Repository navigation
fix(mobile): a tab's bar buttons no longer fade in on its first visit - #37302
Conversation
ApprovabilityVerdict: Needs Human Approval Macroscope's review found this PR requires human approval — The change prewarms iOS tab headers while avoiding full background content loading. It also adds two lint-suppression directives in WorkspaceNativeTabs.ios.tsx, which require human review. No code changes detected at You can add or adjust custom eligibility rules. Learn more. |
|
🤖 Android preview
|
Thread transfer impact✅ Thread transfer remains within every enforced ceiling.
Other scenariosAll 64 scenario measurements match All 64 scenario measurements
Baseline: Scenario and decoded snapshot size10 historical turns, 5 command tools per turn, 878.9 KiB retained MCP result per historical turn, and a 1.05 MiB retained result in the measured turn.
Updated in place by a trusted workflow. PR artifacts are strictly validated and never executed. |
|
🚀 Expo continuous deployment is ready!
|
f2c61cc to
4c7e701
Compare
166c42b to
ceeb4f4
Compare
4c7e701 to
eed5205
Compare
ceeb4f4 to
8803117
Compare
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
UIKit lays out bars and safe areas only for the selected tab, so a prewarmed tab measured a top inset 14pt short (and no bottom inset on iPhone Duo). Its list rendered with that inset, then shifted once the tab was shown. Hidden tabs now ignore their own measurements and start from what the shown tab measured for the same column. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
On iPad, Chats, PRs and Settings use the column layout and the other tabs a single stack. A prewarmed slot rendered the shown tab's tree, so opening a tab with the other layout replaced the prewarmed tree and its header faded in after all. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
68abf0a to
3c8d82e
Compare
**v0.1.10-nightly.20261116.4761** · 23 changes from 9 contributors - **Steer Grok responses.** Send follow-up messages to Grok while it is still responding, without waiting for the turn to finish. - **No project threads.** Send messages in No project threads when New worktree mode is selected. ### Plus 21 more bug fixes. <details> <summary>Full changelog</summary> ## What's Changed * fix(mobile): iPhone New chat stays beside the tabs after a Settings page by @cwilkus1-bit in pingdotgg/t3code#37021 * fix(server): dismissing a background OpenCode subagent's question lets it carry on by @juliusmarminge in pingdotgg/t3code#37319 * fix(mobile): environment filter leaves out switched-off environments by @tomhmoses in pingdotgg/t3code#32946 * fix(server): work a dead provider process was running no longer stays Running by @juliusmarminge in pingdotgg/t3code#37325 * fix(server): remote tasks report their actual model by @yazcaleb in pingdotgg/t3code#37332 * fix(provider-grok): steer queued messages into a running Grok turn by @juliusmarminge in pingdotgg/t3code#37335 * fix(web): terminal decorations stay visible under the block cursor by @StiensWout in pingdotgg/t3code#37206 * fix(clients): command palette skips tasks shown under their parent by @Onkarj012 in pingdotgg/t3code#37315 * fix(mobile): inline audio keeps playing when its row scrolls out of view by @cmcgrabby-hue in pingdotgg/t3code#37340 * fix(server): clear title markers stuck on old subagent and fork threads by @ahatzz11 in pingdotgg/t3code#37364 * fix(server): nested Claude subagent summaries no longer show Claude Code's hand-back frame by @Jacksondr5 in pingdotgg/t3code#37355 * fix(web): file tabs keep their scroll position when you switch back by @juliusmarminge in pingdotgg/t3code#37361 * fix(server): Windows PowerShell terminal reads native UTF-8 output correctly by @juliusmarminge in pingdotgg/t3code#37376 * fix(web): No project threads in New worktree mode send without a base branch by @juliusmarminge in pingdotgg/t3code#37392 * fix(mobile): Stop stays reachable while a question card is expanded by @juliusmarminge in pingdotgg/t3code#37398 * fix(server): an ACP agent's open tool call ends when its process dies mid-turn by @juliusmarminge in pingdotgg/t3code#37377 * fix(web): a disabled Merge back button shows why it is disabled by @Jacksondr5 in pingdotgg/t3code#37395 * fix(server): a delegated child's turn end re-offers only its own parent run's wake by @juliusmarminge in pingdotgg/t3code#37399 * fix(server): service protect and status say when the kernel has no memory controller by @juliusmarminge in pingdotgg/t3code#37410 * fix(server): a wake stuck on one thread's parent lock no longer holds up other threads' wakes by @juliusmarminge in pingdotgg/t3code#37416 * fix(mobile): first visit to an iOS tab no longer flashes black by @juliusmarminge in pingdotgg/t3code#37083 * fix(server): one busy thread's lock no longer stalls other threads' queued runs by @juliusmarminge in pingdotgg/t3code#37417 * fix(mobile): a tab's bar buttons no longer fade in on its first visit by @juliusmarminge in pingdotgg/t3code#37302 ## New Contributors * @cwilkus1-bit made their first contribution in pingdotgg/t3code#37021 * @yazcaleb made their first contribution in pingdotgg/t3code#37332 * @cmcgrabby-hue made their first contribution in pingdotgg/t3code#37340 **Full Changelog**: pingdotgg/t3code@v0.1.10-nightly.20261115.4749...v0.1.10-nightly.20261116.4761 </details> Upstream release: https://github.com/pingdotgg/t3code/releases/tag/v0.1.10-nightly.20261116.4761
On iOS 26, a tab's bar buttons blurred in over about half a second the first time you opened it. UIKit fades every bar item in the first time it enters a window. This isn't a Core Animation animation, so turning animations off doesn't stop it.
Follows #37083, which fixed the black frame and the title fade on a cold tab switch.
Fix
Once the first tab is up and the JS thread is idle, every other tab mounts only its header: title, search and bar buttons. The body still mounts on first selection, so no thread lists, pull requests or usage are loaded early. Usage's header reads the environment list without asking any environment for usage, and limits auto-refresh stays off.
A new iOS
Tabs.Screenprop in the Screens fork,prewarmRequest, then puts that hidden tab into the window behind the visible one for 0.4s. No appearance events are sent, so the fade happens out of sight. Header items rebuilt later, such as Settings' filter menu once its data loads, are new items and would fade again. Each header write in a tab that hasn't been shown yet requests another prewarm. (pingdotgg/react-native-screens#13, stacked on pingdotgg/react-native-screens#12 for the #37083 change.)A prewarmed tab also measures its layout while hidden, and UIKit lays out bars and safe areas only for the selected tab: the hidden PRs tab measured a top inset 14pt short, so its list rendered under the title and dropped into place a few frames after the first visit. Hidden tabs now ignore their own measurements and start from what the shown tab measured for the same column, until they're shown and measure for real.
Prewarming waits while a detail screen is pushed. Every tab renders the pushed routes too, so prewarming then would mount a thread in each hidden tab.
Evidence
iPhone 17 simulator (iOS 27.2). Each column shows the first frames after tapping a tab for the first time.
The middle state, a single prewarm at mount, still blurred Settings, because its menu was rebuilt after its data loaded: recording. With prewarming repeated after each header write: recording. A launch recording shows no frame change from the prewarm.
First PRs visit on an iPhone Duo simulator, before and after the inset fix:
Native change: needs a new binary, not an OTA.
🤖 Generated with Claude Code (Claude Opus 5.5)