It compiles, the test is green, and the screen is a grey box
A structural shell passes an AST scan and a unit test while rendering nothing the mockup specified. Only a rendered check can tell the difference.
compose
129 posts
A screen nothing navigates to still compiles
A route with no caller compiles, registers and never runs. Reachability is a graph property, so no per-file check can see it and nothing fails.
A hardcoded dp is a token you did not name
A literal spacing value is a token nobody named. The cost is not inconsistency now but that a redesign cannot distinguish values that happen to match.
The string literal that cannot be translated
A hardcoded string is not just untranslated — it is invisible to the sweep that finds untranslated text, so coverage reports complete while it is not.
Compose is doing what you told it; the instruction was wrong
Recomposition scope is set by where a value is read, not where it is declared. Most Compose performance problems are a read placed higher than it needed to be.
A testTag is a contract between the screen and its test
Tests that assert on visible text break on copy edits, which trains people to weaken assertions. A test tag is a deliberate contract that survives rewording.
The Scaffold parameter everyone ignores is the one that breaks the layout
Scaffold arranges the app shell and hands back the insets its own bars consume. Ignoring that PaddingValues is the single most common Compose layout bug, and edge-to-edge makes it worse.
An external display is a second window, not a bigger one
Android 16 makes external displays a mainstream concern. The layout work is mostly window size classes you already have; the new parts are density, input capability and deciding what belongs on which screen.
Wiring ML Kit GenAI into a Compose screen without blocking it
Wiring an on-device GenAI capability into Compose is mostly state design. Model the availability and failure states before the happy path, not after.
Streaming tokens into Compose without recomposing the world
Streaming output updates state far faster than a normal screen. Narrow what reads the changing value, and let the smallest possible scope recompose.
How do you test a screen whose text is generated?
Snapshot tests die the moment output varies per run. Asserting invariants instead of strings, the vacuous-check trap, and a verdict that means could-not-determine.
adb install -r doesn't kill your app, so your screenshot is a lie
Reinstalling over a running app can leave the old composition on screen. A screenshot taken without a force-stop cannot say which build drew it.