Repository navigation
docs: update Krux Installer usage for v0.0.24 - #970
BLVCK-MAMBA-6 wants to merge 3 commits into
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## develop #970 +/- ##
========================================
Coverage 97.60% 97.60%
========================================
Files 84 84
Lines 11001 11001
========================================
Hits 10738 10738
Misses 263 263 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
|
I believe the ci action test runs will resolve once this pr is rebased against "develop". Lots of changes went into it yesterday, including ci action changes that fix black failures. alerting for review: @qlrd @joaozinhom since these docs concern usage of krux-installer. |
6b1e1bd to
d251b29
Compare
|
Thanks, @jdlcdl. I rebased the PR onto |
There was a problem hiding this comment.
/root/.local/krux-installer isn´t a good place to give as example. Prefer another image with a proper example name.
There was a problem hiding this comment.
Now I'm reading /home/user/krux-installer. But IIRC the path is is similar to $HOME/.local/krux-installer (need peer verification from others too). Also the image appear to be very small to me (i put some zoom to read it), is good to read on your video guys?
There was a problem hiding this comment.
Thanks for the clarification, @qlrd
I'll increase the displayed image width to improve readability.
Regarding the path, would /home/user/.local/krux-installer be an appropriate example based on the default $HOME/.local/krux-installer location? I'm currently testing on Windows, so I can't independently verify the Linux default.
There was a problem hiding this comment.
let's wait, i really not remember the correct path on unix-like cc @joaozinhom
There was a problem hiding this comment.
Confirmed: the default is $HOME/.local/krux-installer on both Linux and macOS (Windows: %LOCALAPPDATA%\krux-installer\local). So the /home/user/krux-installer in the screenshot is not the default; a retaken screenshot showing /home/user/.local/krux-installer would match a fresh install.
One thing worth noting about this setting: I checked the code and the "Asset's destination path" entry is a leftover from the download era. Since v0.0.22 the installer bundles the firmware and downloads nothing; the bundled .kfpkg files are unpacked by PyInstaller into its own temp directory and by KTool into a temporary directory at flash time, and nothing reads the destdir.assets config anymore. So the "Destination path of downloaded assets" description in Settings doesn't match any current behavior. I'd suggest not documenting it as a download path; maybe just drop that sentence and keep baudrate and locale.
| Krux Installer bundles one version of the Krux firmware. Once installed, it can | ||
| flash your device offline. To use a different firmware version, download an | ||
| [installer release](https://github.com/selfcustody/krux-installer/releases) that | ||
| bundles it or follow the [manual installation instructions](../from-pre-built-release.md). |
There was a problem hiding this comment.
| Krux Installer bundles one version of the Krux firmware. Once installed, it can | |
| flash your device offline. To use a different firmware version, download an | |
| [installer release](https://github.com/selfcustody/krux-installer/releases) that | |
| bundles it or follow the [manual installation instructions](../from-pre-built-release.md). | |
| Krux Installer bundles one version of the Krux firmware. Once installed, it can | |
| flash your device offline. To use a different firmware version, download an | |
| [installer release](https://github.com/selfcustody/krux-installer/releases) that | |
| bundles it or follow the [manual installation instructions](../from-pre-built-release.md). It should be used to first install or to install after a [*Wipe device* procedure](#air-gapped-update). To update firmware, we recommend to see (*Air-gapped update* section)[#air-gapped-update]. |
|
Hi @qlrd, thanks for the review! I clarified the USB flashing and air-gapped update instructions and updated the Settings screenshot. I couldn’t capture the flash-progress screenshot because I don’t have a compatible device to test with. The strict MkDocs build passes locally. |
|
I widened the Settings screenshot from 640px to 960px so the text is easier to read. For the asset destination path, I’m still waiting on confirmation of the right Linux example. |
qlrd
left a comment
There was a problem hiding this comment.
LGTM i think only confirming the correct path for image. Maybe a follow-up could be done with example for windows and macos, since the paths could diff
|
Thanks, @qlrd! I’ll wait for @joaozinhom confirmation. Platform-specific examples for Windows and macOS would be a good follow-up for sure. |
|
LGTM from my side. I verified the guide against the v0.0.24 code: offline flashing, the dimmed devices (bit is correctly dimmed, it has no bundled kfpkg), the SHA256 check before flashing, the wipe text, and the rewritten sections all match the installer behavior. The internal links also resolve fine. This review surfaced a few loose ends in the installer itself that came from the offline refactor: the destdir.assets setting has no consumer anymore, DEFAULT_DESTDIR in base_krux_installer.py is unused, and the airgap update screens look unreachable from the UI (that is why pointing to the microSD guide in this PR is the right call). I can open a follow-up issue/PR to remove or clean up that dead code. Does that sound good to you, @qlrd? |
What is this PR for?
The Usage guide still showed the old firmware download screens. This PR updates the steps and screenshots for Krux Installer v0.0.24, which bundles the firmware and can flash it offline. The guide also links to the microSD update instructions.
Refs #941
Changes made
What did you test?
I checked the installer screens on Linux. Pre-commit and the strict MkDocs build passed. Pytest reported 1,194 passed and one expected failure. Flashing and wiping on a physical device are still untested.
What is the purpose of this pull request?