Tell whether the network panel's Timeout or Ethernet reading is real
Read /usr/bin/omarchy-network-status on 4.0.4-1: internet_probe=1.1.1.1, ping -n -c 1 -W 1, interface from ip route get, anything without /sys/class/net/<iface>/wireless reported as ethernet, IPv4-only route and address lookups. Issues #9068, #11215 and #7438 are open and say what the record says, and PR #8866 is open. Two claims fail. The quattro copy is not unchanged: it drops the omarchy-cmd-present nmcli guard, though the probe logic is the same. More important, the record says the Wi-Fi network list is empty behind a VPN, and tells the reader to manage Wi-Fi from the terminal for that reason. Issue #7438 does not say this, and Panel.qml binds the list to wifiStationAvailable, which comes from the NetworkManager service's Wi-Fi device (line 607), not from the status script. The issue also says the bar icon stays correct. Symptom, cause and fix corrected to drop that claim. Ran curl https://1.1.1.1/ here and got 301, which matches the verify step. Not exercised: no VPN or IPv6-only network.
Symptom
The Omarchy network panel reports Ping: Timeout and Packet loss: 100% while pages load fine, typically on eduroam, a corporate or hotel network, or behind CGNAT. Or on an IPv6-only Wi-Fi the panel says Timeout while IPv6 browsing works. Or with Mullvad, WireGuard, Cloudflare WARP or another VPN up, the panel header says Ethernet (sometimes Ethernet (10gbit)) and shows the tunnel IP although you are on Wi-Fi. The bar icon still shows Wi-Fi.
Cause
All three come from how /usr/bin/omarchy-network-status, which feeds the panel header and details, measures the link. The probe logic is the same on 4.0.4-1 and on quattro.
It measures internet reachability with one ICMP echo to the hardcoded 1.1.1.1 (ping -n -c 1 -W 1). Networks whose egress policy allows only TCP and UDP ports drop ICMP, so every sample is empty and the panel counts each as lost (issue #9068).
The probe and the route lookup are IPv4 only, so on an IPv6-only or 464XLAT network where IPv4 to 1.1.1.1 does not work the panel reads Timeout while IPv6 traffic flows (issue #11215).
It picks the active interface with ip route get 1.1.1.1 and calls anything without /sys/class/net/<iface>/wireless Ethernet. A VPN that holds the default route, directly or by policy routing, returns its tunnel device, which has no wireless directory, so the script reports Ethernet with the tunnel's address, and a tunnel that reports a bogus link speed shows as 10gbit (issue #7438, PR #8866 open on 2026-10-05). The bar icon and the Wi-Fi list come from NetworkManager directly and stay correct.
In all three cases the connection is fine and only the panel's reading is wrong.
Fix
Check the real state instead of the panel:
curl -sS -o /dev/null -w '%{http_code} %{time_connect}s\n' https://1.1.1.1/ # TCP works where ping is filtered
curl -6 -sS -o /dev/null -w '%{http_code}\n' https://archlinux.org/ # IPv6 path
nmcli -t -f DEVICE,TYPE,STATE,CONNECTION device status # the physical link behind a VPN
omarchy-network-status --verbose # what the panel is being told
If either curl returns an HTTP code, the network is working and the panel is wrong. No setting changes the probe on 4.0.4: the target is hardcoded, and /usr/bin/omarchy-network-status is owned by the omarchy package, so an edit there is overwritten on the next update. The fixes are upstream in PR #8866 for the VPN case and pending for the other two.
Verify
curl https://1.1.1.1/ returns 301 (or, on an IPv6-only network, curl -6 https://archlinux.org/ returns 200) and nmcli device status shows the Wi-Fi device connected, which together show the panel reading is the only fault.
Sources
- https://github.com/omacom/omarchy/issues/9068
- https://github.com/omacom/omarchy/issues/11215
- https://github.com/omacom/omarchy/issues/7438
- https://github.com/omacom/omarchy/blob/quattro/bin/omarchy-network-status
- https://github.com/omacom/omarchy/pull/8866
- https://github.com/omacom/omarchy/blob/quattro/shell/plugins/panels/network/Panel.qml