Wireless USB | Hacker News Reader

Wireless USB

en.wikipedia.org

34 pointsfirefoxd15 comments

I had a wireless USB hub from iogear. I thought it was awesome; I could put my printer on the other side of my apartment. The speed was slow so anything that depended on speed was not great, like a flash disk, although it did work.

The worst part was the drivers. Instead of something totally invisible, it required the installation of these ugly drivers. They were written by a company other than the manufacturer and were never maintained. They required an old version of windows. Eventually I ran it in a VM and shared the printer on the network. When I finally got a wifi printer, I no longer had any use for the thing.

I was so excited for the standard and for devices to be native wireless USB and to have built on drivers on all operating systems. Searching for updates or new products using the standard was very hard because "wireless usb" just turned up wifi dongles.

My suspicion is that WUSB died because it used UWB which was/is very difficult to hack. The signal looks like random noise distributed across a wide spectrum and if you don't know the secret key it's hard to tell if there is even a signal there. If WUSB had made UWB widely available and cheap it seems likely it could have been hacked to create cheap very secure point-to-point wireless links for anyone and everyone. UWB has been around for a decade or so, but there are still few commercial products and the ones that do appear tend to later disappear into the military world (e.g., Aetherwire). The one exception is UWB tracking tags, which Samsung just introduced and Apple is about to introduce (and there are several other companies making them like Nanotron). Precision location has some great applications but also some scary ones. Non-interceptable secure short-range data links may soon follow if the tags/phones turn out to be hackable. It will be interesting to see how this current foray into UWB goes.

Too bad that didn't become a thing. Sounds like a good alternative to bluetooth.

Yea, would have been a neat alternative.

But from my perspective, it lacks a couple of the benefits of USB vs. what we had before.

First, data and power in one connector. Sure, some very low-power devices (like mice) could cheat using the DTR or RTS pins of a serial port, but that wasn't nearly enough for a scanner or hard disk.

Second, (mostly) no-nonsense connection of new devices. Yes, installing USB drivers sucked, but plugging in a cable is still easier that going though menus to pair a wireless device[1]. Even in Windows XP, plugging in a thumb drive pretty much "just worked".

From a practical usability point of view, I think WUSB was a regression for a large class of devices: You still needed a power cable for anything not battery powered[2] like a thumb drive or scanner, and you still needed to go through menus to re-mount a device not permanently attached to your computer (like a camera or GPS).

The remaining use-cases, like game controllers, headsets, keyboards and mice were already (barely-)adequately addressed by Bluetooth, which existed before WUSB.

[1] IMHO Apple's "magical" AirPods juuuust barely meet the threshold of "not completely annoying" when switching devices. And that's, what, 15 years after Bluetooth was invented?

[2] Strangely, USB has become the de facto low-voltage DC charging standard. It would be pretty funny to power your WUSB hard disk from a USB port.

These are all really good points, thank you for typing them out.

I agree so much with wireless not "just working". People call me crazy for not digging through bluetooth menus and waiting 30 seconds for something to pair up.

I have two different bluetooth headphones, one for my phone, one for my computer just to not deal with that nonsense.

Re [2] WUSB would have allowed for a single device (powered by one USB connection) to maintain usb connections to multiple devices.

Sure, but you could add that protocol to wired USB if you wanted.

Actually, I think the killer app in 2009 would have been a WUSB ethernet dongle: ~250Mbps internet for a laptop within a single room. That might have been compelling when 802.11g gave you around 25-30 Mbps in practice and roaming between APs worked less well than today.

An alternative to Bluetooth sounds like pointless fragmentation.

Using something like wifi, to allow network infrastructure to help us connect peripherals & devices, would be incredibly cool.

Wifi's higher throughput, range, and lower latency than bluetooth are all appealing.

I for one think the risk here is limiting your imagination to what we have, to the narrow set of available profiles in Bluetooth.

More DIY than this but there is USB-IP[1]. It needs two PCs but given the small size & low power of an rpi0 that is no longer so problematic. Anecdotally, USB-IP seems to not work a bunch of times.

I was hoping MediaTek was serious about their "CrossMount"[2] idea (2015) to allow devices to work across machines, but that never seemed to go anywhere/materialize/become real.

802.11ad / WiGig (in the 60GHz band) had/has a Wireless Bus Extension (WBE) for PCIe and Wireless Serial Extension (WSE) for USB, back in 2011[3]. Wigig is still sort of a product, 10 years latter, occasionally. I'd really really really love for it to have become a little more available.

It's such a pity how USB over USB ports remains the only way to connect things to computers. Especially sad that wireless connectivity is so absent.

[1] http://usbip.sourceforge.net/

[2] https://corp.mediatek.com/news-events/press-releases/mediate...

[3] https://www.electronicdesign.com/technologies/communications...

A few years ago I have found a wireless usb hub.

I did find the idea to be nice and original, but as there was no working driver for Linux, I was never able to test it.

At that time, I did not find any info about it. No wikipedia page. So I was thinking that it was an original orphan chinese gadget as you can find sometimes. A kind of hack trying to abuse of the usb brand. I did not think that it was based on a standard.

Some people should probably redevelop such a concept I think.

Is a wireless repeater for USB not technically possible? There are so many use cases for that I’m always puzzled it hasn’t been done yet.

I don’t think the timing allows this

i have one, so it is possible for usb 2. it also worked quite well and didn't require special driver.

I'm not sure about usb 3.

For USB 2 the maximum round trip from host to device and back is about 1.5 milliseconds.

No idea if that is attainable with a wireless transmitter in between.

Reply on news.ycombinator.com