Tasks Test Flaky

[ChrisMcD1]

Describe the bug
Intermittently, when running go test ./... -short on my machine I get the following error:

setting to 2
--- FAIL: TestNewCmdTask (0.28s)
    tasks_test.go:143: expected beforeStart to be called 1 times, got 0
    tasks_test.go:143: expected onEndOfInput to be called 1 times, got 0
    tasks_test.go:154: expected writer to receive the following content:
        prefix
        test

        . But instead it received:
--- FAIL: TestNewCmdTaskRefresh (1.70s)
    tasks_test.go:266: total < initialRefreshAfter: expected line counts on refresh: [100], got [0]
    tasks_test.go:266: total == initialRefreshAfter: expected line counts on refresh: [100], got [0]
    tasks_test.go:266: total > initialRefreshAfter: expected line counts on refresh: [50 100], got [0]
    tasks_test.go:266: initialRefreshAfter == -1: expected line counts on refresh: [100], got [0]
    tasks_test.go:266: totalTaskLines < initialRefreshAfter: expected line counts on refresh: [25], got [0]
    tasks_test.go:266: totalTaskLines between total and initialRefreshAfter: expected line counts on refresh: [50 75], got [0]
FAIL
FAIL    github.com/jesseduffield/lazygit/pkg/tasks      2.916s

To Reproduce

1. Run go test ./... -short
2. Get unlucky (Can't reproduce consistently)
3. See error

Expected behavior
I expect the tests to pass consistently on my local machine, like they seem to do in pipelines

Screenshots
N/A

Version info:
Building locally on 1e05055ff
git version 2.25.1

Additional context

Didn't happen when I tried reproducing while writing this bug report. Logs taken from earlier today. I also went all the back to last July and it still failed on a detached head there. So not related to recent changes.

I'm running the tests inside of Ubuntu WSL2 Kernel version: 5.10.102.1 on Windows 10, on go version go1.23.6 linux/amd64.

I'll be poking around this a bit more, but given that my tests started passing when I try to reproduce now... Might just be left open for the next person who searches this error.

[ChrisMcD1]

I'm thinking that the flakiness might be tied to the time.Sleep in the test. When it goes down from 100 milliseconds to 50 milliseconds, it starts failing almost always, and 70 ms makes it fail occassionally.

lazygit/pkg/tasks/tasks_test.go

Lines 256 to 260 in f05f81d

	go func() {
	time.Sleep(100 * time.Millisecond)
	close(stop)
	wg.Done()
	}()