Black screen on connect – macOS Tahoe 26.5.2, Hackintosh/OpenCore VM

[Sage5182]

Setup:
– NoMachine 9.8.2 (server), macOS Tahoe 26.5.2
– Hackintosh/OpenCore (OSX-KVM) VM running on Proxmox/QEMU
– VMware-compatible emulated display device

This exact configuration worked fine under macOS Sequoia (15.x) on the same VM. The problem started immediately after upgrading to Tahoe, with no other changes made.

Symptom:
Connecting from my NoMachine client on my local desktop, the remote desktop window goes solid black the moment the connection completes.

– The mouse cursor visibly moves and responds on the Mac’s own console (viewed via Proxmox’s noVNC) while the NoMachine window stays black — input is reaching the session fine, but no video is being captured or rendered by NoMachine.

– On subsequent connect attempts, instead of the cursor moving, the Mac drops to its login/lock screen, and keyboard input is blocked at the console entirely until I disconnect the NoMachine client.

Troubleshooting already done:

– Removed and re-granted Screen Recording and Accessibility permissions for NoMachine in System Settings > Privacy & Security — no change.

– Confirmed display sleep, screen saver, and “require password after sleep” are all disabled.

– Ruled out a conflict between Proxmox’s noVNC console and NoMachine running simultaneously — same black screen occurs even with noVNC fully closed before connecting.

– Reviewed the nxserver log and found a recurring “NXNODE ERROR! Agent: Window Server is not available” entry logged during an active connected session.

I’ve enabled debug logging per DT07S00243 and reproduced the issue — happy to attach/upload the server log archive here or send it by email,
whichever you prefer.

Has anyone else hit this specifically on Tahoe? Any idea whether this is a permissions/entitlement change on Apple’s side (ScreenCaptureKit) or something NoMachine needs to adjust for on Tahoe?

Thanks!

[Britgirl]

Hi, we aren’t aware of issues with with ScreenCaptureKit on Tahoe.
Logs from the macOS server would also be useful. Go to Server settings > Security > scroll down and Take logs.

You could also try adding the -nosckgrab parameter in the NoMachine node.cfg file on the server side. In /Applications/NoMachine.app/Contents/Frameworks/etc/node.cfg, append DisplayServerExtraOptions "-nosckgrab", save, then restart the nxserver (Server settings > Status > Restart). Does this make any difference?

[Britgirl]

I just noticed you submitted some logs already, we’ll check those and come back to you.

[Author]

Posts