mirror of
https://github.com/jesseduffield/lazygit.git
synced 2026-09-11 08:06:25 -04:00
94e5f570f6
The previous two commits stop gocui from flushing while suspended, but that guard cannot be fully airtight from gocui's side: it is a check-then-act on the suspended flag, so a flush racing the suspend itself (e.g. from a spinner goroutine) could still reach the screen just as it disengages, and tcell's disengageFinish mutates the cell buffer without holding the screen lock. Upstream now closes this at the source (gdamore/tcell#1139): draw() returns immediately on a disengaged screen, and the draw scan loop can no longer stall on the width-0 cells that a released cell buffer reports (#5309). The delta over the previously pinned snapshot is these two fixes, a CSI R input decode fix, a wasm packaging chore, and dependency bumps. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> |
||
|---|---|---|
| .. | ||
| aliases.go | ||
| dll_windows.go | ||
| env_windows.go | ||
| eventlog.go | ||
| exec_windows.go | ||
| memory_windows.go | ||
| mkerrors.bash | ||
| mkknownfolderids.bash | ||
| mksyscall.go | ||
| race.go | ||
| race0.go | ||
| security_windows.go | ||
| service.go | ||
| setupapi_windows.go | ||
| str.go | ||
| syscall.go | ||
| syscall_windows.go | ||
| types_windows.go | ||
| types_windows_386.go | ||
| types_windows_amd64.go | ||
| types_windows_arm.go | ||
| types_windows_arm64.go | ||
| zerrors_windows.go | ||
| zknownfolderids_windows.go | ||
| zsyscall_windows.go | ||