button OTA from GitHub Releases: Settings > System > Firmware update (OtaScreen.h). The latest release's -Wio-Tracker-L2-ota.bin (app image; build-solo-firmwares.yml now publishes *_solo_lvgl) is streamed over TLS verified with the framework's CA bundle into the idle slot of default_16MB.csv (two 6.25 MB app slots, no layout change), chip id checked, then restart. Arduino 3.x builds only.
• The 3.x env is the *_solo_lvgl one (stage 2).
• With the first release that has an L2 asset: test an update end to end.

6. Theme

• A consistent style, written down in Theme.h: accent fill = the primary action (Download, Install, Save, Go), dim accent fill = selected / on (tabs, chips, segments -- also the default theme's CHECKED), accent text = names, counts, modes; one card radius for buttons and rows, pills for chips, RADIUS_SM inside; slightly lighter surfaces. Accent colour selectable (Settings > Display & power > LOOK: amber, orange, coral, violet, cyan, lime; NVS mc_ui).
• Light motion (Anim.h): a new screen emerges from the middle (a background-coloured cover fades while the content drifts 6 px), Home pages slide after a swipe, popups and toasts rise and fade in, buttons shrink a little while pressed.
• Frame time measured on the device (-D UI_PERF_TEST: walks screens by itself, prints render + flush per refresh): ~48 → ~33 ms per transition frame with uncompressed fonts and two 120-line buffers. Internal-RAM / DMA buffers, -O2, hot code in IRAM, two draw threads, system malloc: no gain. The rest is LVGL's software rendering at 320x240.
• Status bar icons in equal cells; muted is a speaker with a cross.
• Under an own message: the age and a small mark as in L1 (✓ n repeaters for channel posts; ✓ / ✗ / ... for DMs) instead of the words.
• The path window of a message: quote, time and hops, a path diagram (sender → repeaters → this device), repeaters that relayed an own post, an own DM's delivery in words.
• Tile download popup fits the screen; its WiFi button went (WiFi is set up in one place, Settings).
• Splash screen (Splash.h): MeshCore wordmark, SOLO, the Solo and the upstream version, build date, loading dots.

7. Research: LVGL for the other displays

• Find out whether LVGL on every display variant gives consistency and a nicer look more easily. Adopt only if the result is better and every feature is kept (L1: 128×64 OLED, nRF52, RAM already 68% used).

Findings (2026-09-26). Solo builds, static RAM / flash used:

Board	MCU	Display	RAM	Flash
Wio Tracker L1	nRF52840	128×64 OLED	68% (74 KB free)	70% (214 KB free)
Wio Tracker L1 e-ink	nRF52840	e-ink	70%	71%
T-Echo Lite	nRF52840	e-ink	69%	64%
GAT562 Mesh Watch13	nRF52840	128×64 OLED	68%	92% (55 KB free)
ProMicro, GAT562 30S	nRF52840	128×64 OLED	(as L1)
Heltec V3 / V4	ESP32-S3	128×64 OLED	56%	44% of 3.2 MB
Cardputer ADV	ESP32-S3, no PSRAM	240×135 colour TFT, keyboard	56%	43% of 3.2 MB

What LVGL costs on the L2: the library ~310 KB of code (full config, PNG decoder and all widgets; a minimal one is ~120-150 KB), the ui-lvgl screens ~225 KB, fonts 360 KB (uncompressed, European + Cyrillic, 12-40 px). ui-new on the L1 is ~137 KB.

• 128×64 OLED (L1, Heltec, ProMicro, GAT562): to be tried. Colour and anti-aliasing don't matter there, but the v2 UI goals do: motion (Home carousel slide, loading-dot wave, animated splash, a bottom drawer, radar sweep), soft corners, graphic indicators instead of text, a bigger font to try out, dithering for large elements. LVGL has the animation engine, shapes, TTF fonts at any size and self-laying-out lists for that; rendered in greyscale and converted to 1 bit in the flush, a Bayer threshold there would turn every fade / translucent fill into dithering while 1-bpp text and icons stay crisp. Against it: a rewrite of ui-new's screens (~15.6k lines; ui-core/ carries over), I2C caps a full frame at ~23 ms either way, and nRF52 memory -- estimated ~20-30 KB RAM (heap + an 8 KB L8 buffer) and ~120-150 KB flash for a trimmed LVGL, against 74 KB RAM / 214 KB flash free on the L1 (ui-new's drawing code would go) and only 55 KB flash on the Watch13. Next step, after this roadmap: a spike on the L1 -- Home carousel with the slide and the bottom drawer, a message list with soft bubbles, the dot wave, dithering in the flush -- measured (RAM, flash, fps) and shown in the sim next to ui-new, then decide.
• E-ink (L1 e-ink, T-Echo Lite): no. Slow full refreshes rule out motion; LVGL's small dirty areas fit partial refresh poorly and would need batching. Same RAM limits as above.
• Cardputer ADV: the one candidate. A 240×135 colour screen now shows ui-new's 128×64 picture scaled up; ESP32-S3 with room to spare (flash 43%). LVGL would use the real resolution and colour, reuse Theme.h, Anim.h, the fonts and much of the ui-lvgl screen code. Open points: no PSRAM (the L2's PSRAM buffers -- tiles, lists, keyboard maps -- need smaller internal ones or dropping, no raster map), no touch (keyboard focus navigation, LVGL groups, instead of taps), a 135 px-high layout. Worth a separate pilot if the Cardputer matters; otherwise skip.

So: e-ink stays on ui-new; the OLED boards get an LVGL spike (above) once this roadmap is done; the Cardputer follows whatever the spike shows.

Backlog (found along the way)

• USB power detection: WioTrackerL2Board::isExternalPowered() reads STATUS0 (0x40) bit 7 of the chip at I2C 0x22 and always gets 0 with USB plugged in (0x40 reads 0x01). Needed for a charging indicator and for skipping the low-battery shutdown on the cable. Find the right chip / register (devices on the bus: 0x14, 0x18, 0x21, 0x22, 0x2c, 0x34, 0x48, 0x5d). Found 2026-09-26. Also wanted for a hint on the GPS screen while on the cable (see below).
• GPS screen (2026-09-26): Home > GPS and Settings > System > GPS details -- sky plot, C/N0 bar per satellite, fix / TTFF / DOPs, per-constellation counts (helpers/sensors/GpsSky.h, -D GPS_SKYVIEW; the sim feeds it made-up NMEA).
• GPS indoors (2026-09-26): diagnosed with the new GPS screen and raw NMEA logs. On the USB cable (charging) the L76K tracks satellites at 25-35 dB-Hz but never decodes navigation data: no UTC in 10 min, no fix in over an hour. On battery, same spot by the window: UTC within seconds, first fix after 377 s. Screen, WiFi, Bluetooth and LoRa TX were ruled out. Cause: noise from USB power / the charger -- a board property, not fixable in firmware. Test GPS on battery.

Ideas to come back to (2026-09-26):

• Speaker click: would keeping the amplifier on at minimum volume remove it?
• Import routes from the SD card (optional extra).
• Battery life without losing features, above all CPU sleep.
• Live tiles: don't keep them, or give them a bounded cache on the card so they never fill it.
• Vector maps and other tile sources (to think through).