Wireless across the room. Ethernet at the Apple TV.

Based on August–September 2026 Codex work records, handover notes and retained programs. Generated system illustrations are not photographs of the installation. Checks are limited to the dates and scope recorded.

The Apple TV was connected directly to the Raspberry Pi 3 Ethernet port, while the Pi used Wi-Fi for its uplink. A VXLAN tunnel between OpenWrt and the Pi bridged the device into the existing LAN.

The aim was to meet a specific connection requirement with a small additional device. Agent work covered interface configuration, startup behavior and troubleshooting notes; the operator checked the physical wiring and playback.

Playback was sampled every five seconds over 25 seconds on 26 August 2026. Recorded download traffic was approximately 13.44–13.50 Mbps, with zero packet loss in the observation and typical stabilized latency around 6–18 ms. Playback was uninterrupted during that check.

These are short observations from one location and wireless environment. They do not establish all-day reliability or equivalent results at other video bitrates. The notes also record a temporary latency increase during initial background synchronization.

The Pi Ethernet port was reserved for the direct Apple TV connection. Connecting that port back to the upstream switch while the bridge is active can form a network loop, so the allowed wiring was documented explicitly.

Troubleshooting follows power and cabling, Wi-Fi, bridge and tunnel state, then the endpoint address. The handover included configuration backups and recovery steps as well as automatic startup behavior.

A repeat evaluation should separately test recovery after reboot, weaker wireless locations, competing downloads and extended playback. Interruption counts, worst observed latency and recovery time can explain usability better than one average throughput value.

No matched baseline was collected for development time. The case therefore focuses on the working configuration, the limits of its measurement and the handover, rather than claiming a productivity multiplier.

The workflow separated Apple TV's physical Ethernet-link requirement from the difficulty of running a cable across the room.

Apple TV connected directly to the Raspberry Pi 3 Ethernet port; the Pi joined the access point over Wi-Fi.

A VXLAN tunnel between the Pi and OpenWrt on the NAS joined both bridges, carrying DHCP and TV traffic over the wireless path.

Live B tv playback and stream traffic were checked. After an AP change interrupted the link, recovery was checked from the Pi through the tunnel to Apple TV.

26 August 2026: the B tv app recognized a wired connection and played live TV.

The 25-second sample recorded an approximately 13.5 Mbps stream and 0% measured packet loss; this is not a long-term guarantee.

The 9 September recovery record confirms Pi/Apple TV reachability, VXLAN and HomeKit announcements. The setup does not change subscription or account entitlement.

NAS OpenWrt → switch → wireless AP ⇢ Wi-Fi / VXLAN ⇢ Raspberry Pi 3 → short Ethernet → Apple TV → B tv