Skip to content

fix(mobile): keep the board clear of the panel navigation - #338

Merged
ttbombadil merged 10 commits into
feature/browser-emulationfrom
fix/p4-4-mobile-board-nav
Oct 24, 2026
Merged

ttbombadil merged 10 commits into
feature/browser-emulationfrom
fix/p4-4-mobile-board-nav

Conversation

@ttbombadil

Copy link
Copy Markdown
Collaborator

Paket

Rest aus P4.4: Mobile-Layout je Board prüfen (seit M2 offen). Gestapelt auf #337 → #336 → #335 → #334. Die Basis ist feature/browser-emulation. Merge-Reihenfolge: #334, #335, #336, #337, dann dieser PR.

Befund

Ich habe Screenshots aller vier Boards bei 390 × 844 (Telefon, Board-Panel aktiv) aufgenommen. Die schwebende Panel-Navigation (mobile-layout.tsx, unten rechts) überdeckt Teile des Boards:

  • beim Uno die Analogpins A0–A5;
  • beim ATtiny85 die Pins PB0–PB2;
  • beim Nano und Mega die rechte untere Ecke.

Die überdeckten Pins ließen sich nicht anklicken.

Umgesetzt

client/src/index.css: Die mobile Board-Ansicht (.workspace-board-panel.workspace-mobile-full-panel im Layout mobile) hält unten 20rem frei. Das entspricht der Navigation mit bis zu fünf Knöpfen à 40 px samt Abständen, Innenabstand und Rand. Das Board wird darüber zentriert. Andere Layouts und Panels sind nicht betroffen.

Statusdatei: P4.4 und P4.2 sind erledigt. Der P4.2-Rest „kein Hardware-Serial“ ist mit P6.4 (#336) umgesetzt.

Abnahmeevidenz

  • Screenshots vorher/nachher für Uno, Nano, Mega 2560 und ATtiny85 (lokal, Playwright mit Chromium): Nachher liegt kein Pin unter der Navigation.
  • e2e/responsive-layout.spec.ts und e2e/visual-full-context.spec.ts: 12/12 grün. Die visuellen Snapshots (Desktop) sind unverändert.
  • Pre-Push mit Sonar grün.

Abweichungen vom Plan

Keine.

Geänderte bestehende Tests

Keine.

🤖 Generated with Claude Code

ttbombadil and others added 8 commits October 23, 2026 19:57
Port B with INT0 and PCINT0, timer 0 with its own registers and one-word
vectors, the ATtiny timer 1, a four-channel ADC that reads REFS2 at bit 4,
EEPROM timed for 8 MHz, watchdog and clock prescaler. INT0 and the pin-change
interrupt share GIMSK and GIFR, which avr8js wires for one of them only; the
profile wraps both hooks. Registered with the board hardware in P6.4.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
TX is decoded from PB0 with the baud rate found from the edge distances and
the bit width measured from them; RX drives PB1 through an analog comparator
model. SoftwareUart offers the USART members SerialPorts uses, for the
ATtiny85 board.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Pins 0-5 as PB0-PB5 with A0-A3 as second names, PWM on 0, 1 and 4, RESET on
PB5 and Serial as a software port on 0/1. SerialPorts takes a USART or the
software UART, the PWM observation reads timer 1 of the ATtiny85, and the
sketch parsers resolve A<n> and analogRead(n) as ATTinyCore does. The pin
map is checked against the core in the toolchain gate.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
DIP-8 board view, LED and analog labels from the board's own names, the
sketch's Serial.begin() rate for the software UART, five corpus probes, an
end-to-end test and the user guide with the board's limits. The software
UART now judges every fitting baud rate on its frames and decodes TX only
while Serial is active. M3 acceptance report.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On phones the floating panel navigation covered pins of the board (Uno
A0-A5, ATtiny85 PB0-PB2). The mobile board view now leaves room for it.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
ttbombadil and others added 2 commits October 23, 2026 22:22
…ny tests

The gateway deployment test listed the pilot's boards by hand and failed once
the ATtiny85 became preview; it now takes them from the catalog. The ATtiny85
board tests stop as soon as their result is there and share one probe run
(16.8 s to 5.6 s under coverage).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@ttbombadil
ttbombadil merged commit 03f4f14 into feature/browser-emulation Oct 24, 2026
6 checks passed
@ttbombadil
ttbombadil deleted the fix/p4-4-mobile-board-nav branch October 24, 2026 08:30
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant