Repository navigation
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #177 +/- ##
==========================================
+ Coverage 91.04% 91.14% +0.10%
==========================================
Files 27 27
Lines 5124 5104 -20
==========================================
- Hits 4665 4652 -13
+ Misses 459 452 -7 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
…ation Breaking: the config page magic moves to SWRP, so every robot reads as uncalibrated until it is cabled with a page baked by a matching host, and a push only commits from a host sending the mask encoding. Flash the bootloader and the net core together. AI-assisted: Claude Opus 5.5
…st its centre AI-assisted: Claude Opus 5.5
AI-assisted: Claude Opus 5.5
Points at the unmerged DotBot-libs branch adding drv/lh2_geometry; move it to the merge commit once that lands. Both bootloaders now link 00drv_lh2_geometry, which the camera point moved into. AI-assisted: Claude Opus 5.5
…r lines Adds swarmit_localization_get_lines, which moves the init_rng, get_position, get_min_tx_interval_us and ipc_isr veneers: every sandbox app must be rebuilt against this cmse_implib.a and flashed after this bootloader. The secure stack grows to 2 KiB, since the line solve runs on it under swarmit_keep_alive. AI-assisted: Claude Opus 5.5
Breaking: the net core no longer accepts 0xA2 and the bootloader no longer sends 0xCA-tagged log events. Raw counts reach the host only through the calibrate apps, which read them over the NSC raw-count call. AI-assisted: Claude Opus 5.5
AI-assisted: Claude Opus 5.5
Breaking: removes request_lh2_capture from the clients and controller, POST /lh2_capture, PayloadLH2Capture and LH2_CALIB_TAG. PyDotBot's corner-calibration collect imports these, so it lands together with the PyDotBot change that removes collect --device. AI-assisted: Claude Opus 5.5
geonnave
force-pushed
the
lh2-no-capture-pr
branch
from
October 16, 2026 14:32
b373872 to
7f314a2
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stacked on #174 (on #173): this PR's own change is its last three commits.
The bootloader still carries a raw LH2 capture from before the calibrate apps existed: in READY mode, with no app running, the host could ask a robot for one raw sample of LH2 counts and get it back in the event log. Corner and spin calibration now both collect through a calibrate app over the air, so nothing uses that path any more, and the compiled-in
lh2_calibration.hit sat next to is unused too. This removes both, from both cores and from the host.Breaking changes.
0xA2LH2_CAPTURErequest is gone and its id is marked reserved (as0xA1already is). The net core no longer forwards it, IPC channel 9 is gone, and the bootloader no longer writes0xCA-tagged capture samples into the event log.request_lh2_captureis removed from every client, as arePOST /lh2_capture,PayloadLH2CaptureandLH2_CALIB_TAG. PyDotBot's calibration session stops calling them in a companion PR, and the two should be released together.Flashing. Bootloader and net core together, by cable, with the calibration baked. The NSC table is unchanged, so sandbox apps already built against #174's import library run as they are and need no reflash. A robot whose bootloader still has the capture is unaffected by a new host apart from the capture call no longer existing.
This PR's own files:
device/bootloader/Source/main.c(+1 -49), and 16 small files:AGENTS.md,device/bootloader/Source/{ipc.h, lh2_calibration.h, localization.c, localization.h, protocol.h},device/bootloader/bootloader-app.emProject,device/network_core/Source/{ipc.h, main.c, protocol.h},swarmit/client/{__init__.py, http.py, local.py},swarmit/testbed/{controller.py, protocol.py, webserver.py}(+5 -106).Validation.
main: this PR's three commits applied without conflicts andgit range-diffshows them identical. The bootloader and net core build clean at this tip (SES, dotbot-v3 Debug). This branch itself has not been flashed.Left for the docs sweep before merge:
doc/wire-protocol.md(the0xA2row and the0xCAparagraph),doc/swarmit.lua(the raw-capture dissector), and the README link to the deleted header.