Nomenclature for USB-connected WiFi devices

Nomenclature for USB-connected WiFi devices

[Alan Stern]

If you've got a WiFi device that connects to the host computer via USB, 
do you refer to it as a "wireless USB device" or as a "USB wireless 
device"?

The second would seem to be more logical, by analogy with things like a 
USB mouse or a USB thumbdrive -- we don't say "mouse USB device" or 
"thumbdrive USB device"!

Furthermore, the first ("Wireless USB") is in fact the name of a defunct 
specification for an Ultra-WideBand interface that would run the USB 
communication protocol over a wireless connection.

Nevertheless there are quite a few places in the kernel source that use 
"wireless USB" where they really mean "USB wireless".  (A few of them 
are gray cases, like "Sierra Wireless USB-to-WWAN", although here the 
word "Wireless" evidently is redundant -- maybe it is part of a brand 
name?)

Would there be any objection to a patch that does a wholesale conversion 
from "wireless USB" to "USB wireless"?

Alan Stern

PS: Similar reasoning applies to the term "wireless PCIe device", 
although here the number of offenders is smaller.

Re: Nomenclature for USB-connected WiFi devices

[Oliver Neukum]

On 23.07.23 18:04, Alan Stern wrote:
> If you've got a WiFi device that connects to the host computer via USB,
> do you refer to it as a "wireless USB device" or as a "USB wireless
> device"?
> 
> The second would seem to be more logical, by analogy with things like a
> USB mouse or a USB thumbdrive -- we don't say "mouse USB device" or
> "thumbdrive USB device"!

If you are doing this, why not go to USB WiFi? We want to know what
kind of wireless device it is.

	Regards
		Oliver

Re: Nomenclature for USB-connected WiFi devices

[Alan Stern]

On Sun, Jul 23, 2023 at 07:05:40PM +0200, Oliver Neukum wrote:
> On 23.07.23 18:04, Alan Stern wrote:
> > If you've got a WiFi device that connects to the host computer via USB,
> > do you refer to it as a "wireless USB device" or as a "USB wireless
> > device"?
> > 
> > The second would seem to be more logical, by analogy with things like a
> > USB mouse or a USB thumbdrive -- we don't say "mouse USB device" or
> > "thumbdrive USB device"!
> 
> If you are doing this, why not go to USB WiFi? We want to know what
> kind of wireless device it is.

Doing that would require more knowledge than I possess.  :-)

I don't know what kinds of wireless device these different drivers are 
meant for.  Presumably most of them are WiFi, but are all of them?

Besides, if the original authors felt it was appropriate to write 
"wireless USB" instead of "WiFi USB", then most likely they will also 
feel it is appropriate to change the text to "USB wireless" instead of 
"USB WiFi".

Alan Stern

Re: Nomenclature for USB-connected WiFi devices

[Greg KH]

On Sun, Jul 23, 2023 at 12:04:50PM -0400, Alan Stern wrote:
> If you've got a WiFi device that connects to the host computer via USB, 
> do you refer to it as a "wireless USB device" or as a "USB wireless 
> device"?

A "USB wifi device", just like the "PCI wifi device" in my desktop.

Or "USB keyboard", or "USB mouse", or "USB thermometer", etc.

> The second would seem to be more logical, by analogy with things like a 
> USB mouse or a USB thumbdrive -- we don't say "mouse USB device" or 
> "thumbdrive USB device"!

Exactly.

> Furthermore, the first ("Wireless USB") is in fact the name of a defunct 
> specification for an Ultra-WideBand interface that would run the USB 
> communication protocol over a wireless connection.

Exactly, let's not get things confused with that obsolete and dead
technology please.

> Nevertheless there are quite a few places in the kernel source that use 
> "wireless USB" where they really mean "USB wireless".  (A few of them 
> are gray cases, like "Sierra Wireless USB-to-WWAN", although here the 
> word "Wireless" evidently is redundant -- maybe it is part of a brand 
> name?)
> 
> Would there be any objection to a patch that does a wholesale conversion 
> from "wireless USB" to "USB wireless"?

I would be glad to take such a patch.

> PS: Similar reasoning applies to the term "wireless PCIe device", 
> although here the number of offenders is smaller.

Hah, I should have read the whole email :)

thanks,

greg k-h