Commit Graph
2 Commits
Author SHA1 Message Date
jaap-janandClaude Opus 5 57d4b30557 Put the app's own mark on the launcher
ci / build and test (push) Failing after 3s
ci / api image (push) Skipped
ci / android head (push) Failing after 2s
The sign-in and locked screens both draw the same thing — a square outline in the accent
with >_ inside it — and the launcher was still showing the stock Android silhouette, so the
icon somebody taps and the icon the app opens onto had nothing to do with each other.

Redrawn as a vector rather than exported from the screen as a bitmap. There is one geometry
here and no set of density buckets to update four of and forget the fifth, and the accent
stays a number that can be diffed against Palette.axaml rather than a colour baked into a
PNG. The hex is written out because an Android resource cannot reference a XAML dictionary
— the same duplication colors.xml already carries for the window background, with the same
obligation attached.

Adaptive only, no raster fallback. Adaptive icons landed in API 26 and this head requires
28, so there is no device it ships to that would need the bitmaps; density buckets exist to
choose between PNGs and there is nothing to choose. The background layer is the same
@color/dodo_window as the window and the status bar, so the mark sits on the app's own
near-black rather than on a second one almost like it.

Two departures from the screen, both because a launcher is looked at much smaller than a
sign-in header. The box is 42 across rather than the 48 that first suggested itself: 72 of
the 108 survives masking, but that is a width, and a square meets a circular mask at its
corners — at 48 they land 33.9 out against a radius of 36 and read as clipped despite
technically clearing it. And the strokes are 2.2 and 2.8 where proportional fidelity to a
1px border on 44px would be 1.0, which a launcher drawing this at 48dp would render as half
a pixel of nothing.

A monochrome layer too, for the Android 13+ themed-icon setting. Without one a launcher
with themed icons on falls back to the full-colour icon, which would leave this the single
green thing on an otherwise recoloured home screen.

Verified in the packaged APK: the icon resolves at all five densities, the three layers
resolve, and the compiled vector carries the geometry above. The launcher rendering itself
was checked against local renders under circular and squircle masks at 144 and 64 px, not
on the device — the phone was locked.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-01 13:46:40 +02:00
jaap-jan fe9d7fc289 Give DodoSSH a phone, and a shared shell for both heads to drive
The Android head from docs/android-port.md, taken as far as its step 6.

Step 3, the spike, is answered and its throwaway screen is gone: libsodium.so and
libe_sqlite3.so are both in the arm64 APK, so NSec resolves its native half on Android
despite shipping no Android build, and the local cache opens. Two findings the audit
could not have had: Avalonia.Controls.WebView only ships net10.0-android36.0, which
settles the open "which Android versions" question at targetSdk 36; and Android has
blocked cleartext HTTP since API 28, so the terminal renderer needs a network security
config scoped to 127.0.0.1 or the WebView loads nothing.

DodoSSH.Client.Shell is new and is why the phone can exist: the view models, the terminal
renderer files and the palette moved there so both heads drive one state machine and draw
from one set of tokens. The desktop head is otherwise untouched and its 144 tests still
pass.

The platform pieces behind interfaces that already existed: the profile directory from
filesDir, a device key wrapped by a StrongBox-backed key that a fingerprint releases, and
a foreground service so a shell outliving a vault lock stays true on a platform that
stops backgrounded processes.

Sign-in is deliberately absent rather than approximated. It needs an app link, because
reusing the desktop loopback listener is the attack RFC 8252 section 8.3 names.
2026-07-31 20:58:48 +02:00