[Feedback]: Main window never becomes visible until the app is launched a second time — ready-to-show does not fire while the window is hidden
Nobody has claimed this yet.
Assessment
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Newbie friendliness
- 72/100
- Issue type
- Bug
- Clarity
- Clearly specified
- Activity status
- Active
- Tech stack
- electron, javascript
- Domain
- desktop
Research direction
Start by reproducing the packaged Windows launch from the issue and inspect out/main/index.js, especially the hidden window setup and ready-to-show, second-instance, deep-link, and notification-click handlers. Check how visibility behaves before the first paint, including background throttling and the reported GPU fallback. Done means the first launch reliably shows the main window without requiring a second instance and avoids an unexplained black flash.
Written by the indexing model from the issue text.
Description
What problem are you trying to solve?
Summary
After installing, the first launch appears to do nothing: no window, no taskbar entry, no error — for as long as you wait. Launching the app a second time brings the window up in under half a second. This happens on every launch, so the app can only be opened by launching it twice.
Environment
- Command Code Desktop
0.1.31(packaged) - Installer
CommandCode-0.1.31-x64-setup.exe, size145234320, sha512Kap6h9HtsAdGCqAsGiCo/gddarFzRbdUn+l1a7Gov0gjYyHHb35t3aM362xR9wxNpSuWOWDzfY+pQpp9G33LIA== - Windows 11 (Windows PowerShell 5.1.26100.9482)
- Display 1600x900 @ scale 1.2
- GPU vendor
4318/ device8580, driver32.0.16.1692
Steps to reproduce
- Quit Command Code completely.
- Launch
Command Code.exeonce. - Do not touch anything — just watch the screen.
Expected
The window appears within a couple of seconds.
Actual
No window appears. Verified with a 60-second instrumented watch (user32 EnumWindows polled every 100ms): no window was ever reported visible. Launching a second instance at t+60s made the main window visible at t+60.89s — 0.44s later.
Evidence
Window-visibility timeline alongside main.log:
t+0.00s launched Command Code.exe (pid 16228)
t+0.69s main.log: Command Code starting - v0.1.31 (packaged: true)
t+0.83s main.log: [startup] ... main.harness-loaded +44ms (837ms) <- window created, hidden
t+60.21s *** launched a 2nd instance ***
t+60.29s PROC+ pid=9736 (2nd main process)
t+60.89s WIN+ MAIN APP WINDOW FIRST BECAME VISIBLE
hwnd=2164394 class=Chrome_WidgetWin_1 size=1741x1042 title=[Command Code]
t+61.45s main.log: [startup] ... main.ready-to-show +59988ms (60825ms)
Two observations:
-
ready-to-showfired at+59988ms, i.e. at the moment the second instance arrived — not when the renderer finished a fixed amount of work. Across 5 launches the value tracked only how long the user waited (22012ms, 25038ms, 27024ms, 27930ms, 59988ms). -
Chromium's own log places the second instance immediately before it:
[16952:0917/175428.395:VERBOSE1:chrome\browser\process_singleton_win.cc:114]
Handling STARTUP request from another process
28ms later ready-to-show fired; 130ms after that the window became visible.
Root cause
out/main/index.js — the window is created hidden with a black background:
backgroundColor: "#000000",
show: false,
and is only ever revealed from the ready-to-show handler:
win.on("ready-to-show", () => {
...
if (!win.isDestroyed()) win.show();
});
The only other callers of win.show() are the second-instance, deep-link and notification-click handlers:
app.on("second-instance", (_event, argv) => {
...
const win = pickWindowToReveal(mainWindow, appWindows);
if (!win) { createWindow(); return; }
if (win.isMinimized()) win.restore();
win.show();
win.focus();
});
So visibility depends on ready-to-show firing for a window that is not yet visible. On this machine it never does while hidden, so the window is never shown — until a second instance calls win.show(), at which point the first paint finally happens and ready-to-show fires.
I can't prove from outside the app why the hidden window's first paint is deferred. Worth checking: hidden-window rendering throttling (webPreferences.backgroundThrottling is left at default), and a GPU/compositor problem — the run log shows Chromium later relaunching the GPU process in software mode (--disable-gpu-sandbox --use-gl=disabled).
This also explains the "black window flashes for a moment" that users report: backgroundColor: "#000000" means the window is shown as a solid black rectangle before the UI paints.
What would improve it?
Suggested fix
Do not make the first paint a hard precondition for visibility. A fallback timer is sufficient:
const reveal = () => {
if (!win.isDestroyed() && !win.isVisible()) win.show();
};
win.once("ready-to-show", reveal);
setTimeout(reveal, 3000);
Also worth considering webPreferences: { backgroundThrottling: false }, and reconsidering the show: false + "#000000" combination — if the window is shown early by any path it renders as a black rectangle, which is what users are seeing.
What do you do today?
No response
Product area
None
- Dominant language
- Shell
- Stars
- 80
- Forks
- 2
- PR merge metrics
- No merged PRs in 30d
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from CommandCodeAI/desktop
-
enhancement
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
CommandCodeAI/desktop#94 ·
-
enhancement
Difficulty 2/5 1-3 hours Newbie friendliness 68/100
CommandCodeAI/desktop#66 · 1 comment · 2 reactions ·
-
bug
Difficulty 4/5 3-5 days Newbie friendliness 35/100
CommandCodeAI/desktop#105 ·
-
enhancement
Difficulty 4/5 3-5 days Newbie friendliness 45/100
CommandCodeAI/desktop#104 ·
-
bug
Difficulty 3/5 1-2 days Newbie friendliness 70/100
CommandCodeAI/desktop#103 ·
All issues in CommandCodeAI/desktop
Similar issues
-
Issue-Enhancement Needs-Triage
Difficulty 1/5 Under an hour Newbie friendliness 86/100
PowerShell/PowerShell#28061 · 2 reactions ·
-
Feature Request: Add ability to load custom environment variables in linux-exec-server-installer.sh Open
Difficulty 2/5 1-3 hours Newbie friendliness 84/100
microsoft/vscode-remote-release#11867 ·
-
AuTest Bug Tests
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
apache/trafficserver#13714 ·
-
Update to NCCL 2.32 Open
Difficulty 2/5 1-3 hours Newbie friendliness 70/100
conda-forge/nccl-feedstock#166 ·
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 86/100
vllm-project/agentic-api#358 · 1 comment ·