3 ms·

> In fact, you can even plug in a generic TP-Link USB-C Ethernet and that works, taking priority over WiFi, and running at gigabit Ethernet speeds. Decidedly n

> In fact, you can even plug in a generic TP-Link USB-C Ethernet and that works, taking priority over WiFi, and running at gigabit Ethernet speeds. Decidedly not gigabit speeds... since USB 2 is capped at 480Mbps. Frankly - decent wifi since 802.11N, released in 2009, will outperform that dongle you just recommended. Since N is rated for 600Mpbs, and again... USB 2 caps out at just under 500Mbps. Honestly, the AX standard is fucking 20 times faster than that hardwired dongle. The device itself is slower - with the 10th gen model only supporting up to 2400Mpbs (https://support.apple.com/guide/deployment/ipad-wi-fi-specification-details-depf9bb7e412/web https://support.apple.com/guide/deployment/ipad-wi-fi-specif...) but you'll still do fuck loads better there than you would with the dongle capped at USB 2 speeds. --- Regardless - my point is that they don't want you doing direct device to device connections anywhere they can get away without it. Which is pretty clear, since they've completely hobbled a port that should be capable of up to 10Gbps down to 480 Mbps. Roughly 1/20th of the speed you should be getting out of it.

Terretta 4y ago
First, the premise they don't want you to connect things is a stretch, you can even boot Macs as an external drive and Apple remain more than happy for you to tether an iDevice to a computer to manage it or as hotspot. Second, on the speeds, above I was giving a tip about device-to-device transfer, syncing small files, a very specific scenario. The lower latency and lack of jitter from hardwire, even rate limited hardware, tends to beat WiFi for real workload sustained throughput, for this task. Can also be better for online gaming, especially in the average American home with < 300 Mbps connections. (That dongle, USB-C, runs 980 Mbps attached to any Mac w/ USB-C as well as iPad Pro, so agree, USB 2 is a limiting factor, not the dongle.) Disagree about fuck loads better. The WiFi is very likely to be worse for most readers of this thread. Very few folks know how to position devices to get "AX" ratings from the side of the box, odds are also high someone's going to want to watch Netflix while that backup is transferring (the specific device-to-device scenario I'm talking about). TL;DR: Ratings from specs aren't the best way to estimate sustained 4 - 8 hour transfers of millions of files, and latency and jitter factor more than raw bandwidth for chatty workloads. If most users compare a WiFi file sync to wired file sync, they'll generally find the wired is several times faster, even at 480 Mbps. YMMV, particularly if you tune everything and put devices in fixed positions relative to radios.
- Terretta 4y ago
  See also: https://news.ycombinator.com/item?id=33445178 https://news.ycombinator.com/item?id=33445178

Terretta 4y ago

First, the premise they don't want you to connect things is a stretch, you can even boot Macs as an external drive and Apple remain more than happy for you to tether an iDevice to a computer to manage it or as hotspot. Second, on the speeds, above I was giving a tip about device-to-device transfer, syncing small files, a very specific scenario. The lower latency and lack of jitter from hardwire, even rate limited hardware, tends to beat WiFi for real workload sustained throughput, for this task. Can also be better for online gaming, especially in the average American home with < 300 Mbps connections. (That dongle, USB-C, runs 980 Mbps attached to any Mac w/ USB-C as well as iPad Pro, so agree, USB 2 is a limiting factor, not the dongle.) Disagree about fuck loads better. The WiFi is very likely to be worse for most readers of this thread. Very few folks know how to position devices to get "AX" ratings from the side of the box, odds are also high someone's going to want to watch Netflix while that backup is transferring (the specific device-to-device scenario I'm talking about). TL;DR: Ratings from specs aren't the best way to estimate sustained 4 - 8 hour transfers of millions of files, and latency and jitter factor more than raw bandwidth for chatty workloads. If most users compare a WiFi file sync to wired file sync, they'll generally find the wired is several times faster, even at 480 Mbps. YMMV, particularly if you tune everything and put devices in fixed positions relative to radios.

Terretta 4y ago
See also: https://news.ycombinator.com/item?id=33445178 https://news.ycombinator.com/item?id=33445178