Repository navigation
OnDelay-release 26.09.0 (reviewed and ready) - #948
Conversation
Shown on the About page and, until acknowledged for the running version, before the application can be used. The acknowledged version is stored on flash, so a firmware update shows it again.
The disclaimer shown on the device is now also in the README and in the getting started page. Wording that might resemble marketing was removed.
Although space is limited, the addition of the term, coherent with documentation, help inform those who don't read it. Split the label into a fixed "Estimated entropy:" prefix, drawn once, and a status line with Good, Poor! or Insufficient!, redrawn on change. Reserve room for the extra line: crop the camera preview where it reached the last text line, and move the flash filling progress bar up.
…om camera Camera frames are stored unscrambled, so a flash dump can reconstruct them for verification. Adds a privacy warning to the tamper detection docs and a call to read the documentation at the start of the TC Code setup flow.
…tions TC-code - warn about privacy when filling flash with entropy from camera
FIX: handle back
…-context feat: show units and ranges for numeric settings
…emonic-loading feat: make mnemonic loading format-first
chore: bump embit to v0.8.2
pulling diverged main into develop prior to release
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #948 +/- ##
==========================================
- Coverage 97.79% 97.59% -0.20%
==========================================
Files 84 84
Lines 12043 10977 -1066
==========================================
- Hits 11777 10713 -1064
+ Misses 266 264 -2 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
qlrd
left a comment
There was a problem hiding this comment.
Overall LGTM. Reviewed with:
git fetch pull/948/head && git checkout FETCH_HEAD
git submodule update --recursive
uv sync --frozen
uv run poe pre-commitAnd manual lecture from memory (since reviewed some). I will do more lecture and manual tests and LYN.
As relative note for follow-up: made a patch for uncovered line on #949 if agrees.
| latest_installer_win: krux_installer_v0.0.24_Setup.exe | ||
| latest_installer_mac_arm: krux_installer_0.0.24_arm64.dmg | ||
| latest_installer_sha: krux-installer.SHA256.txt | ||
| latest_installer_sig: krux-installer.SHA256.txt.sig |
There was a problem hiding this comment.
I used .asc as historical usage on last release. Should we provide both or just sig ?
There was a problem hiding this comment.
I must have missed something.
There was a problem hiding this comment.
no problem to me put both or sig, just asking what you think is better (historically on krux, is sig, but the default on many projects is asc, so i'm really in doubt about that)
There was a problem hiding this comment.
I've corrected the shasums filenames to include the version (that I had missed too), and sig now has ascii-armor extension.
|
tACK at 7894669 |
41b9e41 to
28b71bc
Compare
|
Opended jdlcdl#16 to cover the lines for codecov |
…_lines This commit add a `test_init_embed_fire` and a `test_render_image_with_extra_bottom_line` for region cut in `render_image`. Instead of adding boards to the `multiple_devices` fixture, which would grow every parametrized test time, i propose that the `"dock"` branch of the render tests would come a shared config on some display tests, so it could be an `else` branch for other devices (yahboom, wonder_mv, wonder_k share some default values, like region and font size, so they assert same values). They are being silently skipped while being called anyways, so this `else` branch is just another test without significant time increase.
since it doesn't change code, yet fills-in test-coverage for current release w/o altering build reproducibility, I've cherry-picked your commit in PR#949 into this release branch. Thank you again for raising the bar. |
always for team and community! |
qlrd
left a comment
There was a problem hiding this comment.
one thing that came to mind if we could apply (or in a follow-up the proposed refactors on imports did on #930) cc @joaozinhom
|
ACK, code review shared on devs Telegram group |
This commit add an extended hash check for developers while verifying hashes of both firmware binaries as well ktool binaries.
|
re-ACK 89cb71f |
What is this PR for?
Release has been delayed.
All previously proposed+reviewed+ACKed+approved PRs merged into krux "develop", since release 26.08.0 will be merged into "main" for release 26.09.0.
...plus: unseen chores:
Changes made to:
Did you build the code and tested on device?
What is the purpose of this pull request?